[Project] Turing Machine Mk2

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

Moderators: Kent, Joe., luketeaford, lisa

Post Reply
Quinie
Learning to Wiggle
Posts: 40
Joined: Thu Jul 09, 2015 3:14 pm
Location: Netherlands

Re: [Project] Turing Machine Mk2

Post by Quinie » Tue Sep 01, 2020 9:28 am

bostonguy wrote:
Thu Jul 16, 2020 3:49 pm
It is VERY IMPORTANT to use genuine quality semiconductors on the Turing. I would emphasize to anyone building from panel/pcb to source your semis from a reputable source. If you're using a bunch of cheap ebay/aliexpress TL074's that work fine for you in the audio path of other builds, you could find it doesn't work in the Turing. ......
Good advice above.

Just replaced mine with some tl074 from TME. This seems to work in my first test. I think I got it calibrated.
Will do some more test. It was still pretty precice but doable..

Used a high pitched square wave.

skinpop
Learning to Wiggle
Posts: 7
Joined: Tue Sep 15, 2020 12:46 am

Re: [Project] Turing Machine Mk2

Post by skinpop » Sun Oct 11, 2020 9:55 am

I built mine a few weeks ago(full kit from thonk). It works fine I think, though I also had the calibration issue where the "usable" range only seems to be the last 1% or so of the trimmer.
Speaking of range, what kind of range should the big randomness knob have on a well functioning turing machine? Everything after ~2:30 is pretty much locked for me. I expected a bit more room to play with but maybe this is normal? I've tried pushing the calibration a bit further but that seems hopeless.

Pitop
1-Post Wiggler
Posts: 1
Joined: Sun Oct 18, 2020 7:16 am

Re: [Project] Turing Machine Mk2

Post by Pitop » Sun Oct 18, 2020 7:21 am

Hi guys.
I get my Turing machine kit as a full kit from Thonk.
I build it. Check it re check it re re check it everything looks fine. but I can't get it to work.
When I plugged it in and power it all LED are light up and none of the pots or control change anything... any tips or directions to go further ?


Best

User avatar
Thonk Support
aka Monobass
Posts: 8818
Joined: Fri May 29, 2009 5:39 pm
Location: Brighton, UK

Re: [Project] Turing Machine Mk2

Post by Thonk Support » Sun Oct 18, 2020 11:43 am

@Pitop

Best thing to do to start with is take the faceplate off, seperate the two PCBs and send good photos of both sides of both PCB's to support@thonk.co.uk
Thonk - CLICK HERE - Modular Synth DIY + Eurorack Accessories Store

AS of October 2020 this account which used to go by the name of Monobass is now manned by various staff at Thonk.

User avatar
rduif
Learning to Wiggle
Posts: 22
Joined: Thu Sep 06, 2018 8:45 am
Location: Rotterdam, Netherlands

Re: [Project] Turing Machine Mk2

Post by rduif » Tue Oct 20, 2020 1:59 pm

Ok. After a long while not looking at my TM, I have tried for the last few days to get it to work again. The results are... mixed.

I changed a bunch of the IC's and such to make sure I used as much as possible from reputable sources. Right now only the cd4015 and tl431 are from aliexpress (so I guess the first suspects?). I got the module to lock sometimes, but in general it is behaving very inconsistently, but in a wrong way and not in a 'random sequencer' kind of way.

I noticed particularly that the length setting and the amount of steps in the sequences do not match at all, for CW nor CCW. They are usually uneven and never the expected value. Would this indicate a problem in the shift registers? I have tried looking at the schematics but Im a beginner and an amateur so dont have the expertise required to figure this out. I hope someone can point me in the right direction.
Thanks!

User avatar
emmaker
Veteran Wiggler
Posts: 711
Joined: Sat Mar 10, 2012 5:07 pm
Location: PDX

Re: [Project] Turing Machine Mk2

Post by emmaker » Tue Oct 20, 2020 2:42 pm

rduif wrote:
Tue Oct 20, 2020 1:59 pm
Ok. After a long while not looking at my TM, I have tried for the last few days to get it to work again. The results are... mixed.

I changed a bunch of the IC's and such to make sure I used as much as possible from reputable sources. Right now only the cd4015 and tl431 are from aliexpress (so I guess the first suspects?). I got the module to lock sometimes, but in general it is behaving very inconsistently, but in a wrong way and not in a 'random sequencer' kind of way.

I noticed particularly that the length setting and the amount of steps in the sequences do not match at all, for CW nor CCW. They are usually uneven and never the expected value. Would this indicate a problem in the shift registers? I have tried looking at the schematics but Im a beginner and an amateur so dont have the expertise required to figure this out. I hope someone can point me in the right direction.
Thanks!
You might want to get the original Turning Machine V2.1 document. It explains how it works with the only differences being the noise circuit being redone with adjustable gain and the length switch.

