rhythm wolf teardown: any questions?

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Joe., lisa, luketeaford, Kent

Post Reply
User avatar
astrosound
Common Wiggler
Posts: 242
Joined: Wed Aug 06, 2014 11:14 am
Location: New England

Re: rhythym wolf teardown: any questions?

Post by astrosound » Sat Feb 29, 2020 3:09 pm

ihav2p wrote:
Sat Feb 29, 2020 1:28 am
Wow nice work! Let's hear some audio!
https://voca.ro/8aZX5oGOtqq

This demonstrates the three accent levels for the modified snare. You'll notice how loud the noise is relative to the pinged filter. I'm considering attenuating the noise a little before it hits the VCA.

ihav2p
Common Wiggler
Posts: 69
Joined: Wed Apr 28, 2010 2:03 pm

Re: rhythym wolf teardown: any questions?

Post by ihav2p » Sat Feb 29, 2020 4:16 pm

Interesting! I like how it can do short little ghost notes. I don't think the unmodified snare is capable of that. I basically only use the lowest velocity setting only and then crank it. The execution of the mod seems super hairy though. I'm kinda scared to kill something.

vidret
Common Wiggler
Posts: 193
Joined: Sun Jul 30, 2017 10:26 pm
Location: Skogen

Re: rhythym wolf teardown: any questions?

Post by vidret » Sat Feb 29, 2020 4:54 pm

amazing job. Looks like it's gonna be a bit of a hard mod to DIY without a lot of tech skills, but I've always wondered what the hell happened when this came out.

Please do post a video when you're done modding it :sb:

User avatar
guest
Super Deluxe Wiggler
Posts: 4020
Joined: Mon Aug 19, 2013 11:49 am

Re: rhythym wolf teardown: any questions?

Post by guest » Sat Feb 29, 2020 6:10 pm

wow, thats quite an improvement. reducing R134 might be the easiest way of attenuating the noise level, but it might lead to clipping. you could even put a pot in there and have a variable level. putting a resistor in series with the cap going to the base of Q24 would probably be a less a distorted way of attenuating.
openmusiclabs.com

User avatar
astrosound
Common Wiggler
Posts: 242
Joined: Wed Aug 06, 2014 11:14 am
Location: New England

Re: rhythym wolf teardown: any questions?

Post by astrosound » Sat Feb 29, 2020 10:50 pm

ihav2p wrote:
Sat Feb 29, 2020 4:16 pm
The execution of the mod seems super hairy though. I'm kinda scared to kill something.
vidret wrote:
Sat Feb 29, 2020 4:54 pm
Looks like it's gonna be a bit of a hard mod to DIY without a lot of tech skills
With some good light, a pair of tweezers, and modest soldering skills it's not that bad. Took maybe a couple hours, most of which was spent planning where the parts would go and what traces to cut/bridge. Removing the 0603 SMT parts is pretty easy, just put a blob of solder big enough to bridge either side with the tip of the iron and it comes right off. You can use regular through-hole parts for any the replacements. I only used a couple SMT parts because I had em on hand. Where wire is needed I used solid 24AWG tinned copper, bending little 90 degree feet where they contact the solder pads. The kapton tape in the first photo below is helpful but not necessary.
Image
Top side view of snare mod. Notice the middle leg of the pot soldered directly to the top of R213, right in front of it. Care must be taken so the two lifted legs do not touch their old solder pads.
Image
Bottom view of snare mod
Image
Kick modification

Currently working on the VCF. The volume drop with increase resonance is annoying so I'm rerouting the resonance pot.

User avatar
astrosound
Common Wiggler
Posts: 242
Joined: Wed Aug 06, 2014 11:14 am
Location: New England

Re: rhythym wolf teardown: any questions?

