Xoxbox warble resonance
Moderators: altitude, adafruit_support_bill, adafruit, hamburgers, phono

Please be positive and constructive with your questions and comments.

Xoxbox warble resonance

by faybiens on Thu Jan 05, 2017 9:47 am

Happy new year to everyone!

I just wanted to know what you all think about the warbling resonance topic
Seems to be something the xox don't do because of its stronger PSU vs TB303

If anyone did a successful modification on the xox ?

I tried the R1 mod but it does not work, I had to increase it to kilo-ohm range to get an uncontrollable warbling, self travelling between no warbling to lot of warbling randomly

I understood that it is dependent of the filter opening, the pea-whistle effect increases when filter opens
I saw that the PSU is opamp based but with integrated regulators and without a tiny transformer used as a oscillator core to increase voltage to get 12Vdc from 9Vdc

I am about to test a voltage divider after the diode bridge of the xox PSU, hoping to create 'stable' voltage drops to the op amps of the 5.333 and 12Vdc

Thanks for your feedbacks

Posts: 16
Joined: Thu Jul 09, 2015 5:00 pm

Re: Xoxbox warble resonance

by antto on Thu Jan 05, 2017 3:31 pm

the warbling is not caused by the PSU
it's caused by instability in the VCF

there are about 4 or 5 high-pass filters inside the feedback path of the ladder filter
as a result, the filter becomes 9-pole (but you better read Tim Stinchcombe's analysis if you're interested about this in details) when the resonance is increased, and the response is more like a weird band-pass:
weird, because the resonant frequency of the lowpass ladder is variable (controlled by the cutoff/envmod/accent knobs) while the highpass filters are fixed, and create another resonant frequency which stays fixed too, at around 8Hz

the exact frequency and the "Q" factor of this low-frequency resonance depends on the cutoff frequencies of the highpass filters in the feedback path of the ladder
the circuit is sensitive to the exact values of those capacitors (or effective resistances)
usually it's stable (the Q factor is low, so it doesn't resonate much when excited), but if the values move away from "stable" - the Q factor can easily get high and then the low-frequency resonance would resonate for longer (when excited) and thus its amplitude will be able to get louder too - that's when it can get loud enough to start affecting the supply rails and by stressing the supply rails it causes the main VCF cutoff frequency to also go up and down (and if it goes stronger than that it starts affecting the VCO tuning too)

the low-freq resonance's Q factor can even go beyond self-oscillation, which is just terrible

so, to sum it up: this "warble" is not caused by the PSU or any mythical/rare/vintage components which are ommited from the x0xb0x
it's inherited by the design of the ladder filter, but it is normally in the "stable" region
if it warbles - then it's unstable and should be fixed
and since it's caused by the ladder filter design - any clone which copies that part of the circuit will also inherit it

i don't know why some people have praised the warble as a desired "effect" when it really isn't
i've listened to (and monitored) a whole bunch of acid tracks made with 303s from the 80s 90s, and on, and i've not seen any warbling in 97% of them
i've seen warbling on a few, i don't remember which exactly
i've also seen 303s going broken on recordings (unrelated to the warble, Armando - Land of confusion, in the first take, the 303 is playing broken, the slide disappears intermitently and at one point the sequencer crashes i think (the fart sound))
...but there's not any good examples of classic acid tracks with warbling 303s i can think of

PS: if someone reading this strongly disagrees and thinks the warble is great - i have a x0xb0x which warbles and i can't fix it, i could maybe trade it for a non-warbling one heh

now, if after all you still want to get your x0xb0x into the unstable region - desolder the highpass caps in the ladder, and replace them with sockets
then get a bunch of caps with different values and experiment
you should look up the value of each of them in the circuit and change it just slightly, not drastically
since they are already sitting close to each other in pairs and are very near to the unstable region

Posts: 1535
Joined: Thu Apr 15, 2010 3:21 pm

Re: Xoxbox warble resonance

by faybiens on Fri Jan 06, 2017 4:13 pm

Very interesting thank you.

I did some other experimentations on the xox PSU and here are my thoughts, maybe I am totally wrong

I played with a trimmer at J4 on the IO card and on the 5.333V point, it affects the pitch only
On the 12V, it 'creates' a 16Hz vibration on the res, just like what I could measure on some rare samples of TBs that has strong warble, well enough to be seen on a spectral analysis

I think that adding a resistance on the Xox 12V may affect the stability of the xox filter by introducing voltage drops.
When the filter opens, the vibration warble gets stronger, voltage drops may get stronger when filter fully opens
On my side I have adjusted the trimmer just enough to change the resonance tone without having this extreme pea whistle effect

A lower 12V seems to get the filter of the xox in the instable side

The weak PSU may be a reality, as the 12V on the TB is made by a voltage booster with a very small transformer, a zener and others
I don't think it is just about vintage parts but a certain design and then its global behaviour
Any drops on the voltage booster output will affect the 12V first but should not on the 5.33 as the voltage drop should not be as strong

The voltage drop maybe very small on mainly all TBs and stronger on very few because of certainly component ageing.

The trimmer I connected can adjust in a subtle and 'stable' way the tone of the resonance

Posts: 16
Joined: Thu Jul 09, 2015 5:00 pm

Please be positive and constructive with your questions and comments.