Digital FX troubleshooting

i must have messed something up on the digi FX board, as there’s absolutely no sound on the output.
probing around i just found the raw OSC sound and some other, deeper signal (IC2 PIN3(=pin12)) but nothing after 328.
i’ve also swapped the DAC but no change.

there’s a very very low level osc sound on pins 3 and 10 of the IC1 that you probably can’t hear in the video.
also there’s the osc sound on pin 23 of the avr (video)

my voltage is 4.91V, could be too low for the 328? the digital control board works with that level.
using firmware 0.93 with dsp setting.

ok here’s a short video that shows where the sound is:
fx probing

The 328p runs as low as 2.7V

  • Is the TX pin on the digital board connected to the RX pin on the filter board?
  • Has dsp been selected in the filter board selector?
  • Does the green and red LEDs light up when a note is played?

i’ve tested all cable connections and they work, went over a lot of solder points, swapped crystal and tried several cap values there, hmmmmmmmmmmm. midi works fine too btw (it’s no newly assembled digital ctrl board anyways)

oh, just read your post.
i’ll hook up some leds and check the RX/TX connection. as far as i remember they had both an audible sound yesterday, will check that.
yup it’s set to dsp.
thanks, will report soon!

TX/RX: yeah physically connected (beeper), but i’ll check between the spots they both lead to, later.

If you get something at pin 23 (ADC input) of the AVR, but nothing at pin 1 of the DAC, then it’s a digital problem. “Listening” to signals is more of an “analogue” troubleshooting method.

Things that could go wrong:

  • AVR not running, hence not recording any audio: red LED won’t flash.
  • AVR recording audio but not getting the right settings for the filter board: red LED ok, green LED won’t flash.
  • AVR generating the correct audio output but bad connection with the DAC: red and green LED ok, but no sound on pin 1 of the DAC.

If you have a scope, you can also scope the pin labelled >TX on J7 - you should see a squarish waveform, the duty cycle of which depends on the complexity of the selected effect (it must change with the selected effect).

Also, if you have a programmer, try talking to the chip with it (for example, just try to read the content of the chip). You’ll know if there’s a problem with the AVR (damaged chip?) and crystal if the programmer can’t talk to it.

red light flashing in time with midi notes on, env2 release and gate length have impact.
the green light doesn’t light up at all.
that would be case 2 then, which means the AVR is running, at least
i’ll check the RX/TX then.

when i touch the TX header, or pin 17 of the 644, or pin 2 of the 328 it seems to trigger a random setting / mix of tx and osc maybe.
some times there are different effects, filtered sound, a mix of a constant pitched sound and the sequenced melody, also what sounds like a loop. sometimes you hear fast switching while the green led is lit (lights up rythmically too, if the effect decides to appear)

but i can never change anything that would change sound or effect in realtime.
first i thought it’s a bad soldering or broken trace at the output expansion, but reworking the connections didn’t change anything.
when i poke at one of the solder points it seems to just trigger random bugs :slight_smile:
the probe isn’t connected to anything.

i haven’t made a scope test yet, will follow now.
could it be a ground issue maybe?

or maybe a too long cable (15cm)… anyway, i have to pick cherries now :smiley:

you have cherries ???

yeah the sour type are ripe now

So it’s definitely the serial connection between the two boards. The FX board receives filter/FX parameters and control values on this serial link - with no connection it stays at a cutoff and VCA gain of 0. Every time you touch the pin, it seems to cause a random signal parsed as a random sequence of 0 and 1 - causing the filter board to be set at a random combination of gain/FX/filter.

This could be:

  • Older firmware version running on the digital board. When was the last time you updated the firmware on the digital board? (To be sure I have updated all the firmware files online
  • Damaged TX pin on the AVR on the digital board. You need to scope the TX output.
  • Damaged connection. Scope pin 2 of the AVR on the filter board to check if you see the same signal.

yesterday evening, it’s the one from github that’s still called beta, so i’m not 100% sure if it’s the correct one.

Firmware version should be fine, I’ve just removed the “beta” from the name.

ok, i’ve built the SVF and that guy works just fine with the same ctrl board / AVR and cables.

i guess that means the issue must be on my digital fx board then, because the AVR’s TX pin should work, when i can control the SVF, or am i wrong here? or could it still be the AVR?
i’ll do the scope job tomorrow, also flash an AVR from reichelt tomorrow.

No, this could be a problem with the control board TX pin not working.

The SVF filter board is controlled through the shift register output (Q7’ / CLK / EN) - a 74595 is used to “store” the 6 filter configuration bits ; not through the serial connection (TX).

ok, thanks.
i’ll check that tomorrow.
man, the svf is such a cool filter!

what voltage swing should the square wave from the control AVR have?
i don’t get really readable results, but that could also be related to personal unability in reading a scope.

mine is an old analog one, 0-10MHz, i guess that should be enough for that task?

i see a very small kind of triangle wave, but distorted. should i rather film it or just swap the chip?

10 MHz should be fine. What you should see, on the TX pin of the control board, is a signal which is most of the time at +5V with “bursts” of digital data every ms approximately. If you zoom on those burst, you should see a mix of individual 0 and +5V pulses, and changing the cutoff or resonance should change their pattern.

Did you probe the TX pin of the control board, or directly pin 17 of the ATmega644p?