Post by astrosound » Sun Mar 01, 2020 10:31 pm

Addressed the VCF resonance volume drop by relocating the resonance pot. It (P19) was replaced with a fixed 20k resistor. The liberated pot was then wired as a voltage divider, the input of which comes off the buffer output of the first filter stage. The wiper output is sent to the noninverting input of U16A via an 8k resistor. The new resonance control doesn't kill all resonance at minimum and doesn't do the screaming squelching laser oscillation at maximum, but the volume is consistent. This is probably far from the ideal way to do it but it suits my needs.

Considering a crude expo converter for the VCF but it's a low priority.
guest wrote:
Sat Feb 22, 2020 8:10 pm
removing R56 and R57 would eliminate it completely (although the bias level on U1B would have to be adjusted). leaving R56 in would change the compression to expansion. when the voltage across R73 and R76 is less than 0.6V, R56 is not active. but, above this level, it becomes active and draws more current than R73+R76, therefore increasing the gain. it might be an interesting effect. the DC bias on U1B would probably still need to be changed, though. those transistors could also be repurposed to make a perfectly linear control current by wrapping them around U1B, or just using it as a current mirror like a 3080 input.
OK, so went the current mirror route and shorted R77 & R56, and removed R73, R76 and C16. I was planning on putting a trimmer in to adjust the DC bias accordingly, but first I wanted to fire up the synth to see how it sounded. Much to my surprise it sounded fine despite no change to DC bias, though I'm not yet ruling out adding a trimmer. As expected, the lack of CV compression means the accent level has greater impact on the volume. It's nice. The lowest accent level will be useful for adding echos to higher accent notes. And of course the decay sounds much better now that the envelope isn't getting mangled.

Comparing the pre and post VCA signals, I noticed a considerable amount of soft clipping post VCA that was independent of accent level. I think the soft clipping is where some of the character comes from but it was a little excessive to my ears. At high resonance settings (now at proper volume) those nice peaks were getting squashed. I increased R58 to 80k to further attenuate VCA input signal, which tamed the clipping a bit.

Moving on to the envelopes next. My goals are to reduce the 100-200ms(?) flat spot at the peaks, and to get the decay times for VCA and VCF better matched across the range of the decay knob.

This is fun!

User avatar
guest
Super Deluxe Wiggler
Posts: 4020
Joined: Mon Aug 19, 2013 11:49 am

Re: rhythym wolf teardown: any questions?

Post by guest » Mon Mar 02, 2020 12:21 am

awesome work! some things for the VCF:

1. how hot is the signal coming out of the VCF? the gain of the VCF is set by the P19/(R144+Rs) ratio, where Rs is the source impedance (R43 or R67//R79, depending upon the waveform). so putting a 20k in for P19 would give the VCF a gain of 2 at DC, which might be too much. a 10k might be better here.

2. the Q is set by R174, R124, and R144. it might be good to increase R124 to 50k, as this will give less distortion on the U16A buffer (the current in the buffer will vary less with amplitude), although maybe some distortion isnt bad. this will lower the max Q. putting a resistor from the U16A buffer output to the U16A noninverting input should also decrease the resonance. so if you want to tweak the range of the reso knob, set the reso knob to MIN, then adjust the resistor to the noninverting pin till reso is at the lowest level you want. next, turn the reso knob to MAX, and adjust R174 till max reso is where you want it.

VCA:

1. i just noticed that your mods make a really nice wilson mirror for the VCA - much better than my suggestion.

2. the drive on the VCA is probably 5x what it should be, so knocking down the input is a great idea. if the input signal is +/-2.5V, then you have a hundred millivolts or more at the diffpair input, which is a lot.
openmusiclabs.com

User avatar
astrosound
Common Wiggler
Posts: 242
Joined: Wed Aug 06, 2014 11:14 am
Location: New England

Re: rhythym wolf teardown: any questions?

