o_C drifty - cannot calibrate

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: lisa, luketeaford, Kent, Joe.

Post Reply
stickjam
Learning to Wiggle
Posts: 8
Joined: Thu Jan 23, 2020 11:20 am

o_C drifty - cannot calibrate

Post by stickjam » Wed Jun 24, 2020 8:19 pm

I've got a completed 8HP Ornaments & Crimes build that is impossible to set up because the CV outputs drift continuously in calibration mode.

The VRef output of the DAC (the highest quality TI "D" code variety) is a rock solid 2.5001V. The drift is on all 4 channels and seems to be temperature-related as cold or warm air make the CV outs move a lot! Resistors are the tolerance called for, but the 0.1% 100K and 24.9K are 200PPM thick films. It also seems strange to me that the two 47K resistors used to derive Vbias from Vref are spec'ed at a higher tolerance.

The drift seems drastic enough that it could be more than just resistor self-heating doing it, unless the fact that I bought the cheapest 0.1%ers that Mouser sells might not be be up to spec.

I'm just wondering what fellow o_C DIYers have experienced before I place an order for much more expensive 5PPM thin films for all the 0.1%ers as well as the two 47k's to make sure.

Thanks
-Bob

User avatar
emmaker
Veteran Wiggler
Posts: 630
Joined: Sat Mar 10, 2012 5:07 pm
Location: PDX

Re: o_C drifty - cannot calibrate

Post by emmaker » Wed Jun 24, 2020 8:36 pm


pld
Ultra Wiggler
Posts: 942
Joined: Thu Mar 05, 2015 5:15 am
Location: Germany

Re: o_C drifty - cannot calibrate

Post by pld » Thu Jun 25, 2020 7:47 am

Yeah, you'll probably get more eyes on it in one of the existing build threads.

The different resistors serve different purposes, and you shouldn't need 5ppm ones. AFAIK people have built them with plain 1% as well (although there are plenty of uncalibrated ones in the wild so that's not saying much). There's been some (> 1 :)) builds with a drift issue but I don't recall them ever posting a resolution; at one point IIRC the theories included fake opamps.

- Drift up or down, how much is "a lot"
- What does the vbias 1.25V look like (between the 47K)
- Re-check soldering and/or clean off flux

stickjam
Learning to Wiggle
Posts: 8
Joined: Thu Jan 23, 2020 11:20 am

Re: o_C drifty - cannot calibrate

Post by stickjam » Thu Jun 25, 2020 7:57 am

emmaker wrote:
Wed Jun 24, 2020 8:36 pm
Might try here.

viewtopic.php?f=17&t=189224
Yeah, I've been through that one on the 8HP o_c. It references the "original o_c thread" for a drift problem involving caps. I haven't had the fortitude to read the nearly 100 page thread in its entirety, but my searches of the topic turned up no such subject other than recommending using the C or D grade DAC which has better temperature stability.

Like I said, it's acting like it's directly related to temperature because it stabilizes once it's been powered on for an hour in a room with still air. I tried the freeze spray on a plastic stick trick to produce isolated temperature change on individual components, but it's so humid right now that I think condensation on the stick is causing side effects.

I've got no problem buying a few of the most ridiculously accurate thin film resistors Mouser has in stock*, but I'd like to be a bit more certain before taking that plunge.

(*I might shed a tear should a resistor costing $5 apiece accidentally fly out of the tweezers into oblivion. LOL)

stickjam
Learning to Wiggle
Posts: 8
Joined: Thu Jan 23, 2020 11:20 am

Re: o_C drifty - cannot calibrate

Post by stickjam » Thu Jun 25, 2020 8:16 am

pld wrote:
Thu Jun 25, 2020 7:47 am
Yeah, you'll probably get more eyes on it in one of the existing build threads.

The different resistors serve different purposes, and you shouldn't need 5ppm ones. AFAIK people have built them with plain 1% as well (although there are plenty of uncalibrated ones in the wild so that's not saying much). There's been some (> 1 :)) builds with a drift issue but I don't recall them ever posting a resolution; at one point IIRC the theories included fake opamps.

- Drift up or down, how much is "a lot"
- What does the vbias 1.25V look like (between the 47K)
- Re-check soldering and/or clean off flux
Thanks. It's had several IPA baths leaving no detectable flux residue and solder joints are clean and solid under the microscope. vbias is 1.24780 +/- 30 microvolts although it has been powered on all night. Just disconnected it to let it cool down naturally and will check again in a couple hours.

stickjam
Learning to Wiggle
Posts: 8
Joined: Thu Jan 23, 2020 11:20 am

Re: o_C drifty - cannot calibrate

Post by stickjam » Thu Jun 25, 2020 10:46 pm

stickjam wrote:
Thu Jun 25, 2020 8:16 am
pld wrote:
Thu Jun 25, 2020 7:47 am
Yeah, you'll probably get more eyes on it in one of the existing build threads.

The different resistors serve different purposes, and you shouldn't need 5ppm ones. AFAIK people have built them with plain 1% as well (although there are plenty of uncalibrated ones in the wild so that's not saying much). There's been some (> 1 :)) builds with a drift issue but I don't recall them ever posting a resolution; at one point IIRC the theories included fake opamps.

- Drift up or down, how much is "a lot"
- What does the vbias 1.25V look like (between the 47K)
- Re-check soldering and/or clean off flux
Thanks. It's had several IPA baths leaving no detectable flux residue and solder joints are clean and solid under the microscope. vbias is 1.24780 +/- 30 microvolts although it has been powered on all night. Just disconnected it to let it cool down naturally and will check again in a couple hours.
I found the culprit. My R14 was wildly out of tempco spec. It would have made a great, but inaccurate thermistor. Replaced it and It calibrates! :)

Post Reply

Return to “Music Tech DIY”