LTC2408C [Linear]

4-/8-Channel 24-Bit uPower No Latency ADCs; 4- / 8通道的24位微功耗无延迟的ADC
LTC2408C
型号: LTC2408C
厂家: Linear    Linear
描述:

4-/8-Channel 24-Bit uPower No Latency ADCs
4- / 8通道的24位微功耗无延迟的ADC

文件: 总36页 (文件大小:354K)
中文:  中文翻译
下载:  下载PDF数据表文档文件
LTC2404/LTC2408  
4-/8-Channel 24-Bit µPower  
No Latency ∆ΣTM ADCs  
U
FEATURES  
DESCRIPTIO  
The LTC®2404/LTC2408 are 4-/8-channel 2.7V to 5.5V  
micropower 24-bit A/D converters with an integrated  
oscillator, 4ppm INL and 0.3ppm RMS noise. They use  
delta-sigma technology and provide single cycle digital  
filter settling time (no latency delay) for multiplexed  
applications. The first conversion after the channel is  
changedisalwaysvalid.ThroughasinglepintheLTC2404/  
LTC2408 can be configured for better than 110dB rejec-  
tion at 50Hz or 60Hz ±2%, or can be driven by an external  
oscillator for a user defined rejection frequency in the  
range 1Hz to 120Hz. The internal oscillator requires no  
external frequency setting components.  
Pin Compatible 4-/8-Channel 24-Bit ADCs  
Single Conversion Digital Filter Settling Time  
Simplifies Multiplexing  
4ppm INL, No Missing Codes  
4ppm Full-Scale Error  
0.5ppm Offset  
0.3ppm Noise  
Internal Oscillator—No External Components  
Required  
110dB Min, 50Hz/60Hz Notch Filter  
Reference Input Voltage: 0.1V to VCC  
Live Zero—Extended Input Range Accommodates  
12.5% Overrange and Underrange  
Single Supply 2.7V to 5.5V Operation  
The converters accept any external reference voltage from  
0.1VtoVCC. Withtheirextendedinputconversionrangeof  
–12.5% VREF to 112.5% VREF the LTC2404/LTC2408  
smoothly resolve the offset and overrange problems of  
preceding sensors or signal conditioning circuits.  
Low Supply Current (200µA) and Auto Shutdown  
U
APPLICATIO S  
Weight Scales  
The LTC2404/LTC2408 communicate through a flexible  
4-wire digital interface which is compatible with SPI and  
MICROWIRETM protocols.  
Direct Temperature Measurement  
Gas Analyzers  
Strain-Gage Transducers  
Instrumentation  
Data Acquisition  
Industrial Process Control  
6-Digit DVMs  
, LTC and LT are registered trademarks of Linear Technology Corporation.  
No Latency ∆Σ is a trademark of Linear Technology Corporation.  
MICROWIRE is a trademark of National Semiconductor Corporation.  
U
TYPICAL APPLICATIO  
Total Unadjusted Error vs Output Code  
0.1V TO V  
2.7V TO 5.5V  
CC  
10  
7
4
3
V
2, 8  
8
6
1µF  
V
MUXOUT  
ADCIN  
REF  
CC  
SERIAL DATA LINK  
MICROWIRE AND  
SPI COMPATABLE  
9
CH0  
4
23  
20  
19  
25  
21  
24  
10 CH1  
11 CH2  
12 CH3  
13 CH4*  
14 CH5*  
15 CH6*  
17 CH7*  
CSADC  
CSMUX  
SCK  
2
ANALOG  
INPUTS  
0
24-BIT  
4-/8-CHANNEL  
MUX  
+
∆∑ ADC  
MPU  
CLK  
–2  
–4  
–6  
–8  
–10  
–0.12V  
TO  
REF  
1.12V  
D
REF  
IN  
SDO  
V
CC  
LTC2404/LTC2408  
= INTERNAL OSC/50Hz REJECTION  
= EXTERNAL CLOCK SOURCE  
= INTERNAL OSC/60Hz REJECTION  
26  
6
COM  
F
O
GND  
0
8,338,608  
OUTPUT CODE (DECIMAL)  
16,777,215  
2404/08 TA01  
1, 5, 16, 18, 22, 27, 28  
*THESE PINS ARE NO CONNECTS ON THE LTC2404  
2404/08 TA02  
1
LTC2404/LTC2408  
W W  
U W  
ABSOLUTE MAXIMUM RATINGS  
(Notes 1, 2)  
Operating Temperature Range  
Supply Voltage (VCC) to GND.......................0.3V to 7V  
Analog Input Voltage to GND ....... 0.3V to (VCC + 0.3V)  
Reference Input Voltage to GND .. 0.3V to (VCC + 0.3V)  
Digital Input Voltage to GND........ 0.3V to (VCC + 0.3V)  
Digital Output Voltage to GND ..... 0.3V to (VCC + 0.3V)  
LTC2404C/LTC2408C .............................. 0°C to 70°C  
LTC2404I/LTC2408I ........................... 40°C to 85°C  
Storage Temperature Range ................. 65°C to 150°C  
Lead Temperature (Soldering, 10 sec).................. 300°C  
U
W U  
PACKAGE/ORDER INFORMATION  
ORDER  
ORDER  
TOP VIEW  
TOP VIEW  
PART NUMBER  
PART NUMBER  
1
2
GND  
GND  
1
2
GND  
GND  
28  
27  
26  
25  
24  
23  
22  
21  
20  
19  
18  
17  
16  
15  
28  
27  
26  
25  
24  
23  
22  
21  
20  
19  
18  
17  
16  
15  
GND  
GND  
V
V
CC  
CC  
LTC2404CG  
LTC2404IG  
LTC2408CG  
LTC2408IG  
3
F
3
F
V
V
REF  
O
O
REF  
4
SCK  
4
SCK  
ADCIN  
GND  
ADCIN  
GND  
5
SDO  
5
SDO  
6
CSADC  
GND  
6
CSADC  
GND  
COM  
COM  
7
7
MUXOUT  
MUXOUT  
8
D
8
D
IN  
V
V
IN  
CC  
CC  
9
CSMUX  
CLK  
9
CSMUX  
CLK  
CH0  
CH1  
CH2  
CH3  
NC  
CH0  
CH1  
CH2  
CH3  
CH4  
CH5  
10  
11  
12  
13  
14  
10  
11  
12  
13  
14  
GND  
NC  
GND  
CH7  
GND  
NC  
GND  
CH6  
NC  
G PACKAGE  
28-LEAD PLASTIC SSOP  
G PACKAGE  
28-LEAD PLASTIC SSOP  
TJMAX = 125°C, θJA = 130°C/W  
TJMAX = 125°C, θJA = 130°C/W  
Consult factory for Military grade parts.  
U
CONVERTER CHARACTERISTICS The denotes specifications which apply over the full operating  
temperature range, otherwise specifications are at TA = 25°C. (Notes 3, 4)  
PARAMETER  
CONDITIONS  
2.5V V V , (Note 5)  
MIN  
TYP  
MAX  
UNITS  
Resolution (No Missing Codes)  
Integral Nonlinearity  
24  
Bits  
REF  
CC  
V
V
= 2.5V (Note 6)  
= 5V (Note 6)  
2
4
10  
15  
ppm of V  
ppm of V  
REF  
REF  
REF  
REF  
Offset Error  
2.5V V V  
0.5  
0.01  
4
2
ppm of V  
REF  
CC  
CC  
CC  
CC  
REF  
Offset Error Drift  
Full-Scale Error  
2.5V V V  
ppm of V /°C  
REF  
REF  
2.5V V V  
10  
ppm of V  
REF  
REF  
Full-Scale Error Drift  
Total Unadjusted Error  
2.5V V V  
0.02  
ppm of V /°C  
REF  
REF  
V
V
= 2.5V  
= 5V  
5
10  
ppm of V  
ppm of V  
REF  
REF  
REF  
REF  
Output Noise  
V
= 0V (Note 13)  
1.5  
µV  
RMS  
IN  
Normal Mode Rejection 60Hz ±2%  
(Note 7)  
110  
130  
dB  
2
LTC2404/LTC2408  
CONVERTER CHARACTERISTICS The denotes specifications which apply over the full operating  
U
temperature range, otherwise specifications are at TA = 25°C. (Notes 3, 4)  
PARAMETER  
CONDITIONS  
MIN  
TYP  
130  
100  
110  
110  
MAX  
UNITS  
dB  
Normal Mode Rejection 50Hz ±2%  
Power Supply Rejection DC  
Power Supply Rejection 60Hz ±2%  
Power Supply Rejection 50Hz ±2%  
(Note 8)  
110  
V
V
V
= 2.5V, V = 0V  
dB  
REF  
REF  
REF  
IN  
= 2.5V, V = 0V, (Note 7)  
dB  
IN  
= 2.5V, V = 0V, (Note 8)  
dB  
IN  
U
U
U
U
A ALOG I PUT A D REFERE CE  
The denotes specifications which apply over the full operating  
temperature range, otherwise specifications are at TA = 25°C. (Note 3)  
SYMBOL  
PARAMETER  
CONDITIONS  
MIN  
0.125 • V  
0.1  
TYP  
MAX  
UNITS  
V
V
V
C
C
Input Voltage Range  
(Note 14)  
1.125 • V  
REF  
IN  
REF  
Reference Voltage Range  
Input Sampling Capacitance  
Reference Sampling Capacitance  
Input Leakage Current  
Reference Leakage Current  
On Channel Leakage Current  
MUX On-Resistance  
V
CC  
V
REF  
10  
15  
1
pF  
S(IN)  
pF  
S(REF)  
IN(LEAK)  
REF(LEAK)  
IN(MUX)  
I
I
I
CS = V  
–10  
12  
10  
nA  
nA  
nA  
CC  
V
= 2.5V, CS = V  
1
12  
REF  
CC  
V = 2.5V (Note 15)  
S
±20  
R
I
I
= 1mA, V = 2.7V  
= 1mA, V = 5V  
250  
120  
300  
250  
ON  
OUT  
OUT  
CC  
CC  
MUX R vs Temperature  
0.5  
20  
%/°C  
%
ON  
R vs V (Note 15)  
ON  
S
I
I
t
t
t
MUX Off Input Leakage  
MUX Off Output Leakage  
MUX Break-Before-Make Interval  
Enable Turn-On Time  
Channel Off, V = 2.5V  
±20  
±20  
nA  
nA  
ns  
S(OFF)  
D(OFF)  
OPEN  
ON  
S
Channel Off, V = 2.5V  
D
125  
290  
490  
190  
70  
V = 1.5V, R = 3.4k, C = 15pF  
ns  
S
L
L
Enable Turn-Off Time  
V = 1.5V, R = 3.4k, C = 15pF  
ns  
OFF  
S
L
L
QIRR  
QINJ  
MUX Off Isolation  
V
= 2V , R = 1k, f = 100kHz  
dB  
pC  
pF  
IN  
P-P  
L
Charge Injection  
R = 0, C = 1000pF, V = 1V  
±1  
S
L
S
C
C
Input Off Capacitance (MUX)  
Output Off Capacitance (MUX)  
10  
S(OFF)  
D(OFF)  
10  
pF  
3
LTC2404/LTC2408  
U
U
DIGITAL I PUTS A D DIGITAL OUTPUTS  
The denotes specifications which apply over the full  
operating temperature range, otherwise specifications are at TA = 25°C. (Note 3)  
SYMBOL  
PARAMETER  
CONDITIONS  
2.7V V 5.5V  
MIN  
TYP  
MAX  
UNITS  
V
V
V
V
High Level Input Voltage  
2.5  
2.0  
V
V
IH  
IL  
IH  
IL  
CC  
CS, F  
2.7V V 3.3V  
O
CC  
Low Level Input Voltage  
CS, F  
4.5V V 5.5V  
0.8  
0.6  
V
V
CC  
2.7V V 5.5V  
O
CC  
High Level Input Voltage  
SCK  
2.7V V 5.5V (Note 9)  
2.5  
2.0  
V
V
CC  
2.7V V 3.3V (Note 9)  
CC  
Low Level Input Voltage  
SCK  
4.5V V 5.5V (Note 9)  
0.8  
0.6  
V
V
CC  
2.7V V 5.5V (Note 9)  
CC  
I
I
Digital Input Current  
0V V V  
CC  
–10  
–10  
10  
µA  
µA  
pF  
pF  
V
IN  
IN  
CS, F  
O
Digital Input Current  
SCK  
0V V V (Note 9)  
10  
IN  
IN  
CC  
C
C
V
V
V
V
Digital Input Capacitance  
10  
10  
IN  
CS, F  
O
Digital Input Capacitance  
SCK  
(Note 9)  
IN  
High Level Output Voltage  
SDO  
I = 800µA  
O
V
V
– 0.5V  
– 0.5V  
OH  
OL  
OH  
OL  
CC  
CC  
Low Level Output Voltage  
SDO  
I = 1.6mA  
O
0.4V  
V
High Level Output Voltage  
SCK  
I = 800µA (Note 10)  
O
V
Low Level Output Voltage  
SCK  
I = 1.6mA (Note 10)  
O
0.4V  
10  
V
I
High-Z Output Leakage  
SDO  
–10  
2
µA  
OZ  
+
V
V
H
MUX High Level Input Voltage  
MUX Low Level Input Voltage  
V = 3V  
V
V
IN MUX  
+
L
V = 2.4V  
0.8  
IN MUX  
W U  
POWER REQUIRE E TS  
The denotes specifications which apply over the full operating temperature range,  
otherwise specifications are at TA = 25°C. (Note 3)  
SYMBOL  
PARAMETER  
Supply Voltage  
Supply Current  
CONDITIONS  
MIN  
TYP  
MAX  
UNITS  
V
CC  
2.7  
5.5  
V
I
I
CC  
Conversion Mode  
Sleep Mode  
CS = 0V (Note 12)  
200  
20  
300  
30  
µA  
µA  
CS = V (Note 12)  
CC  
Multiplexer Supply Current  
All Logic Inputs Tied Together  
15  
40  
µA  
CC(MUX)  
V
IN  
= 0V or 5V  
4
LTC2404/LTC2408  
W U  
TI I G CHARACTERISTICS The denotes specifications which apply over the full operating temperature range,  
otherwise specifications are at TA = 25°C. (Note 3)  
SYMBOL  
PARAMETER  
CONDITIONS  
MIN  
2.56  
0.5  
TYP  
MAX  
307.2  
390  
UNITS  
kHz  
µs  
f
t
t
t
External Oscillator Frequency Range  
External Oscillator High Period  
External Oscillator Low Period  
Conversion Time  
EOSC  
HEO  
0.5  
390  
µs  
LEO  
F = 0V  
130.66  
156.80  
133.33  
160  
EOSC  
136  
163.20  
(in kHz)  
ms  
ms  
ms  
CONV  
O
F = V  
O
CC  
External Oscillator (Note 11)  
20480/f  
f
Internal SCK Frequency  
Internal Oscillator (Note 10)  
External Oscillator (Notes 10, 11)  
19.2  
kHz  
kHz  
ISCK  
f
/8  
EOSC  
D
Internal SCK Duty Cycle  
(Note 10)  
(Note 9)  
(Note 9)  
(Note 9)  
45  
55  
%
kHz  
ns  
ISCK  
f
t
t
t
External SCK Frequency Range  
External SCK Low Period  
2000  
ESCK  
250  
250  
1.64  
LESCK  
External SCK High Period  
ns  
HESCK  
DOUT_ISCK  
Internal SCK 32-Bit Data Output Time  
Internal Oscillator (Notes 10, 12)  
External Oscillator (Notes 10, 11)  
1.67  
1.70  
ms  
ms  
256/f  
(in kHz)  
EOSC  
t
t
t
t
t
t
t
External SCK 32-Bit Data Output Time  
CS to SDO Low Z  
CS to SDO High Z  
CS to SCK ↓  
(Note 9)  
32/f  
(in kHz)  
ms  
ns  
ns  
ns  
ns  
ns  
ns  
ns  
ns  
DOUT_ESCK  
ESCK  
0
0
150  
150  
150  
1
2
(Note 10)  
(Note 9)  
0
3
CS to SCK ↑  
50  
4
SCK to SDO Valid  
SDO Hold After SCK ↓  
SCK Set-Up Before CS ↓  
SCK Hold After CS ↓  
200  
KQMAX  
KQMIN  
(Note 5)  
15  
50  
t
t
5
6
50  
Note 1: Absolute Maximum Ratings are those values beyond which the  
life of the device may be impaired.  
Note 2: All voltage values are with respect to GND.  
Note 10: The converter is in internal SCK mode of operation such that  
the SCK pin is used as digital output. In this mode of operation the  
SCK pin has a total equivalent load capacitance CLOAD = 20pF.  
Note 11: The external oscillator is connected to the FO pin. The external  
oscillator frequency, fEOSC, is expressed in kHz.  
Note 3: VCC = 2.7 to 5.5V unless otherwise specified, source input  
is 0.  
Note 12: The converter uses the internal oscillator.  
Note 4: Internal Conversion Clock source with the FO pin tied  
to GND or to VCC or to external conversion clock source with  
fEOSC = 153600Hz unless otherwise specified.  
FO = 0V or FO = VCC  
.
Note 13: The output noise includes the contribution of the internal  
calibration operations.  
Note 5: Guaranteed by design, not subject to test.  
Note 14: For reference voltage values VREF > 2.5V the extended input  
of 0.125 • VREF to 1.125 • VREF is limited by the absolute maximum  
Note 6: Integral nonlinearity is defined as the deviation of a code from  
a straight line passing through the actual endpoints of the transfer  
curve. The deviation is measured from the center of the quantization  
band.  
Note 7: FO = 0V (internal oscillator) or fEOSC = 153600Hz ±2%  
(external oscillator).  
rating of the Analog Input Voltage pin (Pin 3). For 2.5V < VREF  
0.267V + 0.89 • VCC the input voltage range is 0.3V to 1.125 • VREF  
.
For 0.267V + 0.89 • VCC < VREF VCC the input voltage range is 0.3V  
to VCC + 0.3V.  
Note 15: VS is the voltage applied to a channel input. VD is the voltage  
applied to the MUX output.  
Note 8: FO = VCC (internal oscillator) or fEOSC = 128000Hz ±2%  
(external oscillator).  
Note 9: The converter is in external SCK mode of operation such that  
the SCK pin is used as digital input. The frequency of the clock signal  
driving SCK during the data output is fESCK and is expressed in kHz.  
5
LTC2404/LTC2408  
U W  
TYPICAL PERFOR A CE CHARACTERISTICS  
Total Unadjusted Error  
(3V Supply)  
Negative Extended Input Range  
Total Unadjusted Error (3V Supply)  
INL (3V Supply)  
10  
5
10  
5
10  
5
V
V
= 3V  
= 3V  
V
V
= 3V  
= 3V  
V
V
= 3V  
= 3V  
CC  
REF  
CC  
REF  
CC  
REF  
T
= 90°C  
A
T
= –55°C, –45°C, 25°C, 90°C  
A
T
A
= 25°C  
0
0
0
T
T
= 45°C  
= 55°C  
A
A
T
= –55°C, –45°C, 25°C, 90°C  
A
–5  
–5  
–5  
–10  
–10  
–10  
0
0.5  
1.0  
1.5  
2.0  
2.5  
3.0  
0
0.5  
1.0  
1.5  
2.0  
2.5  
3.0  
0
0.05 0.10 0.15 0.20 0.25 0.30  
INPUT VOLTAGE (V)  
INPUT VOLTAGE (V)  
INPUT VOLTAGE (V)  
24048 G01  
24048 G02  
24048 G03  
Total Unadjusted Error  
(5V Supply)  
Positive Extended Input Range  
Total Unadjusted Error (3V Supply)  
INL (5V Supply)  
10  
8
10  
5
10  
5
V
V
= 5V  
= 5V  
CC  
REF  
V
V
= 3V  
REF  
CC  
= 3V  
6
4
T
A
= 55°C  
2
T
A
= 45°C  
0
0
0
T
= –55°C, –45°C, 25°C, 90°C  
A
–2  
–4  
–6  
–8  
–10  
T
= 90°C  
T = 25°C  
A
A
–5  
–5  
–10  
–10  
3.0  
3.1  
3.2  
3.3  
0
1
2
3
4
5
0
1
2
3
4
5
INPUT VOLTAGE (V)  
INPUT VOLTAGE (V)  
INPUT VOLTAGE (V)  
24048 G04  
24048 G05  
24048 G06  
Positive Extended Input Range  
Total Unadjusted Error (5V Supply)  
Negative Extended Input Range  
Total Unadjusted Error (5V Supply)  
Offset Error vs Reference Voltage  
20  
15  
10  
5
10  
5
10  
5
V
V
= 5V  
= 5V  
V
T
= 5V  
V
V
= 5V  
= 5V  
CC  
REF  
CC  