Post by astrosound » Mon Mar 02, 2020 1:50 pm

guest wrote:
Mon Mar 02, 2020 12:21 am
1. how hot is the signal coming out of the VCF? the gain of the VCF is set by the P19/(R144+Rs) ratio, where Rs is the source impedance (R43 or R67//R79, depending upon the waveform). so putting a 20k in for P19 would give the VCF a gain of 2 at DC, which might be too much. a 10k might be better here.
Didn't write it down last night, but I believe it was around +/-5V with resonance at minimum. I've swapped the 20k for a 10k which brought it down to approximately +/-2.5V.
2. the Q is set by R174, R124, and R144. it might be good to increase R124 to 50k, as this will give less distortion on the U16A buffer (the current in the buffer will vary less with amplitude), although maybe some distortion isnt bad. this will lower the max Q. putting a resistor from the U16A buffer output to the U16A noninverting input should also decrease the resonance. so if you want to tweak the range of the reso knob, set the reso knob to MIN, then adjust the resistor to the noninverting pin till reso is at the lowest level you want. next, turn the reso knob to MAX, and adjust R174 till max reso is where you want it.
Excellent, thanks. I increased R124 to 40k as I don't have any 50k 0603s handy. One question that's best accompanied by a picture (phone camera was more expedient than a photo editor):
Image
this is where I'm at now. Lowering the resistor going from pot wiper to the noninverting input *increases* max resonance. Below 8k or so (with Res. pot at max) the signal becomes an unstable garbled mess presumably due to all the positive feedback. Under what circumstances would that resistor decrease Q? Or were you suggesting it would decrease Q vs having no resistor, ie with the wiper directly tied to U16A noninverting input?
1. i just noticed that your mods make a really nice wilson mirror for the VCA - much better than my suggestion.
I mistakenly thought thats what you were suggesting by mention of the 3080 mirror. But the wilson mirror is used in the 13700 correct?

User avatar
guest
Super Deluxe Wiggler
Posts: 4020
Joined: Mon Aug 19, 2013 11:49 am

Re: rhythym wolf teardown: any questions?

Post by guest » Mon Mar 02, 2020 2:15 pm

astrosound wrote:
Mon Mar 02, 2020 1:50 pm
Lowering the resistor going from pot wiper to the noninverting input *increases* max resonance. Below 8k or so (with Res. pot at max) the signal becomes an unstable garbled mess presumably due to all the positive feedback. Under what circumstances would that resistor decrease Q? Or were you suggesting it would decrease Q vs having no resistor, ie with the wiper directly tied to U16A noninverting input?
after i wrote that, i kept turning it around in my head, and wondering if i got it right. i knew there was something wrong. ok, first off, keep the wiper going through a resistor to the inverting pin (so reso control happens at the inverting pin). second, you might not need the resistor to the noninverting pin. check your min/max reso ranges, and if youre happy with them, then its all fine. if not, turn the reso pot so the wiper is at ground (this should be max reso). if this level of resonance is not enough, then add a resistor from the U16A buffer output to the noninverting pin. making this noninverting resistor smaller makes the Q go up (as such, no resistor - or infinite resistor - gives no additional resonance). next, turn the reso so the wiper is at the top of pot (this should be min reso). if this is not low enough, decrease the resistor from the pot wiper to the inverting pin.

i think i had gotten confused over what "min" meant for the pot, as min reso is actually max pot setting, and vice versa, so everything got flipped in my head.
I mistakenly thought thats what you were suggesting by mention of the 3080 mirror. But the wilson mirror is used in the 13700 correct?
yes, the 13700 uses the wilson, which is an improved version. R56,77 could probably be left in, and it would work the same for the majority of the range. i think the initial turn-on phase might be less steep (which is worse in my opinion).
openmusiclabs.com

ihav2p
Common Wiggler
Posts: 69
Joined: Wed Apr 28, 2010 2:03 pm

Re: rhythym wolf teardown: any questions?

Post by ihav2p » Mon Mar 02, 2020 4:12 pm

ok, first off, keep the wiper going through a resistor to the inverting pin (so reso control happens at the inverting pin). second, you might not need the resistor to the noninverting pin. check your min/max reso ranges, and if youre happy with them, then its all fine. if not, turn the reso pot so the wiper is at ground (this should be max reso). if this level of resonance is not enough, then add a resistor from the U16A buffer output to the noninverting pin. making this noninverting resistor smaller makes the Q go up (as such, no resistor - or infinite resistor - gives no additional resonance). next, turn the reso so the wiper is at the top of pot (this should be min reso). if this is not low enough, decrease the resistor from the pot wiper to the inverting pin.
Could you do a drawing for this? I have been following along and now i just returned everything to the way it was in the original schematic.

