Hi,
I am using a DAC8581 for high precision stimulation in a biological setting. I have ordered multiple batches of DAC8581's and have noted a substantial voltage offset. The most recent DAC I measured had an offset of about -6.5mV, which is a fairly typical magnitude. The power inputs were all +/- 5V. I tried fooling around with the DAC such as setting it to the maximum and minimum voltages (about 4.63V and -4.67V) before setting it back to 0, but the offset persisted. I've double checked each of the input power/ground pins and they are all within 1 mV of what they should be.
Is this a known issue? Is there something I could do to fix this aside from individually finding the offset for each DAC and using a lookup table when stimulating?
Thanks,
Guy
Hi Guy,
Where are you measuring the offset voltage at? The DAC8581 being a bipolar device, measuring offset error is a little different; you want to look at -Vref. Are you writing code 0x0000 to the device and comparing it to ground or are you writing 0x8000 to reach -FS and comparing that to -Vref? You want to write 0x8000 to get minus full scale and then look at the voltage value for -FS to compare to the reference voltage to solve for the offset error of a bipolar DAC.
Regards,
Tony Calabria
This offset of about -6.5 mV is when I write 0x0000 and compare it to ground. When I write 0x8000 I get an output from pin 2 of about -4.67V. The negative reference voltage going to pin 3 is -5V.
You are running into a headroom problem between the output voltage swing set by the reference voltage and the analog negative power supply. If you test a few other codes near near the full scale negative supply, 0x8000 (0x8001, 0x8002, 0x8003), you will notice that they all produce the same output of -4.67V without a change. This has to do with the headroom of the built in output buffer amplifier. My recommendation is to try a 4V reference with +/-5V supplies, giving the output swing 1V headroom from both the negative and positive supply. You should then see offset error that reflects the data sheet.
I tried:
+4V applied to pins 1,4
-4V applied to pin 3
+5V applied to pins 13, 14, and 16
The resulting voltage offset was approximately -5.2mV. The negative swing was -3.734V and the positive was 3.699. When I tested other codes near the full scale +/- supplies they did not produce the same output - they made approximately the correct voltage steps (hard to tell exactly though).
You will want to apply +4V to only pin 1 (Vref). Set pin 4 (AVDD) to +5V, pin 3 (AVSS) to -5V and DVDD (pin 13/16) to your logic level voltage. Pull pin 14 to your DVDD level to keep it out of /CLR mode. This will be the best way to measure the offset voltage. Do you have a schematic of the system you are using to measure the offset voltage that I can take a look at?
Tony
I can't do that with my current setup, but I don't think that is the issue. We tried swapping out a different DAC from a new batch on the stimulator we are using and there was no offset. The stimulator we are using uses 5V for the power and +/- 5V for the references, like my test board.
A different DAC from the new batch gives me an offset on a testing board though.
How would you prefer I send you my schematic? Do you have an email I should send it to?
It can be posted on the forum or sent to me directly using a private conversation.
hi....tony and guy...i am also using DAC8581....are there any sample codes available on how to send data to DAC...if yes...cold u plz post the links....thnx
Prateek,
We do not have sample code directly for the DAC8581. Please refer to your other post.
-------------------------------------- Kevin Duke Precision DAC Applications