I just have built a Big Genes Ksoloti, which works fine, except for the fact, that the all pot outputs are really jittery. I got the 0.7 version which came with 50k pots. I tried to use pots, while tinkering with the bass_FX patch from the library. It completely destroys the sound, as the jitter works on the parameters. I tried all sorts of smothing, but to no avail. This way the pots on the thing are completely useless…
Any other trick I can pull? Except for putting in 10k pots, Which is imposible anyway…
Typical raw jitter could reach around +/-0.2, depending on power supply grounding, and so on.
The pots give you the raw unprocessed ADC measurements so yeah they’re jittery. In case you have ever tinkered with an Arduino or Teensy or similar, Big Genes ADC values are already RC filtered and an order of magnitude quieter than those. However smoothing is up to the patch creator, depending on how the pots are used.
You’re going to have to add smoothing, e.g. try drj/math/mov_avg, or the MultiPot 3 object by TSG which has a deadband attribute. Or look into other patches like the recent Ksoloti Gills demos how the pots are handled there.
10k pots are no magical solution, the jitter might be a bit less, but really just a bit.
I will try your smoothing suggestions, before I isolate the Genes to a empty skiff. At the moment the Genes lives in a Rack with a EigenMatrix (the only other that draws 5V) and a couple of other modules, powered by a old Doepfer +/- 12V powersupply and a 5V wallwart that delivers 2,5 A of 5V.
I will let you know ho things work out…
best,
Omnichordman
I tried those and they didn’t help at all. Meanwhile the Genes has been alone in a 4MS Skiff with it’s solid power-supply. Still all the same. So I guess it is not a power/ground problem
I thought this to be the perfect companion to my EaganMatrix. Now I am in doubt…
Here is a WIP for proper pot smoothing out of the box.
Note that it (currently) comes with a few percent DSP load penalty.
Thinking about replacing the actual pot objects for Gills and Big Genes with these to avoid confusion. If raw pot values are desired, a set of pot raw objects could be added.
@Omnichordman could you check if it works for your Big Genes? You can play with “threshold” and “smooth” to find a value that works for you.
Yes, need to revisit my soldering. And I just noticed that the Ksoloti Core is not powered, when the PROG USB port is not used. Have I missed something there?
That’s good news! Note that this is hard coded to pot 1 and needs some work, I’ll upload it as library object when it is ready.
What is your issue with 5V? By the way Big Genes produces its own 5V - it has a 10-pin Euro header, not a 16-pin one, so 5V from the Euro bus is nowhere to be drawn.
The Core and everything else on Big Genes should be powered by your Eurorack power in any case.
Only if the Euro case power is off and you connect PROG USB to your computer or anything, Big Genes should power up in something like “digital only” mode, drawing 5V via USB. Which means you can upload patches and access the SD card etc. but pots 1-4, audio, and CV are not going to work because there is no Euro power.
I just resoldered all relevant 5V, 3,3V and ground pins, and tried to find any point with 5V. Nowhere to be found. I measured around U2 and U5 and could not find 5V or 3.3V. Defective voltage regulator?
Anyway, I found that I can simple use 5V from the rack power supply and connect it to the left side connector of the Core board. Works great but I don’t know how it would influence the jitter behavior. Btw. 10v ref is at 9,97V.
Very odd, although 5V from the power regulator U2 does not “manifest itself” until after the inductor, see screenshot, you’re right though, there should be 5V at least at U5.
This is possible, or any other part in that area. You’d need an oscilloscope to find out. Check first if +12V is found at the red areas and at pin 7 of U2.
If you haven’t done so, check if 5V is there without the Core connected.
With and without the Core connected, check if VBUS_END is “high” i.e. anything higher than 0.5V or so. You can check at the VBUS_END pin on the socket for the Core but also checking at pin 5 of U2 is a good idea.
Check if there is anything going on at pin 8 (PH) of U2. I’m not sure what exactly should be “going on” there but I assume it is a high-frequency PWM signal around 500 kHz.
That’s about it so far. Any parts other than U2 and the inductor L1 are unlikely to fail. It is still possible there is a small error somewhere pulling VBUS_END low.
No 5 V an<ywhere or any Signals around U 2. Seems dead. I feed 5 V to the point in the picture I sent, and everything works, except for the heavy jitter. I compared it with my Axoloti with Music Thing Modular Axo Control. The jitter there is +/- 2 digits, like 2,02 - 2,04 V. Thats way better than with the Big Genes.
So, what now? I guess I got a defective Genes circuit-board where U2 never worked.
Including pin 5 which is VBUS_END? VBUS_END must be high or U2 does not turn on. It is like an enable switch.
If VBUS_END is low it must mean it is being (accidentally) pulled low somewhere else. We can suspect the pin header and socket pair, H2.
As you can see in the schematic VBUS_END is pulled high directly from the +12V rail via a 47k resistor, then limited to 3.3V using a zener diode. And +12V you tested is working and present.
Check if VBUS_END is high, and if there is continuity between the header pin and the orange area.
Also check if VBUS_END happens to be shorted (or has low resistance) to GND (which it must not)
Anyway it is possible that IC is really dead, or it could just be badly reflowed. Do any of the pins move when you (carefully) prod them with a needle or small screwdriver? Are you comfortable with reheating the SMD pins?
This problem existed at the SDRAM chip on early prototypes.
They just didn’t do a good job at the factory I was ordering at back then. But I’ve changed suppliers long ago.
So what else can we do:
I’d send you a new PCB but desoldering parts from the old PCB to transfer them to the new is next to impossible. Those parts in turns are Thonk’s so I have no authority over these.
If you’re comfortable desoldering the SOIC-8 package U2 I can send you a tested one (and also a new inductor L1 while we’re there)
