Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

Missing codes in the output of ADC

Status
Not open for further replies.
P

proton

Member level 3
Joined
May 31, 2001
Messages
67
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,286
Activity points
453
Hi good friends,
I've a question about a circuit that I'm debugging: this is an ADC
board (the ADC is LT1400). If I acquire 1000 samples of the same signal, and plot the results, I see missing codes in the output, which sounds
strange to me. There are many missing codes, typically large one to two channels but I do not see a logic like a missing bit or something like that. The ref and power voltages seems good and almost noise free. Any idea ?
thanks
proton
 

adc missing codes

This is simply something you have to deal with, and very often for 12bit ADCs or more. Frequently a datsheet will say 'no missing codes to 10bits' or something similar. ADCs arnt perfect, you dont get absolute equality of channel widths (unless you use a wilkinson convertor for example) and you do get missing codes.

I couldnt find an LT1400, but I did see a 12bit LTC1400 which was 12bit, so I assume this is normal. If its only an 8bit ADC however, you might want to think about a better chip :)

Check the datasheet for your actual chip, see what its supposed to do.
 

adc no missing codes

What is your input signal. Do you aquire the output samples in the correct timming.

This AD is 12-Bit, 400ksps. If you aquire only 1000 samples you do not have all the possible codes. You need at least 4096 samples, this if your input signal do sweep over the full range with one LSB for each sample. I recommend using much more samples.

For INL and DNL evaluation you need more than 4.26829 million samples.

Joey Doernberg, Hae-Seung Lee and David A. Hodges, “Full-Speed Testing of A/D Converters”, IEEE Journal of Solid State Circuits, Vol. SC-19, No. 6, pp. 820-827, December 1984.

Bastos
 

adc missing code

Hi,
yes sorry it is a LTC1400, 12 bits 400 ksamples SAR ADC from linear technology.
However the datasheet states: No Missing Codes over Temperature
which should mean no missing codes up to 12 bits ...
For what concerns the test, the input signal is almost constant, so it is spread over something like 80 channels, this means that 1000
samples are enough. I tried also 5000 with the same result. The missing
codes are still there and seem almost the same from one run to the other.
Do you really think that it could be the standard working mode for this ADC ?
thanks a lot
proton
 

no missing code

proton said:
Hi,
yes sorry it is a LTC1400, 12 bits 400 ksamples SAR ADC from linear technology.
However the datasheet states: No Missing Codes over Temperature
which should mean no missing codes up to 12 bits ...
For what concerns the test, the input signal is almost constant, so it is spread over something like 80 channels, this means that 1000
samples are enough. I tried also 5000 with the same result. The missing
codes are still there and seem almost the same from one run to the other.
Do you really think that it could be the standard working mode for this ADC ?
thanks a lot
proton
Click to expand...

What is "constant", a DC signal, or a type of sinusoidal. The range of the signal is like 80 channels ?? Whats the range of the signal compare with the reference voltage in volts. I think you probably have noise in the references. Any noise with amplitude close to half LSB could give a missing code.

Bastos
 

missing code in adc

LT1400 should be really no missing codes ADC. For me it looks like your circiut have problem with time of sampling. First try to acquire histogram on lower speed - start from 1k, no matter 1000 or 5000 points. If it make results better then in my oppinion the problem is with impedance of analog circuit - check acquisition time vs source impedance plot on datasheet.
If decrease of acq speed do not change results try to use another chip, maybe you hurt this one. But I think it won't be necassary.

regards
KamW
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top