A
CC  
REF  
= 25°C  
T
= 90°C  
A
T
= 55°C  
= 45°C  
A
T
A
= 25°C  
0
0
T
A
T
= 45°C  
= 55°C  
A
T
A
= 90°C  
T = 25°C  
A
–5  
–5  
T
A
–10  
–10  
0
0
0.05 0.10 0.15 0.20 0.25 0.30  
INPUT VOLTAGE (V)  
5.0  
5.1  
5.2  
5.3  
0
1
2
3
4
5
INPUT VOLTAGE (V)  
REFERENCE VOLTAGE (V)  
24048 G07  
24048 G08  
24048 G10  
6
LTC2404/LTC2408  
U W  
TYPICAL PERFOR A CE CHARACTERISTICS  
RMS Noise vs Reference Voltage  
Offset Error vs VCC  
RMS Noise vs VCC  
5.0  
2.5  
20  
15  
10  
5
5.0  
V
T
= 2.5V  
V
T
= 2.5V  
REF  
A
V
T
= 5V  
REF  
A
CC  
A
= 25°C  
= 25°C  
= 25°C  
0
2.5  
2.5  
0
2.7  
5.0  
0
0
1
2
3
4
5
2.7  
3.2  
3.7  
4.2  
4.7  
5.2  
3.2  
3.7  
4.2  
4.7  
5.2  
V
CC  
V
CC  
REFERENCE VOLTAGE (V)  
24048 G10  
24048 G11  
24048 G12  
RMS Noise vs CODE OUT  
Noise Histogram  
Offset Error vs Temperature  
1500  
1000  
500  
0
5.0  
2.5  
1.00  
0.75  
0.50  
0.25  
0
V
V
V
= 5V  
= 5V  
V
V
V
= 5V  
= 5V  
V
V
V
T
= 5V  
= 5V  
CC  
REF  
IN  
CC  
REF  
IN  
CC  
REF  
IN  
A
= 0V  
= 0V  
= 0.3V TO 5.3V  
= 25°C  
0
2.5  
5.0  
–1.0  
0.5  
0
0.5  
1.0  
1.5  
50 –25  
0
25 50 75 100 125  
0
7FFFFF  
FFFFFF  
TEMPERATURE (°C)  
OUTPUT CODE (ppm)  
CODE OUT (HEX)  
24048 G13  
24048 G15  
24048 G14  
Full-Scale Error  
vs Reference Voltage  
Full-Scale Error vs Temperature  
Full-Scale Error vs VCC  
5.0  
2.5  
6
5
4
3
2
1
0
10.0  
7.5  
5.0  
2.5  
0
V
V
V
= 5V  
REF  
= 5V  
V
V
= 5V  
REF  
CC  
CC  
IN  
= 5V  
= V  
IN  
0
2.5  
V
V
A
= 2.5V  
REF  
IN  
= 2.5V  
T
= 25°C  
5.0  
50 –25  
0
25 50 75 100 125  
2.7  
3.2  
3.7  
4.2  
4.7  
5.2  
0
1
2
3
4
5
TEMPERATURE (°C)  
V
CC  
REFERENCE VOLTAGE (V)  
24048 G16  
24048 G18  
24048 G17  
7
LTC2404/LTC2408  
U W  
TYPICAL PERFOR A CE CHARACTERISTICS  
Sleep Current vs Temperature  
Conversion Current vs Temperature  
PSRR vs Frequency at VCC  
0
230  
30  
25  
20  
15  
V
V
= 4.1V  
= 0V  
CC  
IN  
220  
210  
V
= 5.5V  
= 4.1V  
= 2.7V  
CC  
CC  
–20  
T
= 25°C  
A
O
F
= 0  
V
= 5.5V  
CC  
–40  
–60  
V
200  
190  
180  
170  
160  
V
= 2.7V  
CC  
V
CC  
–80  
–100  
–120  
10  
5
15,360Hz  
10k  
153,600Hz  
1M  
150  
0
1
100  
–25  
0
50  
75 100 125  
50  
100 125  
–50  
25  
–50 –25  
0
25  
75  
FREQUENCY AT V (Hz)  
TEMPERATURE (°C)  
TEMPERATURE (°C)  
CC  
24048 G21  
24048 G19  
24048 G20  
Rejection vs Frequency at VIN  
PSRR vs Frequency at VCC  
PSRR vs Frequency at VCC  
–10  
–30  
0
0
–20  
V
V
T
= 4.1V  
= 0V  
V
V
V
F
= 5V  
V
V
T
= 4.1V  
= 0V  
CC  
IN  
CC  
CC  
IN  
= 5V  
REF  
–20  
= 25°C  
= 2.5V  
= 25°C  
A
0
IN  
A
O
F
= 0  
= 0  
F
= 0  
O
–50  
–70  
–40  
–60  
–40  
–60  
–90  
–110  
–130  
–80  
–100  
–120  
–80  
–100  
–120  
0
50  
100  
150  
200  
250  
1
50  
100  
150  
200  
250  
15200  
15300 15350 15400 15450 15500  
15250  
FREQUENCY AT V (Hz)  
FREQUENCY AT V (Hz)  
FREQUENCY AT V (Hz)  
CC  
CC  
IN  
24048 G22  
24048 G24  
24048 G23  
Rejection vs Frequency at VIN  
Rejection vs Frequency at VIN  
Rejection vs Frequency at VIN  
0
–60  
–70  
0
–20  
–40  
V
V
V
= 5V  
CC  
= 5V  
REF  
–20  
= 2.5V  
IN  
F
= 0  
O
–80  
–40  
–60  
–90  
–60  
–80  
–100  
–110  
–120  
–130  
–140  
–80  
–100  
–120  
–100  
–120  
–140  
SAMPLE RATE = 15.36kHz ±2%  
15100  
15200  
15300  
15400  
15500  
–12  
–8  
–4  
0
4
8
12  
f /2  
S
f
0
S
FREQUENCY AT V (Hz)  
INPUT FREQUENCY DEVIATION FROM NOTCH FREQUENCY (%)  
IN  
INPUT FREQUENCY  
24048 G26  
24048 G25  
24048 F23  
8
LTC2404/LTC2408  
U W  
TYPICAL PERFOR A CE CHARACTERISTICS  
Resolution vs Maximum  
Output Rate  
INL vs Maximum Output Rate  
24  
22  
20  
18  
16  
14  
12  
10  
8
24  
V
V
0
= 5V  
REF  
= EXTERNAL  
F = EXTERNAL  
O
CC  
= 5V  
(20480 × MAXIMUM  
22  
20  
18  
16  
14  
12  
10  
8
F
OUTPUT RATE)  
(20480 × MAXIMUM  
OUTPUT RATE)  
T
T
= 25°C  
= 90°C  
A
A
V
CC  
= V  
REF  
= 5V  
T
A
= 25°C  
V
= V  
= 3V  
CC  
REF  
T
= 90°C  
A
LOG(V /RMS NOISE)  
REF  
*RESOLUTION =  
LOG (2)  
0
5
10 15 20 25 30 35 40 45 50 55 60  
MAXIMUM OUTPUT RATE (Hz)  
24048 G27  
0
5
10 15 20 25 30 35 40 45 50 55 60  
MAXIMUM OUTPUT RATE (Hz)  
24048 G28  
U
U
U
PIN FUNCTIONS  
GND (Pins 1, 5, 16, 18, 22, 27, 28): Ground. Should be  
connected directly to a ground plane through a minimum  
length trace or it should be the single-point-ground in a  
single point grounding system.  
CH4(Pin13):AnalogMultiplexerInput. Noconnectonthe  
LTC2404.  
CH5(Pin14):AnalogMultiplexerInput. Noconnectonthe  
LTC2404.  
VCC (Pins 2, 8): Positive Supply Voltage. 2.7V VCC  
CH6(Pin15):AnalogMultiplexerInput. Noconnectonthe  
LTC2404.  
5.5V. Bypass to GND with a 10µF tantalum capacitor in  
parallel with 0.1µF ceramic capacitor as close to the part  
as possible.  
CH7(Pin17):AnalogMultiplexerInput. Noconnectonthe  
LTC2404.  
VREF (Pin3):ReferenceInput.Thereferencevoltagerange  
is 0.1V to VCC.  
CLK (Pin 19): Shift Clock for Data In. This clock synchro-  
nizes the serial data transfer into the MUX. For normal  
operation, drive this pin in parallel with SCK.  
ADCIN (Pin 4): Analog Input. The input voltage range is  
0.125 • VREF to 1.125 • VREF. For VREF > 2.5V the input  
voltage range may be limited by the pin absolute maxi-  
mum rating of 0.3V to VCC + 0.3V.  
CSMUX (Pin 20): MUX Chip Select Input. A logic high on  
this input allows the MUX to receive a channel address. A  
logic low enables the selected MUX channel and connects  
it to the MUXOUT pin for A/D conversion. For normal  
operation, drive this pin in parallel with CSADC.  
COM(Pin6):SignalGround.Shouldbeconnecteddirectly  
to a ground plane through minimum length trace.  
MUXOUT(Pin7):MUXOutput.Thispinistheoutputofthe  
multiplexer. Tie to ADCIN for normal operation.  
DIN (Pin 21): Digital Data Input. The multiplexer address  
is shifted into this input on the last four rising CLK edges  
before CSMUX goes low.  
CH0 (Pin 9): Analog Multiplexer Input.  
CH1 (Pin 10): Analog Multiplexer Input.  
CH2 (Pin 11): Analog Multiplexer Input.  
CH3 (Pin 12): Analog Multiplexer Input.  
CSADC (Pin 23): ADC Chip Select Input. A low on this pin  
enables the SDO digital output and following each conver-  
sion, the ADC automatically enters the Sleep mode and  
remains in this low power state as long as CSADC is high.  
9
LTC2404/LTC2408  
U
U
U
PIN FUNCTIONS  
A high on this pin also disables the SDO digital output. A  
low-to-high transition on CSADC during the Data Output  
state aborts the data transfer and starts a new conversion.  
Fornormaloperation,drivethispininparallelwithCSMUX.  
isshiftedoutofSDOonthefallingedgeofSCK. Fornormal  
operation, drive this pin in parallel with CLK.  
FO (Pin 26): Digital input which controls the ADC’s notch  
frequencies and conversion time. When the FO pin is  
connectedtoVCC (FO =VCC), theconverterusesitsinternal  
oscillator and the digital filter first null is located at 50Hz.  
When the FO pin is connected to GND (FO = OV), the  
converter uses its internal oscillator and the digital filter  
first null is located at 60Hz. When FO is driven by an  
external clock signal with a frequency fEOSC, the converter  
uses this signal as its clock and the digital filter first null is  
located at a frequency fEOSC/2560. The resulting output  
word rate is fEOSC/20480.  
SDO (Pin 24): Three-State Digital Output. During the data  
output period this pin is used for serial data output. When  
the chip select CSADC is high (CSADC = VCC), the SDO pin  
is in a high impedance state. During the Conversion and  
Sleep periods, this pin can be used as a conversion status  
output. The conversion status can be observed by pulling  
CSADC low.  
SCK(Pin25):ShiftClockforDataOut. Thisclocksynchro-  
nizes the serial data transfer of the ADC data output. Data  
U
U
W
FU CTIO AL BLOCK DIAGRA  
INTERNAL  
OSCILLATOR  
V
CC  
GND  
AUTOCALIBRATION  
AND CONTROL  
F
O
CH0  
CH1  
CH2  
CH3  
CH4  
CH5  
CH6  
CH7  
(INT/EXT)  
SDO  
SERIAL  
INTERFACE  
ADC  
SCK  
COM  
CSADC  
V
REF  
DECIMATING FIR  
CSMUX  
CHANNEL  
SELECT  
DAC  
D
IN  
CLK  
24048 BD  
TEST CIRCUITS  
V
CC  
3.4k  
SDO  
SDO  
3.4k  
C
= 20pF  
LOAD  
C
= 20pF  
LOAD  
HI-Z TO V  
OH  
OH  
V
OL  
V
OH  
TO V  
HI-Z TO V  
TO HI-Z  
24048 TC01  
OL  
V
V
TO V  
OL  
OH  
OL  
TO HI-Z  
24048 TC02  
10  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
Converter Operation Cycle  
CONVERT  
TheLTC2404/LTC2408arelowpower,4-/8-channeldelta-  
sigma analog-to-digital converters with easy-to-use  
4-wire interfaces. Their operation is simple and made up  
of four states. The converter operation begins with the  
conversion, followed by a low power sleep state and  
concluded with the data output (see Figure 1). Channel  
selectionmaybeperformedwhilethedeviceisinthesleep  
state or at the conclusion of the data output state. The  
interface consists of serial data output (SDO), serial clock  
(CLK/SCK), chip select (CSADC/CSMUX) and data input  
(DIN). By tying SCK to CLK and CSADC to CSMUX, the  
interface requires only four wires.  
CHANNEL SELECT  
(SLEEP)  
SLEEP  
CSADC  
AND  
SCK  
1
0
DATA OUTPUT  
(CHANNEL SELECT)  
24048 F01  
Initially, the LTC2404 or LTC2408 performs a conversion.  
Once the conversion is complete, the device enters the  
sleep state. While in the sleep state, power consumption  
is reduced by an order of magnitude. The part remains in  
the sleep state as long as CSADC is logic HIGH. The  
conversion result is held indefinitely in a static shift  
register while the converter is in the sleep state.  
Figure 1. LTC2408 State Transition Diagram  
edge of SCK, see Figure 3. The data output state is  
concluded once 32 bits are read out of the ADC or when  
CSADCisbroughtHIGH.Thedeviceautomaticallyinitiates  
a new conversion and the cycle repeats.  
Channel selection for the next conversion cycle is per-  
formed while the device is in the sleep state or at the end  
of the data output state. A specific channel is selected by  
applyinga4-bitserialwordtotheDIN pinontherisingedge  
ofCLKwhileCSMUXisHIGH,seeFigure3andTable3.The  
channel is selected based on the last four bits clocked into  
the DIN pin before CSMUX goes low. If DIN is all 0’s, the  
previous channel remains selected.  
Through timing control of the CSADC and SCK pins, the  
LTC2404/LTC2408 offer two modes of operation: internal  
or external SCK. These modes do not require program-  
ming configuration registers; moreover, they do not dis-  
turbthecyclicoperationdescribedabove. Thesemodesof  
operation are described in detail in the Serial Interface  
Timing Modes section.  
In the example, Figure 3, the MUX channel is selected  
during the sleep state, just before the data output state  
begins. Once the channel selection is complete, the device  
remains in the sleep state as long as CSADC remains  
HIGH.  
Conversion Clock  
A major advantage delta-sigma converters offer over  
conventional type converters is an on-chip digital filter  
(commonly known as Sinc or Comb filter). For high  
resolution, low frequency applications, this filter is typi-  
cally designed to reject line frequencies of 50 or 60Hz plus  
their harmonics. In order to reject these frequencies in  
excess of 110dB, a highly accurate conversion clock is  
required. The LTC2404/LTC2408 incorporate an on-chip  
highly accurate oscillator. This eliminates the need for  
externalfrequencysettingcomponentssuchascrystalsor  
oscillators.Clockedbytheon-chiposcillator,theLTC2404/  
LTC2408 reject line frequencies (50 or 60Hz ±2%) a  
minimum of 110dB.  
Once CSADC is pulled low, the device begins outputting  
theconversionresult.Thereisnolatencyintheconversion  
result. Since there is no latency, the first conversion  
following a change in input channel is valid and corre-  
sponds to that channel. The data output corresponds to  
theconversionjustperformed. Thisresultisshiftedouton  
the serial data output pin (SDO) under the control of the  
serial clock (SCK). Data is updated on the falling edge of  
SCK allowing the user to reliably latch data on the rising  
11  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
V
CC  
+ 0.3V  
Ease of Use  
9/8V  
The LTC2404/LTC2408 data output has no latency, filter  
settlingorredundantdataassociatedwiththeconversion  
cycle.Thereisaone-to-onecorrespondencebetweenthe  
conversion and the output data. Therefore, multiplexing  
an analog input voltage is easy.  
REF  
V
REF  
ABSOLUTE  
MAXIMUM  
INPUT  
NORMAL  
INPUT  
RANGE  
EXTENDED  
INPUT  
RANGE  
1/2V  
REF  
RANGE  
The LTC2404/LTC2408 perform offset and full-scale cali-  
brations every conversion cycle. This calibration is trans-  
parent to the user and has no effect on the cyclic operation  
described above. The advantage of continuous calibration  
is extreme stability of offset and full-scale readings with  
respect to time, supply voltage change and temperature  
drift.  
0
–1/8V  
REF  
–0.3V  
24048 F02  
Figure 2. LTC2404/LTC2408 Input Range  
The LTC2404/LTC2408 converts input signals within the  
extended input range of 0.125 • VREF to 1.125 • VREF  
.
Power-Up Sequence  
For large values of VREF this range is limited to a voltage  
rangeof0.3Vto(VCC +0.3V).Beyondthisrangetheinput  
ESDprotectiondevicesbegintoturnonandtheerrorsdue  
to the input leakage current increase rapidly.  
The LTC2404/LTC2408 automatically enter an internal  
reset state when the power supply voltage VCC drops  
below approximately 2.2V. When the VCC voltage rises  
above this critical threshold, the converter creates an  
internal power-on-reset (POR) signal with duration of  
approximately 0.5ms. The POR signal clears all internal  
registers within the ADC and initiates a conversion. At  
power-up,themultiplexerchannelisdisabledandshould  
be programmed once the device enters the sleep state.  
TheresultsofthefirstconversionfollowingaPORarenot  
valid since a multiplexer channel was disabled.  
Input signals applied to VIN may extend below ground by  
300mVandaboveVCCby300mV.Inordertolimitanyfault  
current, a resistor of up to 5k may be added in series with  
any channel input pin (CH0 to CH7) without affecting the  
performance of the device. In the physical layout, it is im-  
portanttomaintaintheparasiticcapacitanceoftheconnec-  
tion between this series resistance and the channel input  
pin as low as possible; therefore, the resistor should be  
located as close as practical to the channel input pin. The  
effectoftheseriesresistanceontheconverteraccuracycan  
be evaluated from the curves presented in the Analog In-  