Locking is usually an issue with the noise level so that may need to be adjusted again, especially if you changed U1 (noise TL074). That may cause the pattern issues also.

If the Write switch works correctly I'd say the shift register is OK and check the feedback circuit.

Jay S.

User avatar
rduif
Learning to Wiggle
Posts: 22
Joined: Thu Sep 06, 2018 8:45 am
Location: Rotterdam, Netherlands

Re: [Project] Turing Machine Mk2

Post by rduif » Tue Oct 20, 2020 4:53 pm

emmaker wrote:
Tue Oct 20, 2020 2:42 pm
rduif wrote:
Tue Oct 20, 2020 1:59 pm
Ok. After a long while not looking at my TM, I have tried for the last few days to get it to work again. The results are... mixed.

I changed a bunch of the IC's and such to make sure I used as much as possible from reputable sources. Right now only the cd4015 and tl431 are from aliexpress (so I guess the first suspects?). I got the module to lock sometimes, but in general it is behaving very inconsistently, but in a wrong way and not in a 'random sequencer' kind of way.

I noticed particularly that the length setting and the amount of steps in the sequences do not match at all, for CW nor CCW. They are usually uneven and never the expected value. Would this indicate a problem in the shift registers? I have tried looking at the schematics but Im a beginner and an amateur so dont have the expertise required to figure this out. I hope someone can point me in the right direction.
Thanks!
You might want to get the original Turning Machine V2.1 document. It explains how it works with the only differences being the noise circuit being redone with adjustable gain and the length switch.

Locking is usually an issue with the noise level so that may need to be adjusted again, especially if you changed U1 (noise TL074). That may cause the pattern issues also.

If the Write switch works correctly I'd say the shift register is OK and check the feedback circuit.

Jay S.
Thanks Jay!
I found the document you mentioned, so l'll have a look. On first glance it already explains two types of behaviours I didnt understand at first. My write switch seems to work fine too, so I'll run past the noise circuit again.

User avatar
minyu
Learning to Wiggle
Posts: 3
Joined: Thu Oct 22, 2020 3:18 pm
Location: Istanbul
Contact:

Re: [Project] Turing Machine Mk2

Post by minyu » Thu Oct 22, 2020 3:33 pm

Hello all,

I recently finished building TM MII and was super excited to use it, but whatever I do, I can not manage to calibrate it.
When I patch a square wave to clock input, nothing is happening, there is no sound, no light movement on the leds at all.

I checked the resistors (replaced 2 that looks suspicious), capasitors and solderings. everything looks fine. I changed the amp chips (TL074 and TL072) and still, it's behaving the same.

Without calibration, I tried to use the module to see if everything is working or not. Clock input accepts clock, leds are working, all the outputs are working, but of course there is no magic is happening without the locking and randomization. I tried to do calibrate without hearing the voice, checking every time the randomization and locking working or not but it seems an impossible task :)

Is any of you experience this kind of behaviour? And could I ask, how did you solve it?

This is my second build, I am still a novice but I was super careful with everything. It's a bit disheartening to experience when you can not find what is wrong with it, since everything seems to be working, except the strange behaviour on the calibration process.

Should I try to change the other IC's? I was super eager to have this wonderful module, but now I am feeling a bit frustrated.

Any leads would be appreciated a lot.

Thanks in advance, all the best.

User avatar
rduif
Learning to Wiggle
Posts: 22
Joined: Thu Sep 06, 2018 8:45 am
Location: Rotterdam, Netherlands

Re: [Project] Turing Machine Mk2

Post by rduif » Thu Oct 22, 2020 4:55 pm

rduif wrote:
Tue Oct 20, 2020 4:53 pm
emmaker wrote:
Tue Oct 20, 2020 2:42 pm
rduif wrote:
Tue Oct 20, 2020 1:59 pm
Ok. After a long while not looking at my TM, I have tried for the last few days to get it to work again. The results are... mixed.

I changed a bunch of the IC's and such to make sure I used as much as possible from reputable sources. Right now only the cd4015 and tl431 are from aliexpress (so I guess the first suspects?). I got the module to lock sometimes, but in general it is behaving very inconsistently, but in a wrong way and not in a 'random sequencer' kind of way.

I noticed particularly that the length setting and the amount of steps in the sequences do not match at all, for CW nor CCW. They are usually uneven and never the expected value. Would this indicate a problem in the shift registers? I have tried looking at the schematics but Im a beginner and an amateur so dont have the expertise required to figure this out. I hope someone can point me in the right direction.
Thanks!
You might want to get the original Turning Machine V2.1 document. It explains how it works with the only differences being the noise circuit being redone with adjustable gain and the length switch.

