ADS1246: Problems with Factory Trim

Hi Luke,

Just to summarize the issue you are seeing, please confirm if these statements are correct:

• You have received new ADS1246 devices that have seemingly improper gain scaling coefficients
• On the new ADS1246 devices, the gain scaling coefficient upon powering up the ADC is always 0xFFFFC0 i.e. before anything else is written to the ADC
• On the new ADS1246 devices, the gain scaling coefficient is this same fixed value (0xFFFFC0) for all gains
• There have been no changes to the firmware or board hardware other than the new ADCs
• You can manually reset the gain scaling coefficient by writing to the FSC register, and then the new ADCs behave as expected - however, this value is rewritten to the "incorrect" factory-setting when the ADC is reset, power-cycled, etc

If these statements are true, can you please help answer the following questions:

1. Have you tried performing an A-B-A swap to see if the new devices work in the old boards, and the old devices exhibit the same behavior on the new boards?
2. Where did you purchase these devices from, and when?
3. How many devices did you purchase, and how many are exhibiting this behavior?

Let me know

-Bryan

Hi Bryan,

Thanks you for the quick reply.

1. I have done an A-B-A swap and found the issue to follow with the broken part.

2. My PCB supplier purchased this part from a 'Trusted far east source' called Amos Electronics Ltd. I did not get any details on when but likely within the past 6 months my supplier obtained them and my supplier is asking these questions to their source.

3. I have 100 units here and have tested a small random amount and found all to have this issues. In total my supplier has 600 pieces held for me 

The parts all have the part number of: ADS1246 39KG4 A1CG

Questions for you:

1. I have read back the FSC from my working parts and they are all very close to 0x400000 eg values such as 0x3FED00 (Worst case). This would only give me a small accuracy loss of around 0.1% which I can deal with. What is the typical variation on the FSC, is it typically within ranges such as 0.1%? 

Thanks,

Luke

Hi Luke,

I am looking into providing a definitive answer to your question, but this will be delayed since this is a holiday weekend at TI. I have reached out to several of my colleagues but I would not expect to hear back until Tuesday.

And thanks for the topside marking on this device, that will help us track them over time to try to figure out how they were tested and where they went after we produced them.

In the meantime, it would be helpful to know from your supplier as much as about the history of these devices as you can.

-Bryan

Hi Luke,

I spoke with our test team, they mentioned that the untrimmed gain error on the ADS1246 should be somewhere around ±0.5% max. An interpretation of this information is that if you manually write a value of 400000h into the FSC register for all gains, you can expect that this will result in an error of 0.5% max. However, please be aware that since this untrimmed gain error specification is not included in the ADC datasheet, it is not guaranteed and you use it at your own risk.

I am still communicating with our backend teams to trace the history of the devices you have. Let me know if you were able to track down any more information from your supplier.

-Bryan

Hi Bryan,

Here is the quote from my supplier: 

"ADS1246IPW 600pcs this parts are really 100% new & original, because TI products were very out of stock last year, none of the authorized agents had inventory of this P/N at that time. We bought this P/N from the factory, this parts is the spare surplus material of the factory."

Does this seems like something your factory's do with surplus stock?

I can likely make do with the error, a graph of error % vs likelihood would be useful if you have that information at hand. 

Thanks,

Luke

Hi Luke,

We don't have any additional information to provide regarding the untrimmed gain error. The non-guaranteed specification I provided in my last post is all we can offer you at this point. Again, please use this information at your own risk.

Thanks for providing the statement from your supplier. I am still in the process of tracking down all of the necessary information for the devices you have, I will be able to provide you with a firm update by tomorrow (~24 hours from this post)

Thanks for your patience.

-Bryan