put/Reference Current section. In addition, a series resis-  
torwillintroduceatemperaturedependentoffseterrordue  
to the input leakage current. A 1nA input leakage current  
Reference Voltage Range  
The LTC2404/LTC2408 can accept a reference voltage  
from 0V to VCC. The converter output noise is determined  
by the thermal noise of the front-end circuits, and as such,  
its value in microvolts is nearly constant with reference  
voltage. A decrease in reference voltage will not signifi-  
cantly improve the converter’s effective resolution. On the  
other hand, a reduced reference voltage will improve the  
overall converter INL performance. The recommended  
range for the LTC2404/LTC2408 voltage reference is  
100mV to VCC.  
will develop a 1ppm offset error on a 5k resistor if VREF  
=
5V. This error has a very strong temperature dependency.  
Output Data Format  
TheLTC2404/LTC2408serialoutputdatastreamis32bits  
long. The first 4 bits represent status information indicat-  
ing the sign, input range and conversion state. The next 24  
bits are the conversion result, MSB first. The remaining 4  
bits are sub LSBs beyond the 24-bit level that may be in-  
cludedinaveragingordiscardedwithoutlossofresolution.  
Input Voltage Range  
Theconverterisabletoaccommodatesystemleveloffset  
and gain errors as well as system level overrange  
situations due to its extended input range, see Figure 2.  
12  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
Bit 31 (first output bit) is the end of conversion (EOC)  
indicator. This bit is available at the SDO pin during the  
conversion and sleep states whenever the CSADC pin is  
LOW. This bit is HIGH during the conversion and goes  
LOW when the conversion is complete.  
Bits 3-0 are sub LSBs below the 24-bit level. Bits 3-0 may  
be included in averaging or discarded without loss of  
resolution.  
DataisshiftedoutoftheSDOpinundercontroloftheserial  
clock (SCK), see Figure 3. Whenever CSADC is HIGH, SDO  
remains high impedance and any SCK clock pulses are  
ignored by the internal data out shift register.  
Bit 30 (second output bit) is a dummy bit (DMY) and is  
always LOW.  
Bit 29 (third output bit) is the conversion result sign indi-  
cator (SIG). If VIN is >0, this bit is HIGH. If VIN is <0, this  
bitisLOW.Thesignbitchangesstateduringthezerocode.  
In order to shift the conversion result out of the device,  
CSADC must first be driven LOW. EOC is seen at the SDO  
pinofthedeviceonceCSADCispulledLOW.EOCchanges  
in real time from HIGH to LOW at the completion of a  
conversion. This signal may be used as an interrupt for an  
external microcontroller. Bit 31 (EOC) can be captured on  
the first rising edge of SCK. Bit 30 is shifted out of the  
device on the first falling edge of SCK. The final data bit  
(Bit0)isshiftedoutonthefallingedgeofthe31stSCKand  
may be latched on the rising edge of the 32nd SCK pulse.  
Onthefallingedgeofthe32ndSCKpulse,SDOgoesHIGH  
indicating a new conversion cycle has been initiated. This  
bit serves as EOC (Bit 31) for the next conversion cycle.  
Table 2 summarizes the output data format.  
Bit 28 (forth output bit) is the extended input range (EXR)  
indicator. If the input is within the normal input range  
0 VIN VREF, this bit is LOW. If the input is outside the  
normal input range, VIN > VREF or VIN < 0, this bit is HIGH.  
The function of these bits is summarized in Table 1.  
Table 1. LTC2404/LTC2408 Status Bits  
Bit 31  
EOC  
Bit 30  
DMY  
Bit 29  
SIG  
Bit 28  
EXR  
Input Range  
> V  
V
0
0
0
0
0
0
0
0
1
1
1
0
0
1
IN  
REF  
0 < V V  
IN  
REF  
+
V
V
= 0 /0  
1/0  
0
IN  
IN  
As long as the voltage on the VIN pin is maintained within  
the 0.3V to (VCC + 0.3V) absolute maximum operating  
range, a conversion result is generated for any input value  
from 0.125 • VREF to 1.125 • VREF. For input voltages  
greaterthan1.125VREF,theconversionresultisclamped  
to the value corresponding to 1.125 • VREF. For input  
voltages below 0.125 • VREF, the conversion result is  
< 0  
Bit 27 (fifth output bit) is the most significant bit (MSB).  
Bits 27-4 are the 24-bit conversion result MSB first.  
Bit 4 is the least significant bit (LSB).  
clamped to the value corresponding to 0.125 • VREF  
.
t
CONV  
CSMUX/CSADC  
SDO  
HI-Z  
HI-Z  
EOC  
“0”  
SIG  
EXT  
MSB  
LSB  
BIT 31 BIT 30  
BIT 0  
SCK/CLK  
D
IN  
EN  
D2  
D1  
D0  
DON’T CARE  
24048 F03  
Figure 3. Typical Data Input/Output Timing  
13  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
Table 3. Logic Table for Channel Selection  
Channel Selection  
CHANNEL STATUS  
EN  
0
D2  
X
0
D1  
X
0
D0  
X
0
Typically, CSADC and CSMUX are tied together or CSADC  
is inverted and drives CSMUX. SCK and CLK are tied  
together and driven with a common clock signal. During  
channel selection, CSMUX is HIGH. Data is shifted into the  
DIN pin on the rising edge of CLK, see Figure 3. Table 3  
showsthebit combinations for channelselection. In order  
to enable the multiplexer output, CSMUX must be pulled  
LOW. The multiplexer should be programmed after the  
previousconversioniscomplete.Inordertoguaranteethe  
conversioniscomplete,themultiplexeraddressingshould  
be delayed a minimum tCONV (approximately 133ms for a  
60Hz notch) after the data out is read.  
All Off  
CH0  
1
CH1  
1
0
0
1
CH2  
1
0
1
0
CH3  
1
0
1
1
CH4*  
1
1
0
0
CH5*  
1
1
0
1
CH6*  
1
1
1
0
CH7*  
1
1
1
1
*Not used for the LTC2404.  
Frequency Rejection Selection (FO Pin Connection)  
While the multiplexer is being programmed, the ADC is in  
a low power sleep state. Once the MUX addressing is  
complete, the data from the preceding conversion can be  
read. A new conversion cycle is initiated following the data  
read cycle with the analog input tied to the newly selected  
channel.  
The LTC2404/LTC2408 internal oscillator provides better  
than 110dB normal mode rejection at the line frequency  
and all its harmonics for 50Hz ±2% or 60Hz ±2%. For  
60Hz rejection, FO (Pin 26) should be connected to GND  
(Pin 1) while for 50Hz rejection the FO pin should be  
connected to VCC (Pin 2).  
Table 2. LTC2404/LTC2408 Output Data Format  
Bit 31  
EOC  
Bit 30  
DMY  
Bit 29  
SIG  
Bit 28  
EXR  
Bit 27  
MSB  
Bit 26  
Bit 25  
Bit 24  
Bit 23  
Bit 4  
LSB  
Bit 3-0  
SUB LSBs*  
Input Voltage  
> 9/8 • V  
V
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
0
0
0
0
0
0
0
0
1
1
1
0
0
0
1
1
1
1
0
0
0
0
1
1
1
0
0
0
1
1
0
0
1
1
0
0
1
1
1
0
0
0
1
0
1
0
1
0
1
0
1
1
1
1
1
0
1
0
1
0
1
0
1
0
1
0
0
1
1
0
1
0
1
0
1
0
1
0
1
0
0
...  
...  
...  
...  
...  
...  
...  
...  
...  
...  
...  
...  
...  
...  
1
1
0
1
0
1
0
1
0
1
0
1
0
0
X
X
X
X
X
X
X
X
X
X
X
X
X
X
IN  
REF  
9/8 • V  
1
REF  
V
V
+ 1LSB  
1
REF  
REF  
1
3/4V  
3/4V  
1/2V  
1/2V  
1/4V  
1/4V  
+ 1LSB  
+ 1LSB  
+ 1LSB  
1
REF  
REF  
REF  
REF  
REF  
REF  
1
1
1
1
1
+
0 /0  
1/0**  
–1LSB  
0
0
0
–1/8 • V  
REF  
V
< –1/8 • V  
REF  
IN  
*The sub LSBs are valid conversion results beyond the 24-bit level that may be included in averaging or discarded without loss of resolution.  
**The sign bit changes state during the 0 code.  
14  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
–60  
–70  
The selection of 50Hz or 60Hz rejection can also be made  
by driving FO to an appropriate logic level. A selection  
change during the sleep or data output states will not  
disturb the converter operation. If the selection is made  
during the conversion state, the result of the conversion in  
progress may be outside specifications but the following  
conversions will not be affected.  
–80  
–90  
–100  
–110  
–120  
–130  
–140  
When a fundamental rejection frequency different from  
50Hz or 60Hz is required or when the converter must be  
synchronized with an outside source, the LTC2404/  
LTC2408 can operate with an external conversion clock.  
The converter automatically detects the presence of an  
external clock signal at the FO pin and turns off the internal  
oscillator. The frequency fEOSC of the external signal must  
be at least 2560Hz (1Hz notch frequency) to be detected.  
The external clock signal duty cycle is not significant as  
long as the minimum and maximum specifications for the  
high and low periods tHEO and tLEO are observed.  
–12  
–8  
–4  
0
4
8
12  
INPUT FREQUENCY DEVIATION FROM NOTCH FREQUENCY (%)  
24048 F04  
Figure 4. LTC2404/LTC2408 Normal Mode Rejection When  
Using an External Oscillator of Frequency fEOSC  
LTC2404/LTC2408 operation will not be disturbed if the  
change of conversion clock source occurs during the  
sleep state or during the data output state while the  
converter uses an external serial clock. If the change  
occurs during the conversion state, the result of the  
conversion in progress may be outside specifications but  
the following conversions will not be affected. If the  
change occurs during the data output state and the  
converterisintheInternalSCKmode,theserialclockduty  
cycle may be affected but the serial data stream will  
remain valid.  
While operating with an external conversion clock of a  
frequency fEOSC, the LTC2404/LTC2408 provide better  
than 110dB normal mode rejection in a frequency range  
fEOSC/2560 ±4% and its harmonics. The normal mode  
rejection as a function of the input frequency deviation  
from fEOSC/2560 is shown in Figure 4.  
Whenever an external clock is not present at the F pin the  
O
converter automatically activates its internal oscillator  
Table 4 summarizes the duration of each state as a  
function of FO.  
and enters the Internal Conversion Clock mode. The  
Table 4. LTC2404/LTC2408 State Duration  
State  
Operating Mode  
Duration  
CONVERT  
Internal Oscillator  
F = LOW (60Hz Rejection)  
133ms  
160ms  
20480/f  
O
F = HIGH (50Hz Rejection)  
O
External Oscillator  
F = External Oscillator  
O
(In Seconds)  
EOSC  
with Frequency f  
kHz  
EOSC  
(f  
EOSC  
/2560 Rejection)  
SLEEP  
As Long As CS = HIGH Until CS = 0 and SCK  
DATA OUTPUT  
Internal Serial Clock  
External Serial Clock with  
F = LOW/HIGH  
(Internal Oscillator)  
As Long As CS = LOW But Not Longer Than 1.67ms  
(32 SCK cycles)  
O
F = External Oscillator with  
As Long As CS = LOW But Not Longer Than 256/f  
ms  
EOSC  
O
Frequency f  
kHz  
(32 SCK cycles)  
EOSC  
As Long As CS = LOW But Not Longer Than 32/f ms  
SCK  
Frequency f  
kHz  
(32 SCK cycles)  
SCK  
1
MAXIMUM OUTPUT  
WORD RATE  
OWR =  
inHz  
t
CONVERT + tDATAOUTPUT  
15  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
Using an External Clock for Faster Conversion Times  
The DC specifications are guaranteed for fEOSC up to a  
maximum of 307.2kHz, resulting in a maximum output  
wordrateofapproximately15Hz.However,forfasterrates  
at reduced performance, frequencies up to 1.22MHz can  
be used on the FO pin. Figures 5 and 6 show the INL and  
Resolution vs Output Rate.  
The conversion time of the LTC2404/LTC2408 is deter-  
mined by the conditions on the FO pin. If FO is connected  
to GND for 60Hz rejection, the conversion time is 133µs.  
IfFO isconnectedtoVCC, theconversiontimeis160µs. For  
an externally supplied frequency of fEOSC(kHz), the con-  
version time is:  
SERIAL INTERFACE  
tCONV = 20480/fEOSC (kHz)  
The resulting frequency rejection is:  
Notch Frequency = 8/tCONV  
The LTC2404/LTC2408 transmit the conversion results,  
program the channel selection, and receive the start of  
conversion command through a synchronous 4-wire in-  
terface(SCK=CLK,CSADC=CSMUX).Duringtheconver-  
sion and sleep states, this interface can be used to assess  
the converter status. While in the sleep state this interface  
may be used to program an input channel. During the data  
output state it is used to read the conversion result.  
The maximum output word rate is:  
1
OWR =  
inHz  
t
CONVERT + tDATAOUTPUT  
24  
22  
20  
18  
16  
14  
12  
10  
8
V
V
0
= 5V  
REF  
= EXTERNAL  
CC  
ADC Serial Clock Input/Output (SCK)  
= 5V  
F
(20480 × MAXIMUM  
OUTPUT RATE)  
The serial clock signal present on SCK (Pin 25) is used to  
synchronizethedatatransfer.Eachbitofdataisshiftedout  
the SDO pin on the falling edge of the serial clock.  
T
= 25°C  
A
In the Internal SCK mode of operation, the SCK pin is an  
output and the LTC2404/LTC2408 creates its own serial  
clock by dividing the internal conversion clock by 8. In the  
External SCK mode of operation, the SCK pin is used as  
input. The internal or external SCK mode is selected on  
power-up and then reselected every time a HIGH-to-LOW  
transition is detected at the CSADC pin. If SCK is HIGH or  
floatingatpower-uporduringthistransition,theconverter  
enters the internal SCK mode. If SCK is LOW at power-up  
or during this transition, the converter enters the external  
SCK mode.  
T
= 90°C  
A
0
5
10 15 20 25 30 35 40 45 50 55 60  
MAXIMUM OUTPUT RATE (Hz)  
24048 G27  
Figure 5. INL vs Maximum Output Rate  
24  
F
= EXTERNAL  
O
(20480 × MAXIMUM  
22  
20  
18  
16  
14  
12  
10  
8
OUTPUT RATE)  
T
T
= 25°C  
= 90°C  
A
A
Multiplexer Serial Input Clock (CLK)  
V
CC  
= V  
= 5V  
REF  
Generally, this pin is externally tied to SCK for 4-wire op-  
eration.OntherisingedgeofCLK(Pin19)withCSMUXheld  
HIGH,dataisseriallyshiftedintothemultiplexer.IfCSMUX  
is LOW the CLK input will be disabled and the channel  
selection unchanged.  
V
CC  
= V  
= 3V  
REF  
LOG(V /RMS NOISE)  
REF  
*RESOLUTION =  
LOG (2)  
0
5
10 15 20 25 30 35 40 45 50 55 60  
MAXIMUM OUTPUT RATE (Hz)  
24048 G28  
Serial Data Output (SDO)  
The serial data output pin, SDO (Pin 24), drives the serial  
data during the data output state. In addition, the SDO pin  
Figure 6. Resolution vs Maximum Output Rate  
16  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
is used as an end of conversion indicator during the  
shows the logic table for channel selection. In order to  
select or change a previously programmed channel, an  
enable bit (DIN = 1) must proceed the 3-bit channel select  
serialdata. TheusermaysetDIN =0tocontinuallyconvert  
on the previously selected channel.  
conversion and sleep states.  
When CSADC (Pin 23) is HIGH, the SDO driver is switched  
to a high impedance state. This allows sharing the serial  
interface with other devices. If CSADC is LOW during the  
convert or sleep state, SDO will output EOC. If CSADC is  
LOW during the conversion phase, the EOC bit appears  
HIGH on the SDO pin. Once the conversion is complete,  
EOC goes LOW. The device remains in the sleep state until  
