MCP4725 offset trimming

Thread Starter

Hamed T

Joined Dec 28, 2018
18
I'm going to use MCP4725 DAC to set voltage/current in a digital lab PSU, when I set the output of the DAC to 0 in Arduino there's a 23mV on DAC output and it's almost linear. I need to trim this offset error or at least reduce it to 1-2mV.
I did googled a bit and find this op amp configuration:

caLEt.png
But it's not removing any offset from the output, any idea how can I remove this damned 23mV or at least reduce it to 2mV?
 

MrChips

Joined Oct 2, 2009
30,720
Take a closer look at your setup. I don't see anything wrong with the circuit schematic and you ought to be able to shift the DC output downwards by 50mV with R5 trimmer.
 

ci139

Joined Jul 11, 2016
1,898
usually the datasheet of the DAC provides it's adjustment methods . . . . . . nope - your one doesn't

Haa , ? how you power your op amp . . . if it's negative supply pin is at signal ground - it likely cannot drop it's output even to the signal ground (which is also the Vss or Vee in this case)

whats wrong with the offset however you can adjust your voltage setting and current limiting circuitry to a variety of dac output ranges /// although - you cannot get more near to ground Zero ← if this is your lowest voltage in the system . . .
 
Last edited:

ci139

Joined Jul 11, 2016
1,898
Simple virtual test -- the source impedance (basically all the signals should be level) , the referencing voltage each op amp favours (Vdd for TLV314 // Vss for LMV821 - - what it would be in real cannot be reliably predicted by this Spice test)
Op-Amp - Offset-Adj-TEST - aa1.pngOp-Amp - Offset-Adj-TEST - aa1 - PLOT.png
___________

▼↓▼ Versus ~ DAC ► referenced to varying "Vrf" -- Vrf referenced DAC ► referenced to VBrf2 (e.g. to the median of 5V = 2.5V) /// all inputs buffered . . . each of the 6 op amps may require additional offset correction -- e.g. -- it's VBrf or VBrf2 shifted by milli-Volts ← in ideal those shifts should be done by op amps or voltage references . . . is how the component count for a simple setup gets sky-rocketing while the precision is in question ▼↓▼
/// about http://www.ti.com/product/LM101A-N /// or perhaps https://www.analog.com/en/products/ad8628.html

Op-Amp - Offset-Adj-TEST - aa2.pngOp-Amp - Offset-Adj-TEST - aa3.png

___________

Edit -- dont follow the last test = the current mirrors make the adjust voltages dependent on / relative to the supply rails ← that are dependent on dynamic power draw of the Pos and Neg supply . . . /// just didn't want to go nuts with references . . . /// but basically -- is one possible way to trim offsets . . . if the Adj. references is replaced by ones that stay firm relative to signal ground

Op-Amp - Offset-Adj-TEST - aa4.png
 

Attachments

Last edited:
Top