rhythm wolf teardown: any questions?

[Mungo]

There is a generally accepted transient/attack period during which pitch is indeterminate, atop the complicating non-linearity of the envelope modulation.

Looking at the C-E vs C-B current flows there is a reverse beta measurable. I think the methods of calculating instantaneous pitch are possibly leading us astray, as the proposed reverse C-E flow would put the pitch increase on the downward edge of the output. Measuring the peaks of the physical result by hand shows the pitch up being both on the falling edges and proportional to amplitude, which ties in with your reverse conduction theory.

[guest]

thats an interesting question, how much is C-E veresus C-B flow. ill think on that for a bit. ill have to make some guesses as to what the reverse beta is.

[YashN]

Very cool thread!

Especially because, look what I have here and what I started to do:

ATC

The analyses and schematics by @guest can come in super handy.

Any chance you also try your hand at the Tom Cat?

For RW and TC, I think two very worthwhile things to do on top of the individual outs are the following but no one talks about them:

1. Reduce the underlying hiss caused by the Howl/Maul circuit. On my TC, Phones outputs get very hissy as from half way of moving the Maul pot on Max Vol. and for the Mix Outs, as from 3/4 of the way. By default though, if you set the Maul pot to 0, the ATC does a very good job of being very quiet overall. As a secondary mod., I've seen people reduce the Howl range of the pot on some RWs. That can be a good thing too, but I think the hiss added is really bad, way too bad in fact.

2. A separate +/- 12V Linear Regulated PSU for the analogue section. End results can be bulkier, require cutting wires and either two power inputs or a new single multi-wire connector, depending on how you approach it, but this is the kind of thing that can make big changes in the overall sound.

Here's to people like @guest and others who delve in deep and coax a lot of cool enhancements out of existing gear.

[guest]

awesome, best of luck with your digging! i found that the noise on the RW was really bad, and it was due to bleed on a few of the voices. basically, if there was a voice that used white noise to create a sound, that white noise was not adequately shut off when the voice was done. when it goes through the distortion, that noise is amplified a lot.

[YashN]

Thanks, I'm also thinking of potential CV input points if those can be useful at all, say on Pitch(es) or other internal patching potential.

On the TC, I find that the Maul on say, the bass drum tail, makes it sound like 'phrrrrll'. Not very musical to me but some people may disagree. An envelope so that the distortion acts mostly on the attacks would be more useful to me.

I am thinking that if there was a way to make the distortion mixable in parallel, that might be more useful, but not sure.

The noise bleed doesn't seem to be as apparent on the TC, although I haven't compared physically with a RW's.

On the TC, I found that bridging the single Diode in the Distortion network, i.e. not touching the two other ones in series, with a 1K resistor dropped the hiss considerably. The distortion effect might be less too though. Worth a try exploring this area in the RW too.

Some kind soul earlier in this thread posted about identifying the triggers on the RW PCB. I'd like to see that please. Could help me identifying the same on the TC PCB as I wouldn't mind getting one or two triggers for outside gear as well, opens up a lot of possibilities in the analogue world for me.

[guest]

all of the data i have is here:
http://wiki.openmusiclabs.com/wiki/RhythmWolf

there arent any pictures, but if you can locate part numbers, maybe some of them are the same? basically, the triggers ran along the middle of the PCB (horizontally), above the drum pads but below all the analog circuitry.

if youre pulling all the individual voices out, you can then mix the disotrtion in parallel, and maybe put it through a VCA or panning circuit to have it envelope controlled.

[YashN]

These are super handy, thanks, and I think I should manage to trace back some triggers, especially going by the trio of transistors near the Kick drum one for instance and then I might need to set a pattern with 3 velocities and attempt to map that out on the scope or just trace back to the trigger with the DVM on continuity.

Also gone back to your site several times over the past years for various projects, really, really useful info and very well explained so thank you for that too.

I've opted for a switched output myself, so basically, when I connect the Snare out to my mixer for special processing, it is removed from the main mix, so not sure I can do the parallel distortion with it internally on the ATC, but it's OK as I can always add a distortion somewhere, e.g. on Aux 2 on the mixer itself.

