ADS1015: Conversion Register Table Bits 3:0 return non 0h value

Hi Nathan,

Sorry to see you are experiencing some trouble. For our 12bit parts in this family, the 4 least significant bits should be 0. However, all 16bits will be present with our 16bit devices in the family. Can you tell us who the manufacturer of the dev boards? It is very possible that an ADS1115 may have been put on the board instead.

Regarding the input to the ADC, was this 2.5V? If so, the full scale range isn't set properly. If not 2.5V, can you provide a known input to the ADC within the full scale range and capture a single conversion using single-shot mode and share the conversion results?

Regards,
Aaron

Thanks Aaron,

manufacturer of the board is HiLetgo: https://www.amazon.com/dp/B01MTU9LJV/ref=cm_sw_r_cp_api_glt_i_W3X1H84V09WWGK8TB699?_encoding=UTF8&psc=1

Markings on the chip indicate that it’s an ADS1015 - so I’m fairly certain it’s the correct IC.

Input is biased to 2.5V but should have only a +-1.4v swing (so 0.6v to 3.9v max voltage on A0/A1). Since it’s an AC signal just needed to bias it up so that it wasn’t sending negative voltage to the board.

I can try single shot mode tomorrow night.

Hi Nathan,

Thanks for the input clarification. If you can measure a DC voltage, that would be great. From there, we can take a look at the conversion results.

One more thing, can you also provide a schematic for me to take a look at? If you don’t want to do so publicly, you can send me a friend request and we can discuss this privately. 

Regards,

Aaron Estrada

Schematic is very similar to the one proposed by Joseph Wu here: ADS1115: Differential Mode, estrange readings when using the 4 channels simultaneously. - Data converters forum - Data converters - TI E2E support forums.

However, since I can't center tap the transformer to bias it, I had to separately bias both lines to +2.5v using a standard voltage divider with 10uF filter cap. The schematic of the dev board is unfortunately not available although all pins appear to be wired through the correct breakout holes and with the recommended resistors/caps.

I can draw something more detailed up if needed, but I suspect we'll get more from the DC voltage test. I should be able to do that tomorrow.

Hi Nathan,

Thanks for the detail and I agree that preforming a DC voltage test would help out. I look forward to your results. 

Regards,

Aaron

Measured DC voltage is +2.05v to GND for both A0 and A1

Single shot, 2400 SPS, 4.096 FS, A0 -> GND
Write 01 C3 A3 
WriteRead 00 2-bytes
Output: 3F D6
Output: 3F CE
Output: 3F CD

Single shot, 2400 SPS, 4.096 FS, A1 -> GND
Write 01 D3 A3
WriteRead 00 2-bytes
Output: 3F CF
Output: 3F C9
Output: 3F C4

Interestingly enough, if I put it into differential mode (e.g. both A0 and A1 read +2.05v) it appears to output correctly.
Single shot, 2400 SPS, 2.048 FS, A0/A1 Differential
Write 01 85 A3
WriteRead 00 2-bytes
Output: 00 00
Output: 00 00
Output: 00 00

The top side marking on the IC is BCMQ which I believe translates to the ADS1015 MSOP10 package, but maybe I'm mistaken?

Hi Nathan,

The single shot results for 2.05V on the inputs are consistent with what we would expect for one of the 16-bit parts. For the ADS1115, the LSB is 125uV for a FSR of +/-4.096V. Doing the conversion, the data shows ~2.04V at the input. 

I did a part marking look up on our website for BCMQ and did not turn anything up. 

Is that the only marking on the part? Can you share a clear image so we can take a look?

Regards,

Aaron

Interesting - and would explain the odd output I've been seeing. Clearest shot I can get at the moment.

ADS1015QDGSRQ1 datasheet - Texas Instruments ADS1013, ADS1014, and ADS1015 (digchip.com) - this is originally where I translated the marking

I honestly wouldn't be surprised if this is a fraudulent chip - I need to stop buying dev boards off Amazon from unknown manufacturers for my home lab

Hi Nathan,

Understood! Let see if I can get confirmation still. I will get back to you soon.

Regards,

Aaron