the first rising edge of SCK occurs while CSADC = 0.  
SERIAL INTERFACE TIMING MODES  
The LTC2404/LTC2408’s 4-wire interface is SPI and  
MICROWIRE compatible. This interface offers two modes  
of operation. These include an internal or external serial  
clock. The following sections describe both of these serial  
interface timing modes in detail. For both cases the  
converter can use the internal oscillator (FO = LOW or FO  
= HIGH) or an external oscillator connected to the FO pin.  
Refer to Table 5 for a summary.  
ADC Chip Select Input (CSADC)  
TheactiveLOWchipselect,CSADC(Pin23),isusedtotest  
the conversion status and to enable the data output  
transfer as described in the previous sections.  
In addition, the CSADC signal can be used to trigger a new  
conversion cycle before the entire serial data transfer has  
been completed. The LTC2404/LTC2408 will abort any  
serial data transfer in progress and start a new conversion  
cycle anytime a LOW-to-HIGH transition is detected at the  
CSADC pin after the converter has entered the data output  
state (i.e., after the first rising edge of SCK occurs with  
CSADC = 0).  
External Serial Clock (SPI/MICROWIRE Compatible)  
This timing mode uses an external serial clock (SCK) to  
shift out the conversion result, see Figure 7. This same  
external clock signal drives the CLK pin in order to pro-  
gram the multiplexer. A single CS signal drives both the  
multiplexer CSMUX and converter CSADC inputs. This  
common signal is used to monitor and control the state of  
the conversion as well as enable the channel selection.  
Multiplexer Chip Select (CSMUX)  
The serial clock mode is selected on the falling edge of  
CSADC. To select the external serial clock mode, the serial  
clock pin (SCK) must be LOW during each CSADC falling  
edge.  
For 4-wire operation, this pin is tied directly to CSADC or  
the output of an inverter tied to CSADC. CSMUX (Pin 20)  
is driven HIGH during selection of a multiplexer channel.  
On the falling edge of CSMUX, the selected channel is  
enabled and drives MUXOUT.  
The serial data output pin (SDO) is HI-Z as long as CSADC  
is HIGH. At any time during the conversion cycle, CSADC  
may be pulled LOW in order to monitor the state of the  
converter. While CSADC is LOW, EOC is output to the SDO  
pin. EOC = 1 while a conversion is in progress and EOC =  
0 if the device is in the sleep state. Independent of CSADC,  
the device automatically enters the low power sleep state  
once the conversion is complete.  
Data Input (DIN)  
The data input to the multiplexer, DIN (Pin 21), is used to  
program the multiplexer. The input channel is selected by  
serially shifting a 4-bit input word into the DIN pin under  
the control of the multiplexer clock, CLK. Data is shifted  
into the multiplexer on the rising edge of CLK. Table 3  
Table 5. LTC2404/LTC2408 Interface Timing Modes  
Conversion  
Cycle  
Data  
Connection  
and  
SCK  
Output  
Control  
Configuration  
External SCK  
Internal SCK  
Source  
External  
Internal  
Control  
Waveforms  
CS and SCK  
CS and SCK  
Figures 7, 8, 9  
Figures 10, 11  
CS ↓  
CS ↓  
17  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
2.7V TO 5.5V  
V
CC  
= 50Hz REJECTION  
= EXTERNAL OSCILLATOR  
= 60Hz REJECTION  
V
F
O
CC  
LTC2404/LTC2408  
0.1V  
CC  
V
CSMUX  
CSADC  
SCK  
CS  
REF  
TO V  
–0.12V  
CH0  
TO CH7  
REF  
REF  
TO 1.12V  
SCK  
MUXOUT  
ADCIN  
GND  
CLK  
D
IN  
SDO  
CSADC/  
CSMUX  
SCK/CLK  
SDO  
TEST EOC  
TEST EOC  
BIT31  
BIT30 BIT29 BIT28 BIT27 BIT26  
SIG EXR MSB  
BIT4  
LSB  
BIT0  
TEST EOC  
HI-Z  
SUB  
LSB  
HI-Z  
HI-Z  
D1  
D
DON’T CARE  
EN  
D2  
D0  
DON’T CARE  
IN  
24048 F07  
Figure 7. External Serial Clock Timing Diagram  
While the device is in the sleep state, prior to entering the  
data output state, the user may program the multiplexer.  
As shown in Figure 7, the multiplexer channel is selected  
by serial shifting a 4-bit word into the DIN pin on the rising  
edge of CLK (CLK is tied to SCK). The first bit is an enable  
bit that must be HIGH in order to program a channel. The  
next three bits determine which channel is selected, see  
Table 3. On the falling edge of CSMUX, the new channel is  
selectedandwillbevalidforthefirstconversionperformed  
followingthedataoutputstate.Clocksignalsappliedtothe  
CLK pin while CSMUX is LOW (during the data output  
state) will have no effect on the channel selection. Further-  
more, if DIN is held LOW or CLK is held LOW during the  
sleep state, the channel selection is unchanged.  
At the conclusion of the data cycle, CSADC may remain  
LOW and EOC monitored as an end-of-conversion inter-  
rupt. Alternatively, CSADC may be driven HIGH setting  
SDO to HI-Z. As described above, CSADC may be pulled  
LOWatanytimeinordertomonitortheconversionstatus.  
For each of these operations, CSMUX may be tied to  
CSADC without affecting the selected channel.  
At the conclusion of the data output cycle, the converter  
enters a user transparent calibration cycle prior to actually  
performing a conversion on the selected input channel.  
Thisenablesa66ms(for60Hznotchfrequency)lookahead  
time for the multiplexer input. Following the data output  
cycle, the multiplexer input channel may be selected any  
time in this 66ms window by pulling CSADC HIGH and  
serial shifting data into the DIN pin, see Figure 8.  
When the device is in the sleep state (EOC = 0), its  
conversion result is held in an internal static shift regis-  
ter. The device remains in the sleep state until the first  
rising edge of SCK is seen while CSADC is LOW. Data is  
shifted out the SDO pin on each falling edge of SCK. This  
enables external circuitry to latch the output on the rising  
edge of SCK. EOC can be latched on the first rising edge  
of SCK and the last bit of the conversion result can be  
latchedonthe32ndrisingedgeofSCK.Onthe32ndfalling  
edge of SCK, the device begins a new conversion. SDO  
goesHIGH(EOC=1)indicatingaconversionisinprogress.  
While the device is performing the internal calibration, it is  
sensitive to ground current disturbances. Error currents  
flowing in the ground pin may lead to offset errors. If the  
SCK pin is toggling during the calibration, these ground  
disturbances will occur. The solution is to either drive the  
multiplexer clock input (CLK) separately from the ADC  
clock input (SCK), or program the multiplexer in the first  
1ms following the data output cycle. The remaining 65ms  
may be used to allow the input signal to settle.  
18  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
CSADC/  
CSMUX  
SCK/CLK  
TEST EOC  
TEST EOC  
BIT31 BIT30 BIT29 BIT28BIT27BIT26  
SIG EXR MSB  
BIT4  
LSB  
BIT0  
SUB  
LSB  
SDO  
HI-Z  
D
IN  
DON’T CARE  
EN D2  
D1  
D0 DON’T CARE  
CONVERTER  
STATE  
CONV  
SLEEP  
DATA OUTPUT  
INTERNAL CALIBRATION  
66ms LOOK AHEAD  
CONVERSION ON SELECTED CHANNEL  
66ms CONVERT  
133ms CONVERSION CYCLE (OUTPUT RATE = 7.5Hz)  
24048 F08  
Figure 8. Use of Look Ahead to Program Multiplexer After Data Output  
2.7V TO 5.5V  
V
CC  
= 50Hz REJECTION  
= EXTERNAL OSCILLATOR  
= 60Hz REJECTION  
V
F
O
CC  
LTC2404/LTC2408  
0.1V  
CC  
V
CSMUX  
CSADC  
SCK  
CS  
REF  
TO V  
–0.12V  
CH0  
TO CH7  
REF  
REF  
TO 1.12V  
SCK  
MUXOUT  
ADCIN  
GND  
CLK  
D
IN  
SDO  
CSADC/  
CSMUX  
SCK/CLK  
SDO  
TEST EOC  
TEST EOC  
BIT31  
BIT30 BIT29 BIT28 BIT27 BIT26  
SIG EXR MSB  
BIT9 BIT8  
LSB  
HI-Z  
HI-Z  
D1  
D
IN  
DON’T CARE  
EN  
D2  
D0  
DON’T CARE  
24048 F09  
Figure 9. External Serial Clock with Reduced Data Output Length Timing Diagram  
Typically, CSADC remains LOW during the data output  
state. However, the data output state may be aborted by  
pulling CSADC HIGH anytime between the first rising edge  
and the 32nd falling edge of SCK, see Figure 9. On the  
rising edge of CSADC, the device aborts the data output  
state and immediately initiates a new conversion. This is  
useful for systems not requiring all 32 bits of output data,  
aborting an invalid conversion cycle or synchronizing the  
start of a conversion.  
Internal Serial Clock  
This timing mode uses an internal serial clock to shift out  
the conversion result and program the multiplexer, see  
Figure 10. A CS signal directly drives the CSADC input,  
while the inverse of CS drives the CSMUX input. The CS  
signal is used to monitor and control the state of the  
conversion cycles as well as enable the channel selection.  
The multiplexer is programmed during the data output  
19  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
2.7V TO 5.5V  
V
CC  
= 50Hz REJECTION  
= EXTERNAL OSCILLATOR  
= 60Hz REJECTION  
V
F
O
CC  
LTC2404/LTC2408  
0.1V  
CC  
V
CS  
CSMUX  
REF  
TO V  
10k  
CH0  
TO CH7  
–0.12V  
TO 1.12V  
CSADC  
SCK  
REF  
REF  
MUXOUT  
ADCIN  
GND  
CLK  
D
IN  
SDO  
CSMUX  
CSADC  
SCKCLK  
SDO  
t
EOCtest  
TEST EOC  
TEST EOC  
BIT31 BIT30 BIT29 BIT28 BIT27 BIT26  
SIG EXR MSB  
BIT4 BIT3 BIT2 BIT1 BIT0  
TEST EOC  
SUB SUB SUB SUB  
LSB  
LSB LSB LSB LSB  
HI-Z  
HI-Z  
HI-Z  
D
DON’T CARE  
EN D2  
D1  
D0  
DON’T CARE  
IN  
24048 F10  
Figure 10. Internal Serial Clock Timing Diagram  
state. Theinternalserialclock(SCK)generatedbytheADC  
is applied to the multiplexer clock input (CLK).  
state if CSADC remains LOW. In order to prevent the  
device from exiting the low power sleep state, CSADC  
must be pulled HIGH before the first rising edge of SCK. In  
the internal SCK timing mode, SCK goes HIGH and the  
device begins outputting data at time tEOCtest after the  
falling edge of CSADC (if EOC = 0) or tEOCtest after EOC  
goes LOW (if CSADC is LOW during the falling edge of  
EOC). The value of tEOCtest is 23µs if the device is using its  
internal oscillator (F0 = logic LOW or HIGH). If FO is driven  
by an external oscillator of frequency fEOSC, then tEOCtest is  
3.6/fEOSC. IfCSADCispulledHIGHbeforetimetEOCtest, the  
device remains in the sleep state. The conversion result is  
held in the internal static shift register.  
In order to select the internal serial clock timing mode, the  
serial clock pin (SCK) must be floating (HI-Z) or pulled  
HIGHpriortothefallingedgeofCSADC.Thedevicewillnot  
enter the internal serial clock mode if SCK is driven LOW  
on the falling edge of CSADC. An internal weak pull-up  
resistor is active on the SCK pin during the falling edge of  
CSADC; therefore, the internal serial clock timing mode is  
automatically selected if SCK is not externally driven.  
The serial data output pin (SDO) is HI-Z as long as CSADC  
is HIGH. At any time during the conversion cycle, CSADC  
may be pulled LOW in order to monitor the state of the  
converter. Once CSADC is pulled LOW, SCK goes LOW  
and EOC is output to the SDO pin. EOC = 1 while a  
conversion is in progress and EOC = 0 if the device is in the  
sleep state.  
If CSADC remains LOW longer than tEOCtest, the first rising  
edge of SCK will occur and the conversion result is serially  
shiftedoutoftheSDOpin. Thedataoutputcyclebeginson  
this first rising edge of SCK and concludes after the 32nd  
rising edge. Data is shifted out the SDO pin on each falling  
edgeofSCK.Theinternallygeneratedserialclockisoutput  
WhentestingEOC,iftheconversioniscomplete(EOC=0),  
thedevicewillexitthesleepstateandenterthedataoutput  
20  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
to the SCK pin. This signal may be used to shift the  
conversion result into external circuitry. EOC can be  
latchedonthefirstrisingedgeofSCKandthelastbitofthe  
conversionresultonthe32ndrisingedgeofSCK. Afterthe  
32nd rising edge, SDO goes HIGH (EOC = 1), SCK stays  
HIGH, and a new conversion starts.  
Typically, CSADC remains LOW during the data output  
state. However, the data output state may be aborted by  
pulling CSADC HIGH anytime between the first and 32nd  
rising edge of SCK, see Figure 11. On the rising edge of  
CSADC, the device aborts the data output state and  
immediately initiates a new conversion. This is useful for  
systems not requiring all 32 bits of output data, aborting  
an invalid conversion cycle, or synchronizing the start of  
a conversion. If CSADC is pulled HIGH while the con-  
verter is driving SCK LOW, the internal pull-up is not  
available to restore SCK to a logic HIGH state. This will  
cause the device to exit the internal serial clock mode on  
the next falling edge of CSADC. This can be avoided by  
adding an external 10k pull-up resistor to the SCK pin or  
by never pulling CSADC HIGH when SCK is LOW.  
While operating in the internal serial clock mode, the SCK  
output of the ADC may be used as the multiplexer clock  
(CLK). DIN is latched into the multiplexer on the rising  
edge of CLK. As shown in Figure 10, the multiplexer  
channel is selected by serial shifting a 4-bit word into the  
DIN pin on the rising edge of CLK. The first bit is an enable  
bitwhichmustbeHIGHinordertoprogramachannel.The  
next three bits determine which channel is selected, see  
Table 3. On the rising edge of CSADC (falling edge of  
CSMUX), the new channel is selected and will be valid for  
the next conversion. If DIN is held LOW during the data  
outputstate,thepreviouschannelselectionremainsvalid.  
Whenever SCK is LOW, the LTC2404/LTC2408’s internal  
pull-up at pin SCK is disabled. Normally, SCK is not  
externally driven if the device is in the internal SCK timing  
2.7V TO 5.5V  
V
CC  
= 50Hz REJECTION  
= EXTERNAL OSCILLATOR  
= 60Hz REJECTION  
V
F
O
CC  
LTC2404/LTC2408  
0.1V  
CC  
V
CS  
CSMUX  
REF  
TO V  
10k  
CH0  
TO CH7  
–0.12V  
TO 1.12V  
CSADC  
SCK  
REF  
REF  
MUXOUT  
ADCIN  
GND  
CLK  
D
IN  
SDO  
CSMUX  
CSADC  
SCKCLK  
SDO  
t
EOCtest  
TEST EOC  
TEST EOC  
BIT31 BIT30 BIT29 BIT28 BIT27 BIT26  
SIG EXR MSB  
BIT12 BIT11 BIT10 BIT9 BIT8  
TEST EOC  
HI-Z  
HI-Z  
HI-Z  
D
DON’T CARE  
EN D2  
D1  
D0  
DON’T CARE  
IN  
24048 F11  
Figure 11. Internal Serial Clock with Reduced Data Output Length Timing Diagram  
21  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
mode.However,certainapplicationsmayrequireanexter-  
nal driver on SCK. If this driver goes HI-Z after outputting  
a LOW signal, the LTC2404/LTC2408’s internal pull-up  
remains disabled. Hence, SCK remains LOW. On the next  
falling edge of CSADC, the device is switched to the  
external SCK timing mode. By adding an external 10k pull-  
up resistor to SCK, this pin goes HIGH once the external  
driver goes HI-Z. On the next CSADC falling edge, the  
