Products > Test Equipment
16-bit ADC testing equipment
<< < (3/3)
dobsonr741:
Before you start testing - take a look how the device should be prepared to avoid the usual pitfalls: https://www.st.com/resource/en/application_note/cd00211314-how-to-get-the-best-adc-accuracy-in-stm32-microcontrollers-stmicroelectronics.pdf

If you did the prepwork right, you might get close to the datasheet specs. Probably a bit below it, as the specs reflect an ideal environment of a billion dollar ST lab.   
Doctorandus_P:

--- Quote from: iheb on May 02, 2023, 01:40:40 pm ---the best budget friendly approach to test this ADC ...

--- End quote ---

You can get a pretty good DC accuracy from a PWM output and filtering. Almost any microcontroller can generate an 16bit PWM output, and this has an inherently good INL. To preserve that, you have to use a filter which does not add any DC offset (so a simple two or three stage RC filter works, especially with low leakage capacitors). Another limitation is the reference voltage. The uC's own power voltage is "dirty", so using a buffer and an external reference voltage is needed. If you can use the same reference voltage as the ADC, then this is always the best to do radiometric comparisons.

A year and a half ago I started the thread below. I believe it could work quite nicely but I still have not put much effort in it. That thread also has some links to proffesional measurement equipment service manuals (with schematics) that uses this method.

https://www.eevblog.com/forum/projects/making-a-pwm-based-voltage-standard/msg3663823/#msg3663823
cncjerry:
Exactly what STM32 part are you testing?  If performance is critical, you could use an external ADC as those by LT have known performance characteristics, available board designs, etc.

As others have pointed out, generating signals that exceed the performance of the device would be a challenge in itself.
coppice:

--- Quote from: Frex on May 02, 2023, 02:49:45 pm ---I highly recommend you this reading:
https://www.edn.com/test-18-bit-adcs-with-an-ultrapure-sine-wave-oscillator/

--- End quote ---
Testing with a pure sine wave is not a bad approach, but a much better approach is to use 2 pure sine waves, with a small difference in frequency. Mix the two, digitise the result, and look for intermodulation products in the captured data.
Navigation
Message Index
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod