ADS1210: ADS1210U does not behave like ADS1210P

Hi Renan,

The only difference between the P device and the U device is the package type.  Can you give me more specific details on what an 'altered function below a programmed conversation rate' actually means?  Also, were the U devices bought through TI or our normal distribution partners?  Or were they also bought from Rochester?

Best regards,
Bob B

Above the 300Hz data rate, behavior is the same. My circuit is a little different than usual because of optical isolation, but it has been trouble-free for over 20 years. The crystal frquency is 9.8 MHz.

We are running in turbo16 mode at rates from 50 to 1000Hz. We don't use background calibration, digital gain, or the internal reference.

Hi user6364447,

Welcome to the E2E forum!  You haven't stated the particular issues you are seeing. 

• Can you give specific detail as to what you are expecting and what you are seeing? Do you have data that you can send me?
• Can you send me your schematic or at least tell me what is connected to the input? 
• Are you seeing differences on a single device or multiple devices? 
• Are you using the internal reference or an external reference?
• Have you modified your board layout to accommodate the U package?  Or are you using an interface adapter board (SOP to DIP)?

Best regards,

Bob B

Conversion results become near-random numbers, instead of voltages with over 22 bits of effective resolution. That is the case for individual conversions as well as for continuous mode conversions.

There are different products involved, but all are mature, reliable products. We need high input impedance (10^13 or higher) so we need a buffer in front of the chip. We also need better than 90 dB common mode rejection at DC, so we have an external in-amp. That all means we need an external multiplexer to bring in input signals and calibration signals. We do our own calibrations in firmware.  So, inputs go to a differential multiplexer, which feeds a differential amplifier, with level shifting to take advantage of the full voltage range of the ADS1210. It is all under the control of a microprocessor.

I have ordered additional parts through Arrow in order to determine if I have a bad batch of parts. Those parts will arrive tomorrow or Monday.

I use an AD586JN external reference.

I changed thelayout to accomodate the surface mount part. I know the circuit is correct because it runs just like the P part above 300 Hz.

Did I cover everything?

Hi user6364447,

You have explained your configuration and system design, but really all you have told me is the ADS1210 works as expected at 300sps and above but not at less than 300sps.  You also said it worked with the P version but not with the U version.  I have no idea what kind of error you are actually seeing as you have presented no data.  What would be helpful would be to see continuous conversions and contiguous data for 128 to 256 data points so I can observe the behavior at both the working and non-working scenarios.

Here is what I suspect is happening.  You are seeing excessive noise which results in random readings.  Based on your settings and clock speed, you are expecting a very high effective resolution.  Where and how did you get the parts you are now testing?

Best regards,

Bob B

I can send data if you require it, but I suggest that is not informative. With dc voltages at the inputs, correct behavior simply means columns of voltages steady to within a few microvolts.

Incorrect data looks like random numbers, generated by a 5-bit A/D converter with an open input mixed in with other random numbers. Our existence as a company is based on getting better usable resolution than anyone else. NASA uses our stuff for calibrations. There is no way that we would mistake random numbers for noise.

The parts I have now did not come from an authorized distributor. Except for discontinued parts, everything we buy for production comes from authorized distribution.  The first thing I did when I saw this problem was order parts from Arrow.  I didn't do that in the first place because of sticker shock. I paid over twice as much for the ADS1210U as I ever paid in 20 years for an ADS1210P.  That difference reduces the margin on the finished product by over 20%, but that is another issue.

Hi user6364447,

I understand your frustration and truly want to understand what might be happening.  One of the reasons I would like to know more details and to see data is to see if I can replicate the same issues.  I'm going to close this post for now.  Once you have tried the new devices let me know what you are seeing.  You can either respond to this post or start a new one.

Best regards,

Bob B

We got the ADS1210U chips today from Arrow. Those parts
work as they should. The original parts were obviously defective.

At a 60 Hz data rate, with our active offset drift suppression,
I now see the expected 22 bits of effective resolution. We do a
running average of pairs of conversions to get an approximation
of 24 bit total resolution because in turbo16 mode, the two
least significant bits are always the same. (That has been true
from the start.)

Now, if I can find an authorized distributor who will sell the
U parts at something approaching the P price, my problems are
over.

Thanks,
Tom

Hi Tom,

I'm glad you had a better response to your testing using the parts from Arrow.  We appreciate the dedication and use of some of our older products by our customers.  The ADS1210 is now 20+ years old, and like many other things in the electronics world the ADS1210 is getting more difficult and costly to produce.

For future designs I would highly recommend moving to a newer device.  We have no plans for obsoleting the ADS1210, but as I said it is getting more difficult and costly to produce.  One possible alternative is the ADS1255.

Let me know if there is anything more we can do to help, but unfortunately I can't help with the pricing.

Best regards,

Bob B

It wasn't a question of testing, the issue was whether I would be able to ship systems to our customers. We still support our Apple][ products from 40 years ago. That gives you an idea of where we stand on the obsolecence scale. I am aware that the ADS1210 is over 20 years old. I first designed it into a product in 1997, a product we still sell in quantity.

Interesting how the single-source parts get more expensive, and the multi-source parts get less so.

We make data systems that go into other people's products. The final product may be an analytical instrument or a process control system or a mechanical tester or a hundred other things. We can change our product so long as the customer can't tell it is any different. We support our hardware with API's that allow the user to easily match our hardware with their software. Anytime the computer interface or operating system changes, we supply a free update, if needed, to maintain compatibility. We support all the data systems we've sold that use the ADS1210 all the way back to Win98.

If I could not get the ADS1210U to work, I would have needed to pick some other part, design the new data system around it, and then arrange for the footpint, specifications, functionality and computer interface to exactly match the existing systems. Then repeat for four other products. The match would include the price, which I am not free to raise at will.

So, you see why I am glad you have no plans to obsolete  the ADS1210U.

Thank you,

Tom