Heres a dirty hack mod for the noise gate time (on the percussion voice). Replace the lowest leg of the noise pot with a diode with the cathode facing down, i.e. towards the noise source. Then connect R6 with a resistor to the other end of the diode. I chose 500 ohms but the resistor value sets the gate length. You could use a pot. The original noise pot fades between the perc and noise.

It's based on the swing vca idea i've read about here and on electro-music.

User avatar
guest
Super Deluxe Wiggler
Posts: 4020
Joined: Mon Aug 19, 2013 11:49 am

Re: rhythym wolf teardown: any questions?

Post by guest » Mon Mar 02, 2020 5:34 pm

ok, here is a drawing:

Image

basically RA and RB are the 2 resistors that would set the min/max of the reso, and RA might not be needed.
openmusiclabs.com

ihav2p
Common Wiggler
Posts: 69
Joined: Wed Apr 28, 2010 2:03 pm

Re: rhythym wolf teardown: any questions?

Post by ihav2p » Tue Mar 03, 2020 5:15 am

...
Last edited by ihav2p on Tue Mar 03, 2020 2:30 pm, edited 4 times in total.

User avatar
astrosound
Common Wiggler
Posts: 242
Joined: Wed Aug 06, 2014 11:14 am
Location: New England

Re: rhythym wolf teardown: any questions?

Post by astrosound » Tue Mar 03, 2020 1:44 pm

guest wrote:
Mon Mar 02, 2020 2:15 pm
after i wrote that, i kept turning it around in my head, and wondering if i got it right. i knew there was something wrong. ok, first off, keep the wiper going through a resistor to the inverting pin (so reso control happens at the inverting pin). second, you might not need the resistor to the noninverting pin. check your min/max reso ranges, and if youre happy with them, then its all fine. if not, turn the reso pot so the wiper is at ground (this should be max reso). if this level of resonance is not enough, then add a resistor from the U16A buffer output to the noninverting pin. making this noninverting resistor smaller makes the Q go up (as such, no resistor - or infinite resistor - gives no additional resonance). next, turn the reso so the wiper is at the top of pot (this should be min reso). if this is not low enough, decrease the resistor from the pot wiper to the inverting pin.
Thank you for clarifying!


I got impatient and put the RW back together last night. Further modification is intended (ihave2p's hihat pot rewiring sounded interesting), but I'm having a lot of fun with it right now. Below is a list of modifications I've done thus far, some taken from guest's notes/suggestions earlier in the thread. Some values shown leave much room for personal taste, so this is by no means an assertion of The Best Way To Do It, just what I happened to settle on.

Synth:

Reduce VCA env decay time to more useful range
Replace C147 (22uF) with 2.2uF
Reduce minimum VCA env decay
replace R265 with 300R-1k (forgot to write down the value I used)
Reduce minimum VCF env decay
replace R96 with 10k
Reduce maximum VCF env decay, bend taper of VCF decay knob
put 300k resistor across P26B (VCF decay pot, outside legs). This one is highly subjective and is worth experimenting with anything from 200k-750k
VCA envelope AD/AR switch
Cut trace on gate side of R311, wire switch to tie R311 to either gate or ground (or wire resistors to the switch and remove R311)
Eliminate VCA CV compression
short R77 & R56, remove R73, R76. This makes the VCA output actually resemble the exponential decay of the envelope controlling it. Also accent will have much greater impact on synth volume.
Reduce overdrive at VCA input -
replace R58 with 50k-100k. Higher resistance further attenuates input signal thus reducing overdrive. I wouldn't do this until after you've modified the VCF/resonance as that can impact the level sent to the VCA input. I went with 80k and slightly regret it as the soft clipping is largely gone now.
increase maximum VCF cutoff
replace R169 & R123 with 10k.
Slightly increase range of cutoff knob
short R204, R226
Improve VCO stability
Remove R140, R94. Tie now-unconnected leg of Tune pot to ground. These two modifications will raise the VCO frequency considerably, and it must be brought back down. This was accomplished by replacing R104 with a 6.8k resistor in series with a 5k trimmer. The trimmer is adjusted to the satisfaction of the tuning utility and shouldn't need to be touched after that.
reduce VCO shutoff threshold
replace R11 with 100R (OR increase R10 accordingly)
VCF resonance
see posts above, I posted a photo of what I did but it's not ideal