device will remain in the internal SCK timing mode.  
develop a 2.5µV offset signal. For a reference voltage VREF  
= 2.5V, this represents a 1ppm offset error.  
Inanalternativeconfiguration,theCOMpinoftheconverter  
canbethesingle-point-groundinasinglepointgrounding  
system. The input signal ground, the reference signal  
ground, the digital drivers ground (usually the digital  
ground)andthepowersupplyground(theanalogground)  
should be connected in a star configuration with the com-  
mon point located as close to the COM pin as possible.  
Asimilarsituationmayoccurduringthesleepstatewhen  
CSADC is pulsed HIGH-LOW-HIGH in order to test the  
conversion status. If the device is in the sleep state  
(EOC = 0), SCK will go LOW. Once CSADC goes HIGH  
(within the time period defined above as tEOCtest), the  
internal pull-up is activated. For a heavy capacitive load  
on the SCK pin, the internal pull-up may not be adequate  
to return SCK to a HIGH level before CSADC goes LOW  
again. This is not a concern under normal conditions  
where CSADC remains LOW after detecting EOC = 0. This  
situation is easily avoided by adding an external 10k pull-  
up resistor to the SCK pin.  
The power supply current during the conversion state  
should be kept to a minimum. This is achieved by restrict-  
ing the number of digital signal transitions occurring  
during this period.  
While a digital input signal is in the 0.5V to (VCC – 0.5V)  
range, the CMOS input receiver draws additional current  
from the power supply. It should be noted that, when any  
one of the digital input signals (FO, CSADC, CSMUX, DIN,  
CLK and SCK in External SCK mode of operation) is within  
this range, the LTC2404/LTC2408 power supply current  
mayincreaseevenifthesignalinquestionisatavalidlogic  
level. For micropower operation and in order to minimize  
the potential errors due to additional ground pin current,  
it is recommended to drive all digital input signals to full  
CMOS levels [VIL < 0.4V and VOH > (VCC – 0.4V)].  
DIGITAL SIGNAL LEVELS  
The LTC2404/LTC2408’s digital interface is easy to use.  
Its digital inputs (FO, CSADC, CSMUX, CLK, DIN and SCK  
in External SCK mode of operation) accept standard TTL/  
CMOS logic levels and can tolerate edge rates as slow as  
100µs.However,someconsiderationsarerequiredtotake  
advantageofexceptionalaccuracyandlowsupplycurrent.  
Severe ground pin current disturbances can also occur  
due to the undershoot of fast digital input signals. Under-  
shoot and overshoot can occur because of the imped-  
ance mismatch at the converter pin when the transition  
time of an external control signal is less than twice the  
propagation delay from the driver to LTC2404/LTC2408.  
For reference, on a regular FR-4 board, signal propaga-  
tion velocity is approximately 183ps/inch for internal  
traces and 170ps/inch for surface traces. Thus, a driver  
generating a control signal with a minimum transition  
time of 1ns must be connected to the converter pin  
through a trace shorter than 2.5 inches. This problem  
becomes particularly difficult when shared control lines  
areusedandmultiplereflectionsmayoccur.Thesolution  
is to carefully terminate all transmission lines close to  
their characteristic impedance.  
The digital output signals (SDO and SCK in Internal SCK  
mode of operation) are less of a concern because they are  
not generally active during the conversion state.  
InordertopreservetheaccuracyoftheLTC2404/LTC2408,  
it is very important to minimize the ground path imped-  
ance which may appear in series with the input and/or  
reference signal and to reduce the current which may flow  
through this path. The COM pin (Pin 6) should be con-  
nected to a low resistance ground plane through a mini-  
mum length trace. The use of multiple via holes is recom-  
mended to further reduce the connection resistance. The  
LTC2404/LTC2408’spowersupplycurrentflowingthrough  
the 0.01resistance of the common ground pin will  
Parallel termination near the LTC2404/LTC2408 input  
pins will eliminate this problem but will increase the driver  
22  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
powerdissipation.Aseriesresistorbetween27and56Ω  
placed near the driver or near the LTC2404/LTC2408 pin  
will also eliminate this problem without additional power  
dissipation. The actual resistor value depends upon the  
trace impedance and connection topology.  
performance of the device. It simply results in an offset/  
full-scale shift, see Figure 13. To simplify the analysis of  
input dynamic current, two separate cases are assumed:  
large capacitance at VIN (CIN > 0.01µF) and small capaci-  
tance at VIN (CIN < 0.01µF).  
If the total capacitance at VIN (see Figure 14) is small  
(<0.01µF), relativelylargeexternalsourceresistances(up  
to 20k for 20pF parasitic capacitance) can be tolerated  
withoutanyoffset/full-scaleerror.Figures15and16show  
a family of offset and full-scale error curves for various  
Driving the Input and Reference  
The analog input and reference of the typical delta-sigma  
analog-to-digital converter are applied to a switched ca-  
pacitornetwork.Thisnetworkconsistsofcapacitorsswitch-  
ing between the analog input (ADCIN), COM (Pin 6) and  
the reference (VREF). The result is small current spikes  
seenatbothADCINandVREF. Asimplifiedinputequivalent  
circuit is shown in Figure 12.  
The key to understanding the effects of this dynamic input  
current is based on a simple first order RC time constant  
model. Using the internal oscillator, the internal switched  
capacitor network of the LTC2404/LTC2408 is clocked at  
153,600Hz corresponding to a 6.5µs sampling period.  
Fourteentimeconstantsarerequiredeachtimeacapacitor  
is switched in order to achieve 1ppm settling accuracy.  
TUE  
0
V
/2  
V
REF  
REF  
V
24048 F13  
IN  
Therefore, the equivalent time constant at VIN and VREF  
should be less than 6.5µs/14 = 460ns in order to achieve  
1ppm accuracy.  
Figure 13. Offset/Full-Scale Shift  
R
SOURCE  
CH0 TO  
CH7  
INTPUT  
Input Current (VIN)  
C
PAR  
20pF  
SIGNAL  
C
IN  
LTC2404/  
LTC2408  
SOURCE  
If complete settling occurs on the input, conversion re-  
sultswillbeuneffectedbythedynamicinputcurrent. Ifthe  
settling is incomplete, it does not degrade the linearity  
24048 F14  
Figure 14. An RC Network at CH0 to CH7  
ADCV  
CC  
(PIN 2)  
R
SW  
I
I
REF  
REF  
5k  
V
REF  
MUXV  
ADCV  
CC  
(PIN 2)  
CC  
(PIN 8)  
SELECTED  
CHANNEL  
±I  
AVERAGE INPUT CURRENT:  
DC  
R
75Ω  
R
SW  
SW  
5k  
I
I
I
I
= 0.25(V – 0.5 • V ) • f • C  
IN(MUX)  
IN(MUX)  
IN(LEAK)  
IN(LEAK)  
DC  
IN REF EQ  
CHX  
MUXOUT  
ADCIN  
C
EQ  
I
10pF (TYP)  
R
SW  
5k  
24048 F12  
COM  
f
f
= 50Hz, INTERNAL OSCILLATOR: f = 128kHz  
= 60Hz, INTERNAL OSCILLATOR: f = 153.6kHz  
OUT  
OUT  
EXTERNAL OSCILLATOR: 2.56kHz f 307.2kHz  
Figure 12. LTC2404/LTC2408 Equivalent Analog Input Circuit  
23  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
300  
250  
50  
C
= 1µF  
= 10µF  
V
V
V
= 5V  
= 5V  
IN  
= 25°C  
IN  
CC  
REF  
V
V
V
= 5V  
= 5V  
CC  
REF  
IN  
C
IN  
= 0V  
= 0V  
40  
30  
20  
10  
0
T
A
T
= 25°C  
A
200  
150  
C
= 0.1µF  
IN  
C
= 0pF  
= 100pF  
IN  
C
IN  
C
IN  
= 1000pF  
100  
50  
0
C
= 0.01µF  
C
IN  
= 0.01µF  
IN  
0
100 200 300 400 500 600 700 800 900 1000  
()  
1
10  
100  
1k  
10k  
100k  
R
SOURCE  
R
()  
SOURCE  
24048 F17  
24048 F15  
Figure 15. Offset vs RSOURCE (Small C)  
Figure 17. Offset vs RSOURCE (Large C)  
0
0
V
V
V
T
= 5V  
= 5V  
V
V
V
T
= 5V  
= 5V  
CC  
REF  
IN  
A
CC  
REF  
IN  
C
= 0.01µF  
IN  
= 5V  
= 5V  
–50  
–10  
–20  
–30  
–40  
–50  
= 25°C  
= 25°C  
A
–100  
–150  
C
= 0.1µF  
C
IN  
= 1000pF  
= 0pF  
IN  
IN  
C
IN  
= 100pF  
C
C
IN  
= 1µF  
IN  
C
= 10µF  
C
= 0.01µF  
IN  
–200  
–250  
–300  
1
10  
100  
R
1k  
10k  
100k  
0
200  
400  
R
600  
()  
800  
1000  
()  
SOURCE  
SOURCE  
24048 F16  
24048 F18  
Figure 16. Full-Scale Error vs RSOURCE (Small C)  
Figure 18. Full-Scale Error vs RSOURCE (Large C)  
small valued input capacitors (CIN < 0.01µF) as a function  
capacitance applied to the MUXOUT/ADCIN results in  
linearity errors. The 75on-resistance of the multiplexer  
switch is nonlinear with input voltage. If the capacitance at  
node MUXOUT/ADCIN is less than 0.01µF, the linearity is  
not degraded. On the other hand, excessive capacitance  
(>0.01µF) results in incomplete settling as a function of  
the multiplexer on-resistance. Hence, the nonlinearity of  
the multiplexer switch is seen in the overall transfer  
characteristic.  
of input source resistance.  
For large input capacitor values (CIN > 0.01µF), the input  
spikesareaveragedbythecapacitorintoaDCcurrent. The  
gain shift becomes a linear function of input source  
resistance independent of input capacitance, see Figures  
17 and 18. The equivalent input impedance is 1.66M.  
This results in ±1.5µA of input dynamic current at the  
extreme values of VIN (VIN = 0V and VIN = VREF, when  
VREF = 5V). This corresponds to a 0.3ppm shift in offset  
and full-scale readings for every 1of input source  
resistance.  
In addition to the input current spikes, the input ESD  
protection diodes have a temperature dependent leakage  
current. This leakage current, nominally 1nA (±10nA  
max), resultsinafixedoffsetshiftof10µVfora10ksource  
resistance.  
While large capacitance applied to one of the multiplexer  
channel inputs may result in offset/full-scale shifts, large  
24  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
Reference Current (VREF  
)
constants tied to the reference input. If the capacitance at  
node VREF is small (CVREF < 0.01µF), the reference input  
can tolerate large external resistances without reduction  
in INL, see Figure 21. If the external capacitance is large  
(CVREF > 0.01µF), the linearity will be degraded by  
0.15ppm/independent of capacitance at VREF, see  
Figure 22.  
Similar to the analog input, the reference input has a  
dynamic input current. This current has negligible effect  
on the offset. However, the reference current at VIN = VREF  
is similar to the input current at full-scale. For large values  
of reference capacitance (CVREF > 0.01µF), the full-scale  
error shift is 0.3ppm/of external reference resistance  
independent of the capacitance at VREF, see Figure 19. If  
the capacitance tied to VREF is small (CVREF < 0.01µF), an  
input resistance of up to 20k (20pF parasitic capacitance  
at VREF) may be tolerated, see Figure 20.  
In addition to the dynamic reference current, the VREF ESD  
protection diodes have a temperature dependent leakage  
current.Thisleakagecurrent,nominally1nA(±10nAmax),  
results in a fixed full-scale shift of 10µV for a 10k source  
resistance.  
Unlike the analog input, the integral nonlinearity of the  
device can be degraded with excessive external RC time  
600  
50  
V
V
V
= 5V  
V
V
= 5V  
REF  
T = 25°C  
A
CC  
CC  
= 5V  
= 5V  
REF  
500  
40  
= 5V  
IN  
T
= 25°C  
A
C
VREF  
VREF  
= 0pF  
VREF  
C
= 100pF  
400  
300  
30  
20  
C
= 1000pF  
C
= 10µF  
VREF  
C
= 0.01µF  
VREF  
C
= 1µF  
VREF  
200  
100  
0
10  
0
C
= 0.1µF  
VREF  
C
= 0.01µF  
VREF  
–10  
0
200  
400  
600  
800  
1000  
1
10  
100  
1k  
10k  
()  
100k  
RESISTANCE AT V  
()  
RESISTANCE AT V  
REF  
REF  
24048 F19  
24048 F21  
Figure 19. Full-Scale Error vs RVREF (Large C)  
Figure 21. INL Error vs RVREF (Small C)  
160  
140  
120  
100  
80  
50  
V
V
V
= 5V  
= 5V  
V
V
= 5V  
= 5V  
CC  
REF  
IN  
CC  
REF  
T = 25°C  
A
40  
= 5V  
T
= 25°C  
A
C
= 0.1µF  
= 1µF  
VREF  
C
VREF  
30  
20  
C
VREF  
= 100pF  
= 1000pF  
VREF  
C
C
= 10µF  
VREF  
C
= 0.01µF  
VREF  
60  
10  
C
VREF  
= 0.01µF  
40  
0
C
VREF  
= 0pF  
20  
–10  
–20  
0
–20  
10  
100  
10k  
100k  
0
800  
1000  
1
1k  
200  
400  
600  
RESISTANCE AT V  
()  
RESISTANCE AT V ()  
REF  
REF  
24048 F22  
24048 F20  
Figure 20. Full-Scale Error vs RVREF (Small C)  
Figure 22. INL Error vs RVREF (Large C)  
25  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
ANTIALIASING  
The LTC2408’s Resolution and Accuracy Allows You  
to Measure Points in a Ladder of Sensors  
One of the advantages delta-sigma ADCs offer over con-  
ventional ADCs is on-chip digital filtering. Combined with  
a large oversampling ratio, the LTC2404/LTC2408 signifi-  
cantly simplify antialiasing filter requirements.  
In many industrial processes, for example, cracking tow-  
ers in petroleum refineries, a group of temperature mea-  
surements must be related to one another. A series of  
platinum RTDs that sense slow changing temperatures  
canbeconfiguredintoaresistiveladder,usingtheLTC2408  
to sense each node. This approach allows a single excita-  
tion current passed through the entire ladder, reducing  
total supply current consumption. In addition, this ap-  
proach requires only one high precision resistor, thereby  
reducingcost. Agroupofuptoseventemperaturescanbe  
measured as a group by a single LTC2408 in a loop-pow-  
ered remote acquisition unit. In the example shown in  
Figure 24, the excitation current is 240µA at 0°C. The  
LTC2408 requires 300µA, leaving nearly 3.5mA for the  
remainder of the remote transmitter.  
The digital filter provides very high rejection except at  
integer multiples of the modulator sampling frequency  
(fS), see Figure 23. The modulator sampling frequency is  
256 • FO, where FO is the notch frequency (typically 50Hz  
or 60Hz). The bandwidth of signals not rejected by the  
digitalfilterisnarrow(0.2%)comparedtothebandwidth  
of the frequencies rejected.  
As a result of the oversampling ratio (256) and the digital  
filter, minimal (if any) antialias filtering is required in front  
of the LTC2404/LTC2408. If passive RC components are  
placed in front of the LTC2404/LTC2408, the input dy-  
namic current should be considered (see Input Current  
section). In cases where large effective RC time constants  
are used, an external buffer amplifier may be required to  
minimize the effects of input dynamic current.  
The resistance of any of the RTDs (PT1 to PT7) is deter-  
mined from the voltage across it, as compared to the  
voltage drop across the reference resistor (R1). This is a  
ratiometricimplementationwherethevoltagedropacross  
R1 is given by VREF – VCH1. Channel 7 is used to measure  
the voltage on a representative length of wire. If the same  
type and length of wire is used for all connections, then  
errors associated with the voltage drops across all wiring  
