search

adafruit / Adafruit_ADS1X15

Driver for TI's ADS1015: 12-bit Differential or Single-Ended ADC with PGA and Comparator
Other
289 stars 301 forks source link

Bug in lib or defect in chip ? 16 bit ADS1115 behaves like 12 bit chip #51

Closed ullix closed 4 years ago

ullix commented 4 years ago

I am using an ADS1115 with a ESP32 and see strange responses, which make me think that only 12 bits are used instead of 16. Since this lib was extended from the 12 bit ADS1015, I am wondering whether something has been forgotten in the lib for the 16 bit variant? Or is the chip defect? I have tried three different ones, all from one shipment, all with the same problem.

I am feeding some ~1.6V - half the 3.3V supply, obtained via a 10:10k divider - to channel A0 of the ADS1115 and measure single-ended with gain GAIN_TWO.

Most of the time I get ADC readouts of 26368 (= 1.648V). I would expect to get some jitter of at least 1 bit up and down, but the only other readout I get is 26352 on 2 of the 3 chips, which is lower by 16, or I get 26384 on 1 of the 3 chips, which is higher by 16.

    // #include <Adafruit_ADS1015.h>
    // Adafruit_ADS1115 ads;  /* Use this for the 16-bit version */

    // 2 out of 3 chips:
    AIN(0): 26352, Voltage: 1647.000 mV
    AIN(0): 26368, Voltage: 1648.000 mV
    AIN(0): 26368, Voltage: 1648.000 mV
    AIN(0): 26368, Voltage: 1648.000 mV
    AIN(0): 26368, Voltage: 1648.000 mV
    AIN(0): 26352, Voltage: 1647.000 mV
    AIN(0): 26368, Voltage: 1648.000 mV
    AIN(0): 26368, Voltage: 1648.000 mV
    AIN(0): 26368, Voltage: 1648.000 mV
    AIN(0): 26352, Voltage: 1647.000 mV

    // 1 out of 3 chips
    AIN(0): 26368, Voltage: 1648.000 mV
    AIN(0): 26368, Voltage: 1648.000 mV
    AIN(0): 26384, Voltage: 1649.000 mV
    AIN(0): 26368, Voltage: 1648.000 mV
    AIN(0): 26384, Voltage: 1649.000 mV
    AIN(0): 26368, Voltage: 1648.000 mV
    AIN(0): 26384, Voltage: 1649.000 mV
    AIN(0): 26368, Voltage: 1648.000 mV
    AIN(0): 26384, Voltage: 1649.000 mV
    AIN(0): 26368, Voltage: 1648.000 mV
    AIN(0): 26368, Voltage: 1648.000 mV
    AIN(0): 26368, Voltage: 1648.000 mV

Since 26368 = 0x6700, 26384 = 0x6710, and 26352=0x66F0 it looks like the lowest 4 bits might have been ANDed out? I checked the lib, but I could not find the problem.

The chip itself has an imprint of

92 [TI logo] BOGI

which I believe confirms it as a ADS1115.

ladyada commented 4 years ago

@ullix what is the exact breakout you purchased?

ullix commented 4 years ago

This one: https://www.amazon.de/gp/product/B07QHWLTTS/ref=ppx_yo_dt_b_asin_title_o03_s00?ie=UTF8&psc=1

On the back of my chips there is an imprint "QIFEI" with a logo unknown to me. The closest to that was this on amazon: https://www.amazon.de/sp?_encoding=UTF8&asin=B07BFVFJH6&isAmazonFulfilled=1&isCBA=&marketplaceID=A1PA6795UKMFR9&orderID=&seller=A19GB54XH1C2M3&tab=home&vasStoreID=; not sure if related.

ladyada commented 4 years ago

since you're the only person with this issue, please contact the vendor first

freestyledork commented 3 years ago

I also experience the same issue.. Did you ever get it figured out?

ullix commented 3 years ago

Yes. Without any doubt my chips were forgeries! 12 bit ADS1015 chips very professionally disguised as 16 bit ADS1115.

The steps to detect are described here: https://forum.arduino.cc/index.php?topic=694956 The pages are in German, maybe Google can translate. Otherwise, there is code and data sprinkled in the pages which may sufffice to let you understand.

freestyledork commented 3 years ago

Yes. Without any doubt my chips were forgeries! 12 bit ADS1015 chips very professionally disguised as 16 bit ADS1115.

The steps to detect are described here: https://forum.arduino.cc/index.php?topic=694956 The pages are in German, maybe Google can translate. Otherwise, there is code and data sprinkled in the pages which may sufffice to let you understand.

Thanks for the quick follow-up. I was afraid they were forgeries. I tried another library and had the same 12bit results, so it's not the library. I also saw a bunch of comments from the amazon knock offs that the boards are actually 12bit ADS1015. I will be returning mine.

BBtheEE commented 3 years ago

Yep, I just ran into the same issue. I thought there was a bug in the library at first, but it turns out it's the chip. I checked the markings, which look legit, but apparently they are not.

ullix commented 3 years ago

@BBtheEE @freestyledork Just to check whether there is a common source for the false chips: where did you purchase them? I got mine from AZDelivery via Amazon: https://www.amazon.de/gp/product/B07QHWLTTS

freestyledork commented 3 years ago

@ullix I tried 3 different suppliers. They all ended up being the wrong chip. I purchased from HiLetgo / MakerFocus / Organizer on Amazon. I returned all 3.

ladyada commented 3 years ago

lol think of how much money you saved! 👍

freestyledork commented 3 years ago

lol think of how much money you saved! 👍

All of it? Shipping is free and it doesn't cost anything to return stuff.

ladyada commented 3 years ago

your time is worth more nothing :)