ADC Common-mode range
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content
‎2022-08-09 12:14 AM
Hello,
I am trying to understand the common-mode range in differential mode - according to the data-sheet, it looks like it is quite narrow:
Could you please clarify?
We wish to wok with 0 to 2.5 V signals.
Thanks.
- Labels:
-
ADC
-
STM32G4 Series
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content
‎2022-08-09 1:55 AM
It is what it is. The usage of STM32 ADC in differential mode is limited.
JW
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content
‎2022-08-09 4:25 AM
I never used or looked at it before, but +-180mV is ridiculous! :D
Anyway, especially with an external signal source, I would always put some opamp with the same supply range as the MCU before the MCU pin(s), if space, power and cost allow.
For protection, and many more options, and probably better signal quality.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content
‎2022-08-09 4:50 AM
Yeh, Thanks.
But with such a narrow CM range, a differential-output amp is a must, unless I'm missing something.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content
‎2022-08-09 5:18 AM
> a differential-output amp is a must
Or just use a "normal" (precision/instrumentation) opamp to convert the differential input to single-ended, and feed that to single-ended ADC.
JW
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content
‎2022-08-09 5:41 AM
> never used or looked at it before, but +-180mV is ridiculous! :D
Yes, it is. But what's worse, although understandable, ST does not exactly brag about it, so you have to read the DS very carefully, as this is a surprising information. In 'F3, this was at least in the Errata, but since then in newer families this is simply a feature documented by a single number in the DS.
To be fair, it's probably consequence of using what's primarily cheap digital semiconductor process to try to produce relatively precise analog circuits. This of course has its limitations.
OTOH, I would expect a warning to be present in all ADC appnotes which deal with those STM32 which do have the differential mode in ADC - and it's in none, including the newest AN5346, where's not even a mention of the differential mode at all.
JW
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content
‎2022-08-09 7:48 AM
Thanks a lot.
Unfortunately this is what we'll do to keep the cost reasonable.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content
‎2022-08-09 7:50 AM
Gladly I've ran into it at a very early stage of the design, but still this is unpleasantly surprising.
Kind regards.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content
‎2022-08-09 8:27 AM
Additionally, if frequency/timing and filters allow it, you can capture two single-ended channels and do the subtraction in software. Or, if you have multiple ADC peripherals, you can have a simultaneous conversion of those channels.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content
‎2022-08-11 2:47 AM
Out of range common mode voltage does not break the ADC. But you get substantial changed gain.