Locking is usually an issue with the noise level so that may need to be adjusted again, especially if you changed U1 (noise TL074). That may cause the pattern issues also.

If the Write switch works correctly I'd say the shift register is OK and check the feedback circuit.

Jay S.
Thanks Jay!
I found the document you mentioned, so l'll have a look. On first glance it already explains two types of behaviours I didnt understand at first. My write switch seems to work fine too, so I'll run past the noise circuit again.
I just rebuilt the noise section and replaced the TL074 with one from another batch and Im pretty sure I got it going! Calibration (with 50K trimmer) was suddenly a breeze as well! Step count is correct, locking works both ways and I seem to have a nice range on the slip too :)

I very unscientifically replaced everything at once so dont really know what was actually the culprit, but I suspect it was the TL074 after all because I think it came from one of my first batches which was from a random chinese supplier. Apparently I never checked that properly the dozen times I took it apart... Other than that I also replaced basically everything in the noise circuit, amongst which one of the film capacitors that was reading more than 5% off its design value. But I dont know how significant that was. Perhaps this can help someone with similar problems to mine.

User avatar
rduif
Learning to Wiggle
Posts: 22
Joined: Thu Sep 06, 2018 8:45 am
Location: Rotterdam, Netherlands

Re: [Project] Turing Machine Mk2

Post by rduif » Thu Oct 22, 2020 5:07 pm

minyu wrote:
Thu Oct 22, 2020 3:33 pm
Hello all,

I recently finished building TM MII and was super excited to use it, but whatever I do, I can not manage to calibrate it.
When I patch a square wave to clock input, nothing is happening, there is no sound, no light movement on the leds at all.

I checked the resistors (replaced 2 that looks suspicious), capasitors and solderings. everything looks fine. I changed the amp chips (TL074 and TL072) and still, it's behaving the same.

Without calibration, I tried to use the module to see if everything is working or not. Clock input accepts clock, leds are working, all the outputs are working, but of course there is no magic is happening without the locking and randomization. I tried to do calibrate without hearing the voice, checking every time the randomization and locking working or not but it seems an impossible task :)

Is any of you experience this kind of behaviour? And could I ask, how did you solve it?

This is my second build, I am still a novice but I was super careful with everything. It's a bit disheartening to experience when you can not find what is wrong with it, since everything seems to be working, except the strange behaviour on the calibration process.

Should I try to change the other IC's? I was super eager to have this wonderful module, but now I am feeling a bit frustrated.

Any leads would be appreciated a lot.

Thanks in advance, all the best.
Im not completely sure I understand your problem correctly, but if its not getting clocked by the square wave but it is by a normal clock, check the levels of the square wave youre using. Ive had my TM be unresponsive to a very low amplitude (~1v) squarewave when I was troubleshooting.

Furthermore, earlier in this thread arthurdent recommended the following topic to me when I was having issues:

viewtopic.php?t=207175&highlight=

They suggest running your vco through an lpf to remove some of the very top end which apparently helped.

User avatar
Willzyx
Common Wiggler
Posts: 127
Joined: Mon Jul 18, 2011 12:58 am
Location: Taiwan

Re: [Project] Turing Machine Mk2

Post by Willzyx » Thu Oct 22, 2020 8:23 pm

Noticed that the TU is sensitive to RF, if you put your phone next to it when its locked it will jump.

Not sure if that is what is causing some issues but maybe worth looking into.

EDIT: I should also mention that I am in a high RF environment and unless my phone is right next to the panel its fine.

lolzdoom
Learning to Wiggle
Posts: 3
Joined: Tue Jun 30, 2020 11:26 pm

Re: [Project] Turing Machine Mk2

Post by lolzdoom » Fri Oct 23, 2020 2:27 am

Hi Everyone, I was having a similar issue that most people seem to have. The trimmer was always either to much or to little and only the last 1% seems to make a difference. I tried the following one by one to resolve it. I reflowed everything, I tried 5 different TL074s(tayda), I tried 4 different 3904s(tayda), I tried swapping in a different 4016 (Mouser), I swapped both 4015s(Mouser). At this point I found a bag from mouser labelled 2N3904TFR, so I gave one of those a shot and that seems like it might be working? I got it calibrated so the tone is stable all the way CW, then around 3 oclock it starts to change occasionally shift. The problem I have now, and I am not sure if this is suppose to happen, but a lot of times at 3 oclock I get huge sections of silence very often. Its like the pattern that gets generated is just blank. I would expect this to happen randomly occasionally, but it seems strange that a completely blank pattern is generated every 4-5 times the pattern shifts. Is this okay? It didn't seem like it was doing that in the calibration video. Maybe there is an issue with the some of the logic IC's? It's worth noting that I was noticing blank patterns showing up alot as I was attempting to calibrate between each test, so I don't think this issue is related to the 2n3904 I put in.