Kick
Replaced R205 with 1M, C98 with 0.1uF. Short R188. These are the same values as the 808. It diminishes the pitch rise you hear on long decay settings. See guest's comment earlier in the thread. You can also just swap R205 and R188 and leave C98 alone.

Snare
reduce decay by replacing C106 to 3.3uF.
rewired noise VCA and noise level pot to more closely resemble the 808 snare circuit. See my comment earlier in the thread. This kind of warrants a drawing, but in the meantime pull up the 808 snare schematic and compare.

Note: On snare schematic, 0 ohm resistor on bottom leg of P15 (noise pot) is erroneously labeled R198. That resistor is actually R207.

Other stuff I'd like to try: VCO pulse width pot, do something with the perc voice noise, adjustable pitch of hihat digital noise, hihat choke, separate hihat decay knobs (maybe), tame the "howl" circuit.
Last edited by astrosound on Tue Mar 03, 2020 3:35 pm, edited 2 times in total.

ihav2p
Common Wiggler
Posts: 69
Joined: Wed Apr 28, 2010 2:03 pm

Re: rhythym wolf teardown: any questions?

Post by ihav2p » Tue Mar 03, 2020 3:07 pm

hey astrosound, this is really helpful. Your input has been really helpful for me to learn and the sound of my machine. Thank you for taking the time to put it all on one post.

User avatar
guest
Super Deluxe Wiggler
Posts: 4020
Joined: Mon Aug 19, 2013 11:49 am

Re: rhythym wolf teardown: any questions?

Post by guest » Tue Mar 03, 2020 4:04 pm

great list! and thanks for the catch on R198, ill fix that asap.
openmusiclabs.com

User avatar
astrosound
Common Wiggler
Posts: 242
Joined: Wed Aug 06, 2014 11:14 am
Location: New England

Re: rhythym wolf teardown: any questions?

Post by astrosound » Tue Mar 03, 2020 5:52 pm

ihav2p wrote:
Tue Mar 03, 2020 3:07 pm
hey astrosound, this is really helpful. Your input has been really helpful for me to learn and the sound of my machine. Thank you for taking the time to put it all on one post.
my pleasure. Credit largely goes to guest for tracing the circuit and posting schematics.

pleased to find there's at least...three(?) other RW fans on the internet. :razz: Hopefully more people will be cracking their RW open now that the ice has been broken and schematics are available.

resynthesize
Wiggling with Experience
Posts: 296
Joined: Tue Nov 01, 2016 1:29 pm
Contact:

Re: rhythym wolf teardown: any questions?

Post by resynthesize » Tue Mar 03, 2020 11:01 pm

ordered a rhythm wolf to hack on this weekend based on this thread. thanks everyone!

User avatar
guest
Super Deluxe Wiggler
Posts: 4020
Joined: Mon Aug 19, 2013 11:49 am

Re: rhythym wolf teardown: any questions?

Post by guest » Wed Mar 04, 2020 12:19 am

awesome! ill have to start actually making mods on mine soon. i wanted to do a decent video first of what it sounded like, so i could do a nice A/B. i dont have good recording equipment, though, and am trying to round something up.
openmusiclabs.com

ihav2p
Common Wiggler
Posts: 69
Joined: Wed Apr 28, 2010 2:03 pm

Re: rhythym wolf teardown: any questions?