can be removed in software. The contribution of wiring  
drop can be scaled if wire lengths are not equal.  
The modulator contained within the LTC2404/LTC2408  
can handle large-signal level perturbations without satu-  
rating. Signal levels up to 40% of VREF do not saturate the  
analog modulator. These signals are limited by the input  
ESDprotectionto300mVbelowgroundand300mVabove  
VCC.  
Gain can be added to this circuit as the total voltage drop  
across all the RTDs is small compared to ADC full-scale  
range. The maximum recommended gain is 40, as limited  
by both amplifier noise contribution, as well as the maxi-  
mum voltage developed at CH0 when all sensors are at the  
maximum temperature specified for platinum RTDs.  
0
–20  
–40  
–60  
–80  
Adding gain requires that one of the resistors (PT1 to PT7)  
be a precision resistor in order to eliminate the error asso-  
ciated with the gain setting resistors R2 and R3. Note, that  
if a precision (100to 400) resistor is used in place of  
one of the RTDs (PT7 recommended), R1 does not need  
to be a high precision resistor. Although the substitution  
of a precision reference resistor for an RTD to determine  
gain may suggest that R2 and R3 (and R1) need not be  
precise, temperature fluctuations due to airflow may ap-  
pear as noise that cannot be removed in firmware. Conse-  
–100  
–120  
–140  
f /2  
S
f
0
S
INPUT FREQUENCY  
24048 F23  
Figure 23. Sync4 Filter Rejection  
26  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
5V  
300µA  
R2  
6
+
0.1µF  
47µF  
OPTIONAL  
GAIN  
5V  
LTC1634-2.5  
3
2
BLOCK  
7
+
4
5
6
LTC1050  
R1  
20.1k  
0.1%  
4
R3  
R2  
UP TO SEVERAL  
HUNDRED FEET.  
ALL SAME  
5V  
WIRE TYPE  
OPTIONAL  
PROTECTION  
RESISTORS  
5k MAX  
7
4
3
V
2, 8  
1µF  
PT1  
100Ω  
V
MUXOUT  
ADCIN  
REF  
CC  
9
CH0  
PLATINUM  
RTD  
23  
10 CH1  
11 CH2  
12 CH3  
13 CH4  
14 CH5  
15 CH6  
17 CH7  
CSADC  
20  
CSMUX  
25  
24-BIT  
SCK  
PT2  
PT7  
8-CHANNEL  
MUX  
19  
+
∆∑ ADC  
CLK  
21  
D
IN  
24  
SDO  
TO PT3-PT6  
V
CC  
LTC2408  
GND  
26  
6
COM  
F
O
2404/08 F24  
1, 5, 16, 18, 22, 27, 28  
Figure 24. Measuring Up to Seven RTD Temperatures with One Reference Resistor and One Reference Current  
quently, these resistors should be low temperature coef-  
ficient devices. The use of higher resistance RTDs is not  
recommended in this topology, although the inclusion of  
one 1000RTD at the top on the ladder will have minimal  
impact on the lower elements. The same caveat applies to  
fast changing temperatures. Any fast changing sensors  
should be at the top of the ladder.  
characteristicsorsignalinversion/noninversioncouldbe  
selected. The R/2R ladder can be purchased as a network  
to ensure tight temperature tracking. Alternatively, resis-  
tors in a ladder or as separate dividers can be assembled  
from discrete resistors. In the configuration shown, the  
channel resistance of the multiplexer does not contribute  
much to the error budget, as only input op amp current  
flows through the switch. The LTC1050 was chosen for  
its low input current and offset voltage, as well as its  
ability to drive the input of a ∆Σ ADC.  
The LTC2408’s Uncommitted Multiplexer Finds Use in  
a Programmable Gain Scheme  
If the multiplexer in the LTC2408 is not committed to  
channel selection, it can be used to select various signal-  
processing options such as different gains, filters or at-  
tenuator characteristics. In Figure 25, the multiplexer is  
shown selecting different taps on an R/2R ladder in the  
feedback loop of an amplifier. This example allows selec-  
tion of gain from 1 to 128 in binary steps. Other feedback  
networks could be used to provide gains tailored for  
specific purposes. (For example, 1x, 1.1x, 1.41x, 2x,  
2.028x,5x,10x,40x,etc.)Alternatively,differentbandpass  
Insert Gain or Buffering After the Multiplexer  
SeparateMUXOUTandADCINterminalspermitinsertion  
of a gain stage between the MUX and the ADC. If passive  
filtering is used at the input to the ADC, a buffer amplifier  
is strongly recommended to avoid errors resulting from  
thedynamicADCinputcurrent. Ifantialiasingisrequired,  
it should be placed at the input to the MUX. If bandwidth  
limitingisrequiredtoimprovenoiseperformance, afilter  
with a –3dB point at 1500Hz will reduce the effective total  
27  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
5V  
3
AV = 1, 2, 4...128  
6
V
IN  
+
LTC1050  
2
5V  
0.1V TO V  
3
CC  
7
4
2, 8  
V
CC  
1µF  
10k  
10k  
10k  
10k  
10k  
10k  
10k  
10k  
2
V
REF  
MUXOUT  
ADCIN  
9
CH0  
20k  
20k  
20k  
20k  
20k  
20k  
20k  
23  
20  
25  
19  
21  
24  
10 CH1  
11 CH2  
12 CH3  
13 CH4  
14 CH5  
15 CH6  
17 CH7  
4
CSADC  
CSMUX  
SCK  
8
24-BIT  
∆∑ ADC  
8-CHANNEL  
MUX  
+
CLK  
16  
32  
64  
128  
D
IN  
SDO  
V
CC  
LTC2408  
GND  
26  
6
COM  
F
O
2404/08 F25  
1, 5, 16, 18, 22, 27, 28  
Figure 25. Using the Multiplexer to Produce Programmable Gains of 1 to 128  
5V  
OPTIONAL  
BANDWIDTH  
3
2
7
LIMIT  
+
6
LTC1050  
MAY BE REQUIRED BY OTHER  
AMPLIFIERS (IS REQUIRED BY  
BIPOLAR AMPLIFIERS)  
C1  
0.022µF  
R4  
5K  
4
R3  
200k  
R1  
5.1k  
R2  
5.1K  
C2  
OPTIONAL GAIN  
AND ROLL-OFF  
5V  
7
4
3
V
2, 8  
V
CC  
10µF  
MUXOUT  
ADCIN  
REF  
9
CH0  
23  
20  
25  
19  
21  
24  
10 CH1  
11 CH2  
12 CH3  
13 CH4  
14 CH5  
15 CH6  
17 CH7  
CSADC  
CSMUX  
SCK  
24-BIT  
8-CHANNEL  
MUX  
ANALOG  
INPUTS  
+
∆∑ ADC  
CLK  
D
IN  
SDO  
V
CC  
LTC2408  
GND  
26  
6
COM  
F
O
2404/08 F26  
1, 5, 16, 18, 22, 27, 28  
Figure 26. Inserting Gain Between the Multiplexer and the ADC Input  
28  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
noise bandwidth of the system to 6Hz. The noise band-  
width of the LTC2408 without any input bandwidth lim-  
iting is approximately 150Hz. A roll-off at 1500Hz  
eliminates all higher order images of the base bandwidth  
of 6Hz. In the example shown, the optional bandwidth-  
limitingfilterhasa–3dBpointat1450Hz. Thisfiltercanbe  
inserted after the multiplexer provided that higher source  
impedance prior to the multiplexer does not reduce the  
3dB frequency, extending settling time, and resulting in  
charge sharing between samples. The settling time of this  
filter to 20+ bits of accuracy is less than 2ms. In the pres-  
ence of external wideband noise, this filter reduces the  
apparent noise by a factor of 5. Note that the noise band-  
width for noise developed in the amplifier is 150Hz. In the  
example shown, the gain of the amplifier is set to 40, the  
pointatwhichamplifiernoisegaindominatestheLTC2408  
noise. Input voltage range as shown is then 0V to 125mV  
DC. The recommended capacitor at C2 for a gain of 40  
would be 560pF.  
T
hecodebeginsbydeclaringvariablesandallocatingfour  
memory locations to store the 32-bit conversion result  
andafifthlocationtostoretheMUXchanneladdress.This  
isfollowedbyinitializingPORTDsSPIconfiguration. The  
program then enters the main sequence. It begins by  
sending the MUX channel data. It then activates the  
LTC2408’s serial interface by setting the SS output low,  
sendingalogiclowtoCSADC/CSMUX. Thisalsoactivates  
theselectedMUXchannel.Itnextwaitsinaloopforalogic  
low on the data line, signifying end-of-conversion. After  
the loop is satisfied, four SPI transfers are completed,  
retrieving the conversion. The main sequence ends by  
settingSShigh.ThisplacestheLTC2408’sserialinterface  
in a high impedance state and initiates another conver-  
sion. The program in Figure 30 modifies the MUX channel  
selection routine in Figure 28’s listing for selection of 16  
channels. Figure 29 shows the connections between the  
LTC1391, LTC2408 and the 68HC11 controller.  
Interfacing the LTC2404/LTC2408 to the 68HC11  
Microcontroller  
19  
CLK  
SCK  
SD0  
CSADC  
CSMUX  
68HC11  
SCK (PD4)  
MISO (PD2)  
SS (PD5)  
25  
24  
23  
20  
21  
LTC2408  
The listing in Figure 28 is a simple assembler routine for  
the 68HC11 microcontroller. It uses PORT D, configuring  
it for SPI data transfer between the controller and the  
LTC2408. The program shows how to select and enable a  
MUXchannelandretrieveconversiondata.Figure27shows  
the simple 4-wire SPI connection.  
MOSI (PD3)  
D
IN  
24048 F27  
Figure 27. Connecting the LTC2408 to a  
68HC11 MCU Using the SPI Serial Interface  
**********************************************************  
*
*
*
* This example program loads multiplexer channels selection data into  
* the LTC2408’s internal MUX and then transfers the LTC2408’s 32-bit *  
* output conversion result to four consecutive 8-bit memory locations. *  
*
*
**********************************************************  
*
***************************************  
* 68HC11 register definitions  
*
***************************************  
*
PORTD  
*
DDRD  
SPCR  
*
SPSR  
*
SPDR  
*
EQU  
$1008  
Port D data register  
“ - , - , SS* ,CSK ;MOSI,MISO,TxD ,RxD “  
Port D data direction register  
SPI control register  
“SPIE,SPE ,DWOM,MSTR;SPOL,CPHA,SPR1,SPR0”  
SPI status register  
“SPIF,WCOL, - ,MODF; - , - , - , - “  
SPI data register; Read-Buffer; Write-Shifter  
EQU  
EQU  
$1009  
$1028  
EQU  
EQU  
$1029  
$102A  
* RAM variables to hold the LTC2408’s 32 conversion result  
*
29  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
DIN1  
DIN2  
DIN3  
DIN4  
MUX  
*
EQU  
EQU  
EQU  
EQU  
EQU  
$00  
$01  
$02  
$03  
$04  
This memory location holds the LTC2408’s bits 31 - 24  
This memory location holds the LTC2408’s bits 23 - 16  
This memory location holds the LTC2408’s bits 15 - 08  
This memory location holds the LTC2408’s bits 07 - 00  
This memory location holds the MUX address data  
***************************************  
* Start GETDATA Routine  
*
***************************************  
*
ORG  
LDS  
$C000  
$CFFF  
#$2F  
Program start location  
Top of C page RAM, beginning location of stack  
-,-,1,0;1,1,1,1  
-, -, SS*-Hi, SCK-Lo, MOSI-Hi, MISO-Hi, X, X  
Keeps SS* a logic high when DDRD, bit 5 is set  
-,-,1,1;1,0,0,0  
SS* , SCK, MOSI are configured as Outputs  
MISO, TxD, RxD are configured as Inputs  
*
INIT1  
*
LDAA  
STAA  
LDAA  
STAA  
PORTD  
#$38  
DDRD  
*
* DDRD’s bit 5 is a 1 so that port D’s SS* pin is a general output  
LDAA  
STAA  
#$50  
SPCR  
The SPI is configured as Master, CPHA = 0, CPOL = 0  
and the clock rate is E/2  
(This assumes an E-Clock frequency of 4MHz. For higher  
E-Clock frequencies, change the above value of $50 to a  
value that ensures the SCK frequency is 2MHz or less.)  
*
*
*
*
GETDATA PSHX  
PSHY  
PSHA  
LDX  
#$0  
The X register is used as a pointer to the memory  
locations that hold the conversion data  
*
LDY  
#$1000  
*
*******************************  
* The next routine sends data to the  
* LTC2408 an sets its MUX channel  
*
*
*******************************  
*
LDAA  
ORAA  
STAA  
$MUX  
#$08  
Retrieve MUX address  
Set the MUX’s ENABLE bit  
SPDR  
Transfer Accum. A contents to SPI register to initiate  
serial transfer  
Get SPI transfer status  
WAITMUX If the transfer is not finished, read status  
*
WAITMUX LDAA  
SPSR  
BPL  
*
***************************************  
* Enable the LTC2408  
*
***************************************  
*
BCLR  
PORTD,Y %00100000 This sets the SS* output bit to a logic  
low, selecting the LTC2408  
*
*
***************************************  
* The next short loop waits for the  
* LTC2408’s conversion to finish before  
* starting the SPI data transfer  
*
*
*
***************************************  
*
CONVEND LDAA  
PORTD  
Retrieve the contents of port D  
Look at bit 2  
Bit 2 = Hi; the LTC2408’s conversion is not  
complete  
Bit 2 = Lo; the LTC2408’s conversion is complete  
Branch to the loop’s beginning while bit 2 remains  
high  
ANDA  
#%00000100  
*
*
*
BNE  
CONVEND  
*
30  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
*
***************************************  
* The SPI data transfer  
*
***************************************  
*
TRFLP1  
LDAA  
STAA  
#$0  
Load accumulator A with a null byte for SPI transfer  
This writes the byte into the SPI data register and  
starts the transfer  
SPDR  
*
WAIT1  
*
LDAA  
BPL  
SPSR  
This loop waits for the SPI to complete a serial  
transfer/exchange by reading the SPI Status Register  
WAIT1  
The SPIF (SPI transfer complete flag) bit is the SPSR’s  
*
*
MSB and is set to one at the end of an SPI transfer. The  
branch will occur while SPIF is a zero.  
Load accumulator A with the current byte of LTC2408 data  
that was just received  
Transfer the LTC2408’s data to memory  
Increment the pointer  
LDAA  
SPDR  
0,X  
*
STAA  
INX  
CPX  
BNE  
#DIN4+1 Has the last byte been transferred/exchanged?  
TRFLP1 If the last byte has not been reached, then proceed to  
the next byte for transfer/exchage  
*
*
BSET  
PORTD,Y %00100000 This sets the SS* output bit to a logic  
high, de-selecting the LTC2408  
Restore the A register  
PULA  
PULY  
PULX  
RTS  
Restore the Y register  
Restore the X register  
Figure 28. LTC2408-68HC11 MCU Digital Interface Routine  
5V  
7
4
3
V
2, 8  
V
CC  
10µF  
ADCIN  
MUXOUT  
REF  
9
TO  
17  
CH0 TO  
CH7  
23  
20  
25  
19  
21  
24  
CSADC  
CSMUX  
SCK  
68HC11  
SS (PD5)  
24-BIT  
+
∆∑ ADC  
CLK  
SCK (PD4)  
MOSI (PD3)  
MISO (PD2)  
LTC1391  
D
IN  
1
2
3
4
5
6
7
8
15  
CH8  
CH9  
S0  
S1  
S2  
S3  
S4  
S5  
S6  
S7  
D
SDO  
V
CC  
LTC2408  
GND  
CH10  
CH11  
CH12  
CH13  
CH14  
CH15  
26  
6
COM  
F
O
10  
11  
13  
12  
CLK  
CS  
1, 5, 16, 18, 22, 27, 28  
D
OUT  
2404/08 F29  
D
IN  
Figure 29. Combining the LTC2408 with the LTC1391 for 16 Input Channels  
31  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
*****************************************************************************  
*
*
*
*
*
*
* This example program loads multiplexer channels selection data into  
* either the LTC2408’s internal MUX or an external LTC1391 MUX. It then  
* transfers the LTC2408’s 32-bit output conversion result to four  
* consecutive 8-bit memory locations.  
*
*
*****************************************************************************  
*
***************************************  
* 68HC11 register definitions  
*
***************************************  
*
PORTD  
*
DDRD  
SPCR  
*
SPSR  
*
SPDR  
*
EQU  
$1008  
Port D data register  
“ - , - , SS* ,CSK ;MOSI,MISO,TxD ,RxD “  
Port D data direction register  
SPI control register  
“SPIE,SPE ,DWOM,MSTR;SPOL,CPHA,SPR1,SPR0”  