So, referring to the TC's Maul hiss, for which the RW's Howl has a corresponding circuit, the diode I was mentioning on the TC corresponds to the RW's D7. I bridged the TC one with a 1K res. yesterday and the hiss was dropped significantly, so give it a try. Not sure I checked if the distortion effect was as pronounced as it was when un-bridged, did that quickly yesterday as I was more interested in making test soldering for the Snare out which then went to my mixer and to a Lexicon LXP-15 I recently bought and restored on Aux 1.

I think the TC's Maul is less interesting for the Kick but more for other sounds, the hats can benefit from the added harmonics for example.

[YashN]

Almost forgot: I did a quick scan of people's RW PCBs and my own TC. The parts are similar or the same, but the numbering differs, e.g. U9 on the RW is something else, like U11, and the layout also differs in places - for instance the ordering of the Instruments' mixing resistors isn't mapped the same in the TC as it is in the RW.

Overall, the general layout is very similar though - they saved a lot by re-using that basic architecture for sure.

[YashN]

Created this so as not to clutter the RW thread: Akai Tom Cat Mods

Included some pics: Main PCB, Maul Section and the Kick for comparisons with the RW.

Checked with your schematics whether I could locate the trigger on the TC Kick.

It looks to me like the Q10, R145, R146 arrangement near P12 (Pitch Env Pot) on the TC is similar to the RW's and I get a small trigger at the top of R146 but not sure, little scope does show a waveform in sync with the sound, but I think it was something like 2.5V or 5V, and no change on accents (might be normal, not sure either).

EDIT: Updated info on the Kick Trigger on the TC is here.

Replies can go to the TC mods thread or here if you don't want to clutter this one further - I'll paste the info there in that case.

Here's the Kick on the leftmost column:

TC Kick

[YashN]

Did I get this right? Is it what you see on the Trigger for the RW?:

The signal I get there is 5ms duration, the amplitude is around 2.5V, with the Top section flat when velocity is High, and slopes down when vel. is med or low

[guest]

for trig, i got a 4ms pulse that went to 3.3V. this then gets combined with the accent, to create an accent pulse of 5ms that varies in hieght with the accent level: 3.4V, 6.8V, 10.2V.

[YashN]

Thanks, guest. I might have it wrong then, but will continue investigating. Tracing that part could be interesting.

EDIT: ah wait, on second thought, that's close enough for the Trigger part which I was looking at. Now I just need to buffer that with a transistor and externalise it on a jack, right?

[YashN]

I don't know if this was mentioned already in this thread: for the RW Kick (potentially other RW instruments?): it can be worthwhile to check what mods were made on the Yocto (it's a clone of the 808 circuits).

I've added some results of explorations for the Attack section of the TC Snare here. Should correspond to the RW in layout despite different numbering.

I think though, that it's the tail end high frequencies which should be tamed in the TC Snare, and the RW is probably similar. A question of balance and perhaps filtering. Looking for a sweet spot to implement, and I'd rather do simple than complex, not cut traces, etc...

[YashN]

ive been looking at the noise at the output as well, and that seems to be a lot of pickup from the switching powersupply, and the gain on the distortion just amplifying the noise floor like crazy. i put a bit of shielding in, and that seemed to help a little bit.

These two are indeed each making their own contributions to background hiss.

Curious as to where you put the shielding - just behind the Volume Pot and that connector socket there?

Here's what I just did tonight to my TC: Twisted that output board with its little cable in place a couple of times. Got rid of the background hiss for Volume at Max positions: nothing I can hear on headphones and everything else is quiet around me.

[guest]

good to know, that sounds like an easy fix! i put some tinfoil sandwiched in paper on top of the entire powersupply board.

[YashN]

good to know, that sounds like an easy fix! i put some tinfoil sandwiched in paper on top of the entire powersupply board.

Thanks, really easy indeed. Actually, I also twisted the two other sets of cables, one yesterday, and the other this morning. The only thing to be mindful about is not to twist too much as the twisting force could damage the cables or the cable + connector assembly.

Good idea about the shielding this way for the PSU.

I have some cool & simple mods going on my TC. It's great as the 'spitty' sound quality at the start of the hats have been tamed, Snare balance is better, Kick is louder, up to some analogue distortion.

