- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
When using the DelSig at high resolution there are many considerations
that come into play. A good example is the design consideration doing a load
cell over large dynamic range.
If we use DelSig at 20 bits, 1.024 Vref, range +/- 1.024, then 1 LSB = ~ 2 uV.
If we look at lead free solder, the seebeck coefiicient is ~ 2 uV / 0K (consult solder
datasheet). So for a 50 C rise in design, each transition junction on the board yields
50 LSBs of error. If in diff mode, and junction count equal on each input, then this
gets eliminated. You can have an imbalance of junctions in chained ground layouts,
signal path nodes not present in ground path, etc., thats when your problems multiply.
Diff mode and layout practices ways of handling this.
Also PSRR in any signal path components, like an OpAmp, can add to this. For
example if PSRR is 60 db, Vdd changes by 100 mV, then that yields OpAmp output
change of 100 uV, another 150 LSBs of error.
Noise, CMRR, thermoelectric effects, INL, DNL, additionally all have to be considered.
Just some of many considerations in attached ap notes, look at AN280 first.
Regards, Dana.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Typo in original post -
This change of 100 uV, another 150 LSBs of error.
should be changed to change of 100 uV, another 50 LSBs of error.
Regards, Dana.