SPI status register  
“SPIF,WCOL, - ,MODF; - , - , - , - “  
SPI data register; Read-Buffer; Write-Shifter  
EQU  
EQU  
$1009  
$1028  
EQU  
EQU  
$1029  
$102A  
* RAM variables to hold the LTC2408’s 32 conversion result  
*
DIN1  
DIN2  
DIN3  
DIN4  
MUX  
*
EQU  
EQU  
EQU  
EQU  
EQU  
$00  
$01  
$02  
$03  
$04  
This memory location holds the LTC2408’s bits 31 - 24  
This memory location holds the LTC2408’s bits 23 - 16  
This memory location holds the LTC2408’s bits 15 - 08  
This memory location holds the LTC2408’s bits 07 - 00  
This memory location holds the MUX address data  
***************************************  
* Start GETDATA Routine  
*
***************************************  
*
ORG  
$C000  
#$2F  
Program start location  
INIT1  
*
LDAA  
-,-,1,0;1,1,1,1  
-, -, SS*-Hi, SCK-Lo, MOSI-Hi, MISO-Hi, X, X  
Keeps SS* a logic high when DDRD, bit 5 is set  
-,-,1,1;1,0,0,0  
SS* , SCK, MOSI are configured as Outputs  
MISO, TxD, RxD are configured as Inputs  
STAA  
LDAA  
STAA  
PORTD  
#$38  
DDRD  
*
* DDRD’s bit 5 is a 1 so that port D’s SS* pin is a general output  
LDAA  
STAA  
#$50  
SPCR  
The SPI is configured as Master, CPHA = 0, CPOL = 0  
and the clock rate is E/2  
(This assumes an E-Clock frequency of 4MHz. For higher  
E-Clock frequencies, change the above value of $50 to a  
value that ensures the SCK frequency is 2MHz or less.)  
*
*
*
*
GETDATA PSHX  
PSHY  
PSHA  
LDX  
#$0  
The X register is used as a pointer to the memory  
locations that hold the conversion data  
*
LDY  
#$1000  
*
***************************************  
* The next routine sends data to the  
* LTC2408 an sets its MUX channel  
*
*
***************************************  
*
LDAA  
TAB  
SUBA  
BLE  
TBA  
ORAA  
BRA  
MUX  
Retrieve MUX address  
Save contents of Accum. A  
Is the MUX address in the low nibble  
#$07  
ENLWMX If it is, branch to enable the LTC2408’s internal MUX  
Restore contents of Accum. A  
#$80  
Enable the LTC1391 external MUX  
MUXSPI Go to SPI transfer2400  
32  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
ENLWMX TBA  
ORAA  
Restore contents of Accum. A  
Set the MUX’s ENABLE bit  
Transfer Accum. A contents to SPI register to initiate  
serial transfer  
Get SPI transfer status  
WAITMUX If the transfer is not finished, read status  
#$08  
MUXSPI STAA  
*
SPDR  
WAITMUX LDAA  
SPSR  
BPL  
*
***************************************  
* Enable the LTC2408  
*
***************************************  
*
BCLR  
PORTD,Y %00100000 This sets the SS* output bit to a logic  
low, selecting the LTC2408  
*
*
***************************************  
* The next short loop waits for the  
* LTC2408’s conversion to finish before  
* starting the SPI data transfer  
*
*
*
***************************************  
*
CONVEND LDAA  
PORTD  
Retrieve the contents of port D  
ANDA  
#%00000100  
Look at bit 2  
Bit 2 = Hi; the LTC2408’s conversion is not  
complete  
Bit 2 = Lo; the LTC2408’s conversion is complete  
Branch to the loop’s beginning while bit 2 remains  
high  
*
*
*
BNE  
CONVEND  
*
*
***************************************  
* The SPI data transfer  
*
***************************************  
*
TRFLP1  
LDAA  
STAA  
#$0  
Load accumulator A with a null byte for SPI transfer  
This writes the byte into the SPI data register and  
starts the transfer  
This loop waits for the SPI to complete a serial  
transfer/exchange by reading the SPI Status Register  
WAIT1 The SPIF (SPI transfer complete flag) bit is the SPSR’s  
MSB and is set to one at the end of an SPI transfer. The  
branch will occur while SPIF is a zero.  
Load accumulator A with the current byte of LTC2408 data  
that was just received  
Transfer the LTC2408’s data to memory  
Increment the pointer  
SPDR  
*
WAIT1  
*
LDAA  
BPL  
SPSR  
*
*
LDAA  
SPDR  
0,X  
*
STAA  
INX  
CPX  
BNE  
#DIN4+1 Has the last byte been transferred/exchanged?  
TRFLP1 If the last byte has not been reached, then proceed to  
the next byte for transfer/exchage  
*
*
BSET  
PORTD,Y %00100000 This sets the SS* output bit to a logic  
high, de-selecting the LTC2408  
Restore the A register  
PULA  
PULY  
PULX  
RTS  
Restore the Y register  
Restore the X register  
Figure 30. LTC2408/LTC1391-684C11 MCU Digital Interface Routine  
An 8-Channel DC-to-Daylight Digitizer  
type, is located some distance from the ADC or operates  
in a high ambient noise environment, the LTC2408’s low  
power dissipation allows circuit operation in close prox-  
imity to the sensor. As a result, conditioning the sensor  
output can be greatly simplified through the use of single-  
ended arrangements. In those applications where differ-  
ential signal conditioning is required, chopper  
ThecircuitinFigure31showsanexampleoftheLTC2408’s  
flexibility in digitizing a number of real-world physical  
phenomena—from DC voltages to ultraviolet light. All of  
the examples implement single-ended signal condition-  
ing. Althoughdifferentialsignalconditioningisapreferred  
approach in applications where the sensor is a bridge-  
33  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
amplifier-based or self-contained instrumentation ampli-  
fiers (also available from LTC) can be used with the  
LTC2408.  
Two channels (CH3 and CH4) of the LTC2408 are used to  
accommodate a 3-wire 100, Pt RTD in a unique circuit  
that allows true RMS/RF signal power measurement from  
audio to gigahertz (GHz) frequencies. The unique feature  
ofthiscircuitisthatthesignalpowerdissipatedinthe50Ω  
termination in the form of heat is measured by the 100Ω  
RTD. Two readings are required to compensate for the  
RTD’s lead-wire resistance. The reading on CH4 is multi-  
plied by 2 and subtracted from the reading on CH3 to  
determine the exact value of the RTD.  
With the resistor network connected to CH0, the LTC2408  
isabletomeasureDCvoltagesfrom1mVto1kVinasingle  
range without the need for autoranging. The 990k resistor  
should be a 1W resistor rated for high voltage operation.  
Alternatively, the 990k resistor can be replaced with a  
seriesconnectionofseverallowercost,lowerpowermetal  
film resistors.  
While the LTC2408 is capable of measuring signals over a  
range of six decades, the implementation (mechanical,  
electrical and thermal) of this technique ultimately deter-  
mines the performance of the circuit. The thermal resis-  
tanceoftheassembly(the50/RTDmasstoitsenclosure)  
will determine the sensitivity of the circuit. The dynamic  
range of the circuit will be determined by the maximum  
temperature the assembly is rated to withstand, approxi-  
mately 850°C. Details of the implementation are quite  
involved and are beyond the scope of this document.  
Please contact LTC directly for a more comprehensive  
treatment of this implementation.  
The circuit connected to CH1 shows an LT1793 FET input  
operational amplifier used as an electrometer for high  
impedance, low frequency applications such as measur-  
ing pH. The circuit has been configured for a gain of 21;  
thus, the input signal range is –15mV VIN 250mV. An  
amplifier circuit is necessary in these applications be-  
cause high output impedance sensors cannot drive  
switched-capacitor ADCs directly. The LT1793 was cho-  
sen for its low input bias current (10pA, max) and low  
noise (8nV/Hz) performance. As shown, the use of a  
driven guard (and TeflonTM standoffs) is recommended in  
high impedance sensor applications; otherwise, PC board  
surface leakage current effects can degrade results.  
In the circuit connected to the LTC2408’s CH5 input, a  
thermistor is configured in a half-bridge arrangement that  
could be used to measure the case temperature of the  
RTD-basedthermalpowermeasurementschemedescribed  
previously. In general, thermistors yield very good resolu-  
tion over a limited temperature range. Measurement reso-  
lution of 0.001°C is possible; however, thermistor  
self-heating effects, thermistor initial tolerance and circuit  
thermal construction can combine to limit achievable  
resolution. For the half-bridge arrangement shown, the  
LTC2408canmeasuretemperaturechangesover5orders  
of magnitude.  
The circuit connected to CH2 illustrates a precision half-  
wave rectifier that uses the LTC2408’s internal ∆Σ ADC as  
an integrator. This circuit can be used to measure 60Hz,  
120Hz or from 400Hz to 1kHz with good results. The  
LTC2408’s internal sinc4 filter effectively eliminates any  
frequency in this range. Above 1kHz, limited amplifier  
gain-bandwidthproductandtransientovershootbehavior  
can combine to degrade performance. The circuit’s dy-  
namicrangeislimitedbyoperationalamplifierinputoffset  
voltage and the system’s overall noise floor. Using an  
LTC1050 chopper-stabilized operational amplifier with a  
VOS of 5µV, the dynamic range of this application covers  
approximately 5 orders of magnitude. The circuit configu-  
ration is best implemented with a precision, 3-terminal,  
2-resistor 10k network (for example, an IRC PFC-D net-  
work) for R6 and R7 to maintain gain and temperature  
stability. Alternatively, discrete resistors with 0.1% initial  
tolerance and 5ppm/°C temperature coefficient would  
also be adequate for most applications.  
Connected to the LTC2408’s CH6 input, an infrared ther-  
mocouple (Omega Engineering OS36-1) can be used in  
limited range, noncontact temperature measurement ap-  
plications or applications where high levels of infrared  
light must be measured. Given the LTC2408’s 0.3ppmRMS  
noise performance, measurement resolution using infra-  
red thermocouples is approximately 0.03°C—equivalent  
to the resolution of a conventional Type J thermocouple.  
Teflon is a trademark of Dupont Company.  
34  
LTC2404/LTC2408  
U
W U U  
APPLICATIONS INFORMATION  
These infrared thermocouples are self-contained: 1) they  
do not require external cold junction compensation; 2)  
they cannot use conventional open thermocouple detec-  
tion schemes; and 3) their output impedances are high,  
approximately 3k. Alternatively, conventional thermo-  
couples can be connected directly to the LTC2408 (not  
shown) and cold junction compensation can be provided  
byanexternaltemperaturesensorconnectedtoadifferent  
channel(seethethermistorcircuitonCH5)orbyusingthe  
LT1025, a monolithic cold-junction compensator IC.  
The photodiode chosen (Hammatsu S1336-5BK) pro-  
duces an output of 500mA per watt of optical illumina-  
tion. The output of the photodiode is dependent on two  
factors: active detector area (2.4mm • 2.4mm) and  
illumination intensity. With the 5k resistor, optical inten-  
sities up to 368W/m2 at 960nM (direct sunlight is ap-  
proximately1000W/m2)canbemeasuredbytheLTC2408.  
With a resolution of 300pA, the optical dynamic range  
covers 6 orders of magnitude.  
The application circuits shown connected to the LTC2408  
demonstrate the mix-and-match capabilities of this multi-  
plexed-input, high resolution ∆Σ ADC. Very low level  
signals and high level signals can be accommodated with  
a minimum of additional circuitry.  
The components connected to CH7 are used to sense  
daylightorphotodiodecurrentwitharesolutionof300pA.  
In the figure, the photodiode is biased in photoconduc-  
tive mode; however, the LTC2408 can accommodate  
either photovoltaic or photoconductive configurations.  
U
PACKAGE DESCRIPTIO  
Dimensions in millimeters (inches) unless otherwise noted.  
G Package  
28-Lead Plastic SSOP (0.209)  
(LTC DWG # 05-08-1640)  
10.07 – 10.33*  
(0.397 – 0.407)  
28 27 26 25 24 23 22 21 20 19 18  
16 15  
17  
7.65 – 7.90  
(0.301 – 0.311)  
5
7
8
1
2
3
4
6
9 10 11 12 13 14  
5.20 – 5.38**  
(0.205 – 0.212)  
1.73 – 1.99  
(0.068 – 0.078)  
0° – 8°  
0.65  
(0.0256)  
BSC  
0.13 – 0.22  
0.55 – 0.95  
(0.005 – 0.009)  
(0.022 – 0.037)  
0.05 – 0.21  
(0.002 – 0.008)  
0.25 – 0.38  
(0.010 – 0.015)  
NOTE: DIMENSIONS ARE IN MILLIMETERS  
*DIMENSIONS DO NOT INCLUDE MOLD FLASH. MOLD FLASH  
SHALL NOT EXCEED 0.152mm (0.006") PER SIDE  
**DIMENSIONS DO NOT INCLUDE INTERLEAD FLASH. INTERLEAD  
FLASH SHALL NOT EXCEED 0.254mm (0.010") PER SIDE  
G28 SSOP 1098  
Information furnished by Linear Technology Corporation is believed to be accurate and reliable.  
However, no responsibility is assumed for its use. Linear Technology Corporation makes no represen-  
tationthattheinterconnectionofitscircuitsasdescribedhereinwillnotinfringeonexistingpatentrights.  
35  
LTC2404/LTC2408  
U
TYPICAL APPLICATION  
DC  
VOLTMETER  
INPUT  
R1  
GUARD RING  
900k  
5V  
0.1%, 1W, 1000 WVDC  
7
ELECTROMETER  
INPUT  
1mV TO 1000V  
3
2
R5  
5k, 1%  
+
R2  
4.7k  
6
(pH, PIEZO)  
LT1793  
0.1%  
0V TO 5V  
4
–60mV TO 4V  
–5V  
R3, 10k  
LT1236CS8-5  
R4  
1k  
6
2
5V  
OUT IN  
8V  
REF  
C1, 0.1µF  
+
+
5V  
MAX  
GND  
4
10µF  
100µF  
3-WIRE R-PACK  
60Hz  
R6  
R7  
5V  
10k, 0.1%  
10k, 0.1%  
AC  
INPUT  
5V  
7
4
3
V
2, 8  
1µF  
SERIAL DATA LINK  
1µF  
R9  
1k  
1%  
R10  
7
V
MUXOUT  
ADCIN  
REF  
CC  
5k  
IN914  
6
IN914  
2
3
1%  
9
CH0  
RT  
MICROWIRE AND  
SPI COMPATABLE  
10 CH1  
11 CH2  
12 CH3  
13 CH4  
14 CH5  
15 CH6  
17 CH7  
LTC1050  
23  
20  
CSADC  
R8  
100, 5%  
+
CSMUX  
CLK  
4
19, 25  
21  
24-BIT  
8-CHANNEL  
MUX  
20mV TO 80mV  
+
∆∑ ADC  
MPU  
–5V  
D
IN  
R11  
24  
SDO  
24.9k, 0.1%  
V
REF  
5V  
LTC2408  
GND  
INTERNAL OSC  
60Hz–RF  
26  
6
COM  
F
O
SELECTED FOR  
RF POWER  
60Hz REJECTION  
<1mV  
J1  
J2  
100Ω  
24048 F31  
1, 5, 16, 18, 22, 27, 28  
Pt RTD  
50Ω  
(3-WIRE)  
FORCE SENSE  
2.7V AT 0°C  
0.9V AT 40°C  
–2.2mV to 16mV  
0V to 4V  
R12  
24.9k, 0.1%  
J3  
V
REF  
5V  
50LOAD  
BONDED TO  
RTD ON  
INSULATED  
MOUNTING  
LOCAL  
TEMP  
THERMISTOR  
10kNTC  
5V  
DAYLIGHT  
HAMAMATSU  
PHOTODIODE  
S1336-5BK  
OMEGA  
0S36-01  
INFRARED  
R13  
5k  
0.1%  
INFRARED  
THERMOCOUPLE  
Fiugre 31. Measure DC to Daylight Using the LTC2408  
RELATED PARTS  
PART NUMBER  
LTC1050  
LT1236  
DESCRIPTION  
COMMENTS  
Precision Chopper Stabilized Op Amp  
Precision Bandgap Reference  
No External Components, 5µV Offset, 1.6µV  
0.05% Max Initial Accuracy, 5ppm/°C Drift  
P–P  
LT1793  
Low Noise JFET Input Op Amp  
24-Bit Micropower ∆Σ ADC in SO-8  
10pA Max Input Bias Current, Low Voltage Noise: 8nV  
<4ppm INL, No Missing Codes, 4ppm Full Scale  
LTC2400  
24048f LT/TP 0100 4K • PRINTED IN USA  
LINEAR TECHNOLOGY CORPORATION 1999  
LinearTechnology Corporation  
1630 McCarthy Blvd., Milpitas, CA 95035-7417  
36  
(408)432-1900 FAX:(408)434-0507 www.linear-tech.com  