PLEASE NOTE: I am not claiming that the 3904 from mouser fixed the calibration issue for me, I need to do more testing. The initial results are promising but I ran out of time. It could be that through all my messing around I got lucky and found a sweet spot on my trimmer. When I have more time next week I'll test the calibration a little more and report back.

User avatar
patched
Common Wiggler
Posts: 79
Joined: Mon Sep 23, 2019 12:42 pm
Location: Sweden

Re: [Project] Turing Machine Mk2

Post by patched » Fri Oct 23, 2020 7:33 am

lolzdoom wrote:
Fri Oct 23, 2020 2:27 am
...The problem I have now, and I am not sure if this is suppose to happen, but a lot of times at 3 oclock I get huge sections of silence very often. Its like the pattern that gets generated is just blank...
I have this issue too. Hope there is an easy solution : )

User avatar
minyu
Learning to Wiggle
Posts: 3
Joined: Thu Oct 22, 2020 3:18 pm
Location: Istanbul
Contact:

Re: [Project] Turing Machine Mk2

Post by minyu » Fri Oct 23, 2020 10:00 am

rduif wrote:
Thu Oct 22, 2020 5:07 pm

Im not completely sure I understand your problem correctly, but if its not getting clocked by the square wave but it is by a normal clock, check the levels of the square wave youre using. Ive had my TM be unresponsive to a very low amplitude (~1v) squarewave when I was troubleshooting.

Furthermore, earlier in this thread arthurdent recommended the following topic to me when I was having issues:

viewtopic.php?t=207175&highlight=

They suggest running your vco through an lpf to remove some of the very top end which apparently helped.
Yes, that's the weird part. It responses to the normal clock, but other than that there is no response from any kind of source.

I tried high, mid, low frequencies. I tried to pass it from an LPF (A-109) also after your comment, but I am afraid no response still. No leds are responding except from a normal clock source - and I can not hear anything from the "Out" output. :despair:

User avatar
rduif
Learning to Wiggle
Posts: 22
Joined: Thu Sep 06, 2018 8:45 am
Location: Rotterdam, Netherlands

Re: [Project] Turing Machine Mk2

Post by rduif » Fri Oct 23, 2020 10:30 am

lolzdoom wrote:
Fri Oct 23, 2020 2:27 am
The problem I have now, and I am not sure if this is suppose to happen, but a lot of times at 3 oclock I get huge sections of silence very often. Its like the pattern that gets generated is just blank. I would expect this to happen randomly occasionally, but it seems strange that a completely blank pattern is generated every 4-5 times the pattern shifts. Is this okay? It didn't seem like it was doing that in the calibration video. Maybe there is an issue with the some of the logic IC's? It's worth noting that I was noticing blank patterns showing up alot as I was attempting to calibrate between each test, so I don't think this issue is related to the 2n3904 I put in.
I had this when clocking at high frequency during calibration a few times. As I was calibrating it seemed to occur less frequently and now I havent had it in a while (though I havent tested too much since I seemingly got it fixed yesterday). As I got closer to calibration, the time it was silent increased, though that seems logical as it locks to a sequence for longer before changing randomly.

I do know that I found the following in the Turing Machine V2.1 document (see this thread: viewtopic.php?t=85924):

"Feeding the module from a multed clock input can sometimes give strange results - clearing a looped sequence. I’ve no idea why this happens. "

So I figured it was due to some aspect present in the square wave I was using. And a quick turn of the knob or push of the write switch would get a new sequence going again.

User avatar
minyu
Learning to Wiggle
Posts: 3
Joined: Thu Oct 22, 2020 3:18 pm
Location: Istanbul
Contact:

Re: [Project] Turing Machine Mk2

Post by minyu » Fri Oct 23, 2020 5:26 pm

With Tom's supervision on github, [https://github.com/TomWhitwell/TuringMachine/issues/142] I proceed a bit on calibration. It's not perfect but at least it's usable now.

This is where I am right now;

5 o'clock locks a sequence,
7 o'clock locks a double sequence.
12 o'clock produces random sequences, everything good so far but;
3 and 9 o'clock positions doesn't generate slightly random sequences. it continues to lock forever. even 1 and 11 is do not make a change, only 12 o'clock seems to differentiate the pattern.

I think I am gonna leave it like this for a while and hope to fix it properly soon. If you have any more thoughts and a suggestion, I would be more than grateful. :hmm:

Post Reply

Return to “Music Tech DIY”