Teletype IN returns fluctuating values with constant input

Depending on what I use for the input, the problem gets worse. For example, using Plaits Shades to produce a constant voltage, I get fluctuations from 0 to about +20 millivolts. Using a 4MS SMR in volt/oct mode, the values fluctuate ±50mvs. As a sanity check, I plugged both of these directly into an oscillator’s v/oct tracking and there were no such fluctuations in the pitch.

I’ve been using this script to debug the issue, triggered by the metronome beating at 500ms
A IN
DEL 100: P 0 A - delay here as I read there is some latency when reading from IN

In a test with constant output from Shades saw the values set in the pattern toggle irregularly between 5632 and 5648 (strangely here no other values were read). Values from the SMR ranged from 3271 to 3387 with many other values in between…

I was able to test the teletype & SMR combo each in a different case, in case it was somehow due to the power supply, but still had this issue.

My teletype is running firmware 3.1. Has anyone else seen this issue/can explain what might be going on/possibly have a workaround? I bought my module second hand a few months ago, hoping that it’s not in need of repair

1 Like

The ADC has three effective inputs:

  • the external signal
  • the analog ground (AGND on the schematic)
  • the reference voltage (ADVREF on the CPU, connected to 3.3V)

If the external signal isn’t varying, one of the other two must be causing the reading to change.

3.3V could be unstable if the bypass cap (10u, C44) for it was out-of-spec. Try having a script that’s pushing more data over I2C (maybe updating a CV output periodically) to see if it makes the fluctuation worse.

The variance with source makes me wonder how the fluctuations are affected by source impedance.

1 Like

I tested with @MeguAzu on my Teletype as well.
Same behavior, so it doesn’t seem a hardware defect on their end.

Hey, thanks for getting back to me. I appended
DEL 100: CV 1 A
DEL 100: CV 2 A
DEL 100: CV 3 A
DEL 100: CV 4 A
to the above script, but the fluctuations appear to be roughly the same.

I also just tried setting a CV out to a specific value in the Init script, patched it to the IN jack, read IN on the metronome tick, and there were no fluctuations at all. I’m not sure what source impedance is (had to google it), sounds like a plausible cause?