Because of the Thanksgiving holiday in the U.S., TI E2E design support forum responses may be delayed the week of Nov. 21. Thank you for your patience.

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

ADS1115-Q1: Conversion time ADS1115 vs. ADS1115-Q1

Part Number: ADS1115-Q1
Other Parts Discussed in Thread: ADS1115, , ADS1015

I have an application that requires measurement of eight single-ended channels preferably at a sample rate of 100 SPS. I have used the ADS1115 in other ADC sampling applications, so I put together a board with two VSSOP-10 ADS1115s.

When I started prototyping I realized that one of the ADS1115s ("B" marked 0B BOGI // ADS1115IDGSR) is much slower than the other one ("A" marked 07 TI 19J6 / ADS1115BQDGSRQ1). The A-chip (Q1) will return single mode conversions in less than one millisecond, while the "B" chip requires 8-9ms between single mode conversions. If I run in continuous mode, but switch between channels at every read, I see the same proportion of latency between the two different chips.

I have scoured the data sheets for both, but I do not see anything in the Q1 data sheet that indicates why it is so much faster. My application will work and I can just barely squeeze out 100 SPS using the Q1's, but not the IDGSR chips. On my bench I've got 3 Q1s and 4 of the others and they all behave the same, so I assume this isn't some type of anomaly. I'm testing them both with the same 400K I2C bus, sampling set to 860 SPS, and a PGA of 6.144.

Can you confirm that the Q1 chips, by design, are almost 10X faster in their conversion processing?

thank you.

  • Hi Mike,

    The Q1 and non-Q1 devices are functionally the same. During your tests, have you confirmed that both Q1 and non-Q1 devices are configured the same? It sounds like it from your post but I would just like to be sure. Additionally, is the analog front end circuitry the same for both tests? I am just trying to find any potential differences that could be causing this conversion timing difference. 

    Regards,
    Aaron Estrada

  • Hi Aaron,

    Yes. The hardware and code is exactly the same between them. If I set up my MCU on the bench with the same compiled config and I2C code and then just swap out the two ADCs I get a different result between the Q1 and non-Q1. For bench testing I am either shorting the inputs to ground or to +5V, but I am doing it the same across the two different ADCs for testing purposes. I am not currently using the alert pin. I can enable that and get more precise in measuring the differences.

    As a side question, if I want to use the ADS1115 to continuously sample all four channels (between ground) what is the theoretical maximum SPS I can get from each channel? It seems me that the time cost to switch channels in continuous mode is about the same as just invoking each channel in single mode conversions, since it does not help me to do multiple samples of one channel before moving to the next channel. Right now I'm at 9ms per "switch", so that gives me an effective SPS of 28 per channel (e.g., 9ms = 111 SPS across all four channels / 4 = 28).

    Thanks,

    Mike 

  • Hi Mike,

    Interesting. Thanks for the details. Let me dig to see if I can find any information on any differences between the Q1 and non-Q1 version. 

    Regarding the max SPS question, you can refer to the following E2E post linked below. You essentially need to determine the time it will take to convert AND receive/send I2C transactions. That time should be the overall SPS.

    https://e2e.ti.com/support/data-converters-group/data-converters/f/data-converters-forum/134959/delta-sigma-ads1115-sample-rate

    With Single-shot mode, things get a little more complicated because the device goes into a low-power state after conversion and there is a wake up time of ~20us. Additionally, you need to add 10% of the data rate period for the internal clock to stabilize after the device wakes up. This should be documented in the E2E post below. 

    https://e2e.ti.com/support/data-converters-group/data-converters/f/data-converters-forum/921565/ads1115-ads1115-s-safe-time-for-switching-between-two-channels

    Regards,

    Aaron Estrada

  • Hi Aaron. I got another batch of ADS1115-Q1's today (had to scrounge them from China) and they are all screaming fast, like the ones I have here. I swapped out the non-Q1's from my PCBs and I am reliably getting single shot reads at 900 microseconds apart (that includes the wake up time). If I rotate between the four channels after each read I can reliably get 2,528 samples per second (632 SPS per channel). It seems to me if someone wants to know the maximum SPS to sample all four channels without "double sampling" before moving channels this is it. FWIW the new Q1 chips I got from China are marked "0* BCOQ" which doesn't show up in the TI parts lookup, but does show up in the Q1 datasheet. For my application I am all set, but it would still be good to know why the Q1 is 10X faster than the non-Q1.  Thx!

  • correction... they were marked "08 BCOQ", which does not show up in the part finder.

  • Hi Mike,

    Glad you were able to get ahold of more parts and are all set. You say that the parts are marked as 08 BCOQ and it is in the data sheet but I cannot find it. Can you show me where it says that? 

    Since you are seeing a 10x speed difference, I am worried that you might have some ADS1015 parts on hand. That device has a max data rate of 3.3kSPS...

    Regards,

    Aaron Estrada

  • I was also concerned I might have a different chip, but I found a reference. Here is a link to an old ADS1115-Q1 data sheet from 2011. On page two it references the BCOQ top-side markings: 

    https://datasheet.octopart.com/ADS1115QDGSRQ1-Texas-Instruments-datasheet-41214608.pdf 

    Again, it doesn't seem to matter if it is the BCOQ marked Q1 or the Q1 marked 19J6, but both perform the same. They seem to stabilize 10X faster than the standard ADS1115 in single-shot mode.

  • Hi Mike,

    Thanks for sharing. This is something that is very interesting to me and I will have to dig in further. Can you share more details on how exactly you are seeing this difference in the conversion time between the Q1 and non-Q1 parts? You mentioned that you are not using the conversion pin so I am curious. Additionally, is there any quantifiable data you can share that shows this difference?

    Regards,

    Aaron Estrada