Post by ihav2p » Wed Mar 04, 2020 5:09 am

I tried the filter res mod according to guest's latest diagram. After trying several resistors, I found that 5k at RB and 20k at RA gave a workable range. The problem is that high resonance does not really sound that good; I looked at a recording in a spectrogram and there are 2 resonant peaks. You can actually get formant-y sounds out of it.

The upside is that the resonance doesn't lower the signal.

So I will be eager to see what guest does with the filter Res!

User avatar
guest
Super Deluxe Wiggler
Posts: 4020
Joined: Mon Aug 19, 2013 11:49 am

Re: rhythym wolf teardown: any questions?

Post by guest » Wed Mar 04, 2020 2:47 pm

does high resonance sound worse than before?
openmusiclabs.com

ihav2p
Common Wiggler
Posts: 69
Joined: Wed Apr 28, 2010 2:03 pm

Re: rhythym wolf teardown: any questions?

Post by ihav2p » Wed Mar 04, 2020 2:53 pm

Yeah unfortunately I think it sounds worse. Here's a recording:

https://voca.ro/iFkgwNU7Pzy

I found this schematic with a youtube video that sounds pretty good.

http://electro-music.com/forum/topic-34426.html

I noticed the resonance feeds back to both the non-inverting and the inverting inputs equally. I don't feel knowledgable enough to implement this strategy in the RW but maybe it will be helpful when you start working on yours.

This has been a crazy crash course in OTAs and op-amps for me, though!

User avatar
guest
Super Deluxe Wiggler
Posts: 4020
Joined: Mon Aug 19, 2013 11:49 am

Re: rhythym wolf teardown: any questions?

Post by guest » Wed Mar 04, 2020 3:31 pm

yeah, whats up with that wierd rumble/warble?? im thinking that maybe AC coupling the feedback might help.

as for the electro-music schematic, thats pretty much the same thing, except with some diodes thrown in to limit resonance.
Last edited by guest on Wed Mar 04, 2020 4:20 pm, edited 1 time in total.
openmusiclabs.com

ihav2p
Common Wiggler
Posts: 69
Joined: Wed Apr 28, 2010 2:03 pm

Re: rhythym wolf teardown: any questions?

Post by ihav2p » Wed Mar 04, 2020 3:36 pm

ass for the electro-music schematic, thats pretty much the same thing, except with some diodes thrown in to limit resonance.
i had kinda guessed it was close. i'm sure it will be possible to get the filter sounding right ...

User avatar
astrosound
Common Wiggler
Posts: 242
Joined: Wed Aug 06, 2014 11:14 am
Location: New England

Re: rhythym wolf teardown: any questions?

Post by astrosound » Wed Mar 04, 2020 9:14 pm

ihav2p, have you increased R58 yet? The overdrive to the VCA is quite significant and could be masking the sound of the resonance. After I did it, the saw wave suddenly sounded like a normal saw wave. Although I did suggest replacing R58 after modifying the filter, it might be worth increasing even temporarily until you get the resonance where you want, then tweak it afterwards. Also (as you're probably well aware) the stock maximum cutoff is kind of low, so that might contribute to the weird sound if you're more accustomed to expo VCFs with wide cutoff ranges. That being said, I kind of know what you mean about the formant-y vibe. Though I kind of like it to be honest.

ihav2p
Common Wiggler
Posts: 69
Joined: Wed Apr 28, 2010 2:03 pm

Re: rhythym wolf teardown: any questions?

Post by ihav2p » Thu Mar 05, 2020 2:43 am

Thinking about distortion, I got rid of the warble by boosting R124 to 50k, but the resonance still leaves a lot to desire. I probed the output of the lm13600 with some headphones and it sounds just the same so I don't think the problem lies at the VCA.

It's gotta have at least some squelch! Here the tone seems to stay constant over the whole note, with only a sharp boost following cutoff frequency, if that makes sense.

Post Reply

Return to “Music Tech DIY”