I have tested 5 pcs of AD8226. Please refer to the attached file. Two of them can be seen like temperature drift.
Is this caused by input offset voltage drift and can I improve this? (No such voltage variation regarding other 3 pcs.)
I have a couple of questions about your setup. Could you clarify what your inputs are? You mention that the inputs to the AD8226 are shorted. Are they shorted to ground? What is the amplifier that you are using to drive the reference? Is there anything else connected to the inputs of the AD8226 or are you testing it on its own? If you could provide your most complete schematic, that would be very helpful.
May I ask what your application is?
I have attached the updated schematic.
Application is a pressure gauge. This issue is not much problem but I would like to know what
causes such phenomemon.
How are you measuring Vout-Vref? It might be a good idea to check Vout and Vref separately so that we can narrow down what the problem is. The AD8226 is a linear part and this type of behavior has not been observed on characterization.
There are some things that you could look at though. At low temperatures, there is some possibility of getting condensation on the board which might create problems with your measurements. You could check your board when it's at 0degrees to see if that's happening. Also, if you have any long wires, noise could easily couple into your circuit. If you have those, you could twist them in order to minimize the problem.
I tested other three pieces of AD8226BRMZ. Please refer to the attached page 2 and 3.
You can see the datail schematic for this test in page 2 and the page 3 shows the test results.
First I tested three samples setting their gain to 9. We found voltage variation (a) on the sample#1 and a little
variation (c) on the sample#3. No such variation on the sample#2. Next I tested them setting their gain to 1 (Rg: OPEN).
In this test no Vout variation on all samples, so I don’t think that there is some condensation on our boards when they are at 0 degC.
As you see that Vref of the sample#1 and #2 had much noise. I changed the decoupling cap of the 3.6Vcc form 1uF to 4.7uF.
After changing it to 4.7uF, I tested sample#3 so its Vref noise is better than those of the sample#1 and #2.
I used a data logger (sampleing interval :10sec) in this test. Also I will test the same thing using a twist cable, may be tomorrow.
1. This Vout variation (a) & (c) seems to relate on gain value. What do you think about this?
2. What is the cause of the Vout variation?
3. What is the cause of sample#1 noise (b)? Is it caused by noise from Vref?
I twisted measuring cables following your advice and don’t find any Vout variation for the moment.
The problem seems to gone away. But I have a question.
Which do you think happened?
AD8226 was affected by noise and its output varied.
AD8226 output was normal (did not vary) but measuring instrument showed varried waveform (false waveform) by noise.
It is more likely that there was some environmental noise that was coupling into your inputs through the wires. Twisting wires is a good practice when evaluating front-end circuits, especially if you are operating in noisy environments. Are your circuit now performing as you expected or is there anything else you need help with?
I will use the AD8226 but has asked a question. Please take care of this case.
Is the AD8226 a chopper type amplifier, if it is not, what kind of type?
I used a chopper type of T.I product before but had a noise problem, so our QA team is not
positivefo us to use a chopper type.
Retrieving data ...