Thanks for your effort, trying to fix this!
Well, for me it comes down to the jitter problem. As I said, I feed 5 V externally and everything works just perfect (except for the jitter). So if a proper working U 2 does not better the jitter behavior, I would just keep it as it is now and use the new pot-object.
-
Not sure what scale this is, there are no volts in Axoloti one could display in the Patcher, just fractional numbers between -64.00 and +64.00.
-
If you’ve measured voltage jitter at the ADC pins, then if on a scale from 0V to 3.3V jitter was 0.02V, it would translate to roughly 0.02V / 3.3V * 64 = 0.388 Axoloti units (unipolar). A believable number.
-
If you’re saying that the Axocontrol pot values stay e.g. between 32.00 and 32.02 if you don’t move the pot, without any software filtering, then I find that hard to believe to be honest, and we may have caught Tom Whitwell practicing black magic. (In the Axocontrol schematic that is up on Github, there is no hardware filtering on the pots.)
-
That being said, it is normal that the simpler a circuit is, the less noise is introduced. Gills for example has quite low jitter. It really has the pot sent straight to the ADC pin through a quite sluggish low pass filter.
-
As soon as you introduce bipolar power, multiple op amp summing and scaling, and the desire for CV to be as fast as possible, then noise will increase at each stage, unless you do heavy hardware filtering, but then you lose speed and everything will act as if there was a glide on it.
-
On top of that, having the USB connection to the computer may add more noise. Computers have A LOT of high frequency stuff going on, plus more and more bluetooth and wifi connections racing through the aether around them.
Long story short, the ADCs are jittery and not hardware filtered because some use cases will require them to be fast. If precise (or non-changing) values are needed, then software filtering is recommended, like for example the pot smooth objects I have been working on and will publish soon.
I am testing jitter right now with B10k and B50k pots and they behave virtually the same.
Below is a video showing jitter with a shi##y AC power supply and cheap DIY Euro power board sing LM317/LM337 (all B50k pots).
I’ll compare with the Toppobrillo Euro Bus I received, of which they claim to have the lowest noise or something, but I am expecting to see about the same jitter.
I am attaching the patch here:
jitter measure.axp (2.3 KB)
I checked with the multimeter, all those points are connected, but I can also measure 75 ohms to ground!
I thoroughly cleaned the pcb off of any residue from soldering. I am waiting now for it to dry. Meanwhile, here’s a photo of U 2 and surroundings from my microscope, before cleaning.
Again, VBUS_END must be high for the 5V regulator to be enabled.
VBUS_END is being pulled low on your PCB, and the only parts where this could happen are
- PIN 24 ON SOCKET H2 and surroundings (Big Genes PCB)
- matching PIN 24 ON HEADER H2 and surroundings (Core)
- zener diode D8 (Big Genes PCB)
- capacitor C38 (Big Genes PCB).
I can’t see anything fishy on your latest photo,
-
maybe worth checking the zener, D8, with the diode setting on the multimeter. Should measure around 0.6V one way, open circuit the other way.
-
Otherwise CHECK PIN 24 ON SOCKET H2 (Big Genes PCB), PIN 24 ON HEADER H2 (Core), and surroundings.
Good idea too. Going back to your earlier photo, I notice this sort of goop on it, where did it come from?
thats a booger (nah, kidding…) it is a bit of flux from resoldering many spots on the board.
D8 measures 0,1142 V and 1,848 V the other way around with my diode tester.
D7 measures 0,1151 V both ways on my diode tester.
My fault. Of course I didn’t mean Volts, but the numbers between -64.00 and 64.00
Now I am only waiting for the new big genes pot smoother…
I’ve created
- a set of pot denoise * objects for Gills and Big Genes, respectively
- GPIO-style gpio/in/analog denoise * objects
- a general-purpose kfilter/denoise object
Sync libraries then search for “denoise” and you’ll be ready to go. let me know if anything is off!
1 Like
THanks for the quick solution. I tried it and it works great with threshold 19 for pots 1-4. For some reason pots 5-8 need more than double the threshold, 42.
Another indicator for bad grounding.