相关型号:

LTC2408CG

4-/8-Channel 24-Bit uPower No Latency ADCs
Linear

LTC2408CG#PBF

暂无描述
Linear

LTC2408CG#TR

LTC2408 - 4-/8-Channel 24-Bit &#181;Power No Latency Delta-Sigma ADC; Package: SSOP; Pins: 28; Temperature Range: 0&deg;C to 70&deg;C
Linear

LTC2408CG#TRPBF

LTC2408 - 4-/8-Channel 24-Bit &#181;Power No Latency Delta-Sigma ADC; Package: SSOP; Pins: 28; Temperature Range: 0&deg;C to 70&deg;C
Linear

LTC2408I

4-/8-Channel 24-Bit uPower No Latency ADCs
Linear

LTC2408IG

4-/8-Channel 24-Bit uPower No Latency ADCs
Linear

LTC2408IG#PBF

LTC2408 - 4-/8-Channel 24-Bit &#181;Power No Latency Delta-Sigma ADC; Package: SSOP; Pins: 28; Temperature Range: -40&deg;C to 85&deg;C
Linear

LTC2408IG#TR

LTC2408 - 4-/8-Channel 24-Bit &#181;Power No Latency Delta-Sigma ADC; Package: SSOP; Pins: 28; Temperature Range: -40&deg;C to 85&deg;C
Linear

LTC2408IG#TRPBF

LTC2408 - 4-/8-Channel 24-Bit &#181;Power No Latency Delta-Sigma ADC; Package: SSOP; Pins: 28; Temperature Range: -40&deg;C to 85&deg;C
Linear

LTC2408_15

4-/8-Channel 24-Bit Power No Latency ADCs
Linear

LTC2410

24-Bit No Latency ADC with Differential Input and Differential Reference
Linear

LTC2410C

24-Bit No Latency ADC with Differential Input and Differential Reference
Linear