I also have a tentative pot for the Snare which gets it lower in frequency. The problem I have with that one is that the whole area where there resonators are nearby seem to have increased in microphony/interference. Not sure I am going to keep it, might have to re-test with a Res in series, or else just remove it altogether. Shame though as the sounds I can get are really, really cool.

Overall mods tame a lot of the top end distortions. Additional reduction of hiss on the Mix Vol. Pot is also good.

What all these end up doing is that the Maul circuit appears much more useful now: because the digititis of the highs have been removed, what the Maul circuit does now adding harmonics sounds rather good instead of even more grating.

Haven't worked on reducing the Maul circuit hiss, although I have a few ideas for how to go about it:

1. One is something I found online by others who did mods for the RW: add either a 22K or 33K bridging Pins 1 and 3 of the Howl Pot. I'd do the same here with a similar value.

2. Replacing that single Diode by a LED or something like that.

3. I might still add another suitable res across that diode or LED.

Having a hard time mapping out the Kick Trigger as an example on the TC. I have a transistor which looks it should go to the Trig signal after a couple of Resistors in a Voltage Divider arrangement as in your Kick schems for the RW.

Instead of that though, I get a lonely Base of the transistor and I can't find where it leads at all.

All this work also make me think that replacing the Analogue Section PSU with a proper Linear Reg. PSU would probably have made all these mods in the TC unnecessary. I do think it's more fundamental.

[guest]

swapping the powersupply is an easy thing to test if you have a bench supply. personally, i think that supply is adding some really high frequency noise, but otherwise not effecting the sound too much. most of my low frequency noise was coming from the niose generators for the hats and snare bleeding through.

[YashN]

swapping the powersupply is an easy thing to test if you have a bench supply. personally, i think that supply is adding some really high frequency noise, but otherwise not effecting the sound too much. most of my low frequency noise was coming from the niose generators for the hats and snare bleeding through.

I should have started from that: I already built a very low noise bipolar 12V for my DIY Analogue. SMPSes always affect all audio frequencies in my experience.

Tonight, I had the idea of attempting a tracing to the back of the PCB, and found that the lone Base does go there.

It goes to a component at the back labelled D45, technically a diode (?), but it has the same three-legged form as the SMD transistors. The other side goes back to a via on the top layer of the PCB once more, and here again, can't trace it more than that for now...

[YashN]

Managed to get a better reading and measurements on the large scope for the Kick Trigger.

I get an amplitude of 3.2 - 3.4V and a duration of 4.10 - 4.20ms

That's more like it!

At least now I am sure that this section of the TC does look like the one you found for the RW regarding the Kick Trigger.

Any idea how to best implement the Trigger Buffer for that going to another jack?

[YashN]

Following our discussions on shielding and cable twisting today, I am thinking that we could potentially need to shield the Maul circuit from the nearby cable (or actually it's shielding the latter from the former or even both). There's also the other end of the cable, shielding the protruding pins could be good too.

Worth a try.

[guest]

do you want the accent out (basically the trig with height variable), or just the trig? the accent would be easy, as just a 1k resistor at the output of the opamp and some protection diodes would do.

[YashN]

do you want the accent out (basically the trig with height variable), or just the trig? the accent would be easy, as just a 1k resistor at the output of the opamp and some protection diodes would do.

Just the Trig so far. Will keep the accent info in mind just in case I need it down the line. Thanks.

[guest]

i think the trig is common to all voices, and the accent is per voice. this synchronises all voices. so there are 2 seperate "trigs" that go to the accent circuit. they make a 2 bit DAC, so either one may not be active depending upon accent level. perhaps the easiest thing to do is to follow the accent opamp output with a transistor/comparator/schmitt trigger, etc to condition it to 5V or something regardless of accent level.

[YashN]

That does sound like the combined Trigger I came across earlier at that point.

Thanks for the advice on the circuit and location, looks like this is what I'd need to do - the output definitely must be clamped to +5V max. I'd really like two different triggers and space definitely is a constraint inside the box.

[YashN]

for trig, i got a 4ms pulse that went to 3.3V. this then gets combined with the accent, to create an accent pulse of 5ms that varies in hieght with the accent level: 3.4V, 6.8V, 10.2V.

At the output of the Op Amp B in the TC, I get 5ms of 3.4V, 7.12V and 10.4V for the three velocities respectively. Now on to the recommended circuit.