[Project] Turing Machine Mk2

[airfrankenstein]

With regards to the lack of locking, have you performed the calibration procedure? Or if you have and have used your TM in a while, maybe doing it again might help.

[video][/video]

Does the second LED ever light?[/url]

Thank you so much for your post. Turing MKII now working flawlessly - and to think it was so easy!

[Faye]

So I finished the Turing Machine main module and I'm stuck at the calibration process.
I'm trying to feed it a digital square oscillator signal via the Gate1 output of CVPal.
Now even if I turn the trimmer fully CCW, the LEDs are still flickering hard and the output is only heavy noise.
I read that people who experienced similar calibration issues opted for a smaller 50k trimmer.
Would that possibly solve my problem as well in this case, or am I totally on the wrong track by even trying to do the calibration with a signal from CVPal?

[arthurdent]

So I finished the Turing Machine main module and I'm stuck at the calibration process.
I'm trying to feed it a digital square oscillator signal via the Gate1 output of CVPal.
Now even if I turn the trimmer fully CCW, the LEDs are still flickering hard and the output is only heavy noise.
I read that people who experienced similar calibration issues opted for a smaller 50k trimmer.
Would that possibly solve my problem as well in this case, or am I totally on the wrong track by even trying to do the calibration with a signal from CVPal?

Why are you going through the CVPal, instead of just feeding the oscillator directly into the TM as the instructions in the video tell you?

[Faye]

Because I don't have an oscillator yet...
I just started building up my modular system.

EDIT: I just tried connecting it to the CV output of a x0xb0x instead. Unfortunately, the TM doesn't register any signals from that source. And when I connect it to the Gate output of the x0xb0x, the Clk led at least lights up on the TM when a signal is sent from x0xb0x, but it doesn't trigger anything on the TM. It is not even gated to the TM output (The 'Seq' jack).

[Faye]

A quick update:
So it seems that TM actually can receive input from the x0xb0x CV output. It was my mistake yesterday to only play single notes in keyboard mode below F#2. TM only seems to register notes from x0xb0x played from F#2 upwards.
When I play a sequence on x0xb0x:
- I hear a click noise on the TM audio output on every played step while TM clk input is connected to the x0xb0x gate output.
- I hear a click noise for random steps on the TM audio output while TM clk input is connected to the x0xb0x CV output.

I suppose this is the intended behavior, right?

But my main goal is something else atm: I want to do the calibration (preferably with any other oscillator source than the CVPal).
That's why tried to connect the TM clk input to the x0xb0x audio output. Unfortunately nothing happens at all.
Does anyone have an idea how I can a proper square signal for the calibration with the equipment that I have at hand atm? (CVPal, x0xb0x, MFOS Sound Lab Mini Synth => SLMS only has a audio output. Doesn't trigger anything on the TM.)

[rduif]

So I finished the Turing Machine main module and I'm stuck at the calibration process.
I'm trying to feed it a digital square oscillator signal via the Gate1 output of CVPal.
Now even if I turn the trimmer fully CCW, the LEDs are still flickering hard and the output is only heavy noise.
I read that people who experienced similar calibration issues opted for a smaller 50k trimmer.
Would that possibly solve my problem as well in this case, or am I totally on the wrong track by even trying to do the calibration with a signal from CVPal?

Perhaps I'm not the best person to give advise as I haven't gotten my TM working properly yet either ( ), but here's my 2ct.

If you have the trimmer fully CCW and you're still getting heavy noise/randomness, something is wrong. The trimmer should attenuate the noise and at fully CCW should remove the noise completely. This means the noise output should be silent and the scale output has no randomness. A smaller trimmer value isn't going to help you there, that only works to make the trimming process less sensitive and easier to find the sweet spot. Most people who are having calibration issues can't find a position between fully random even when locked and no random output at all.

Regarding the clock input, I initially had the same problem where I tried to clock it from the 1kHz pulse wave output of my pocket oscilloscope, which didnt trigger the TM. I believe amplifying the signal helped there, so if you have a way to do that you might be golden. In any case I think that the audio rate clock input step of the calibration is just meant to find the global position quicker. I'd think you should be able to get it calibrated with a clock rate pulse wave only, BUT I may be wrong.

[EASYBOY]

I fucked up in step 9
I soldered a 3x14 pin IC socket into a 4x16 pin spot (4015 to be specific).

Are there any chances I can desolder that somehow?
All my attempts failed...

[emmaker]

I fucked up in step 9
I soldered a 3x14 pin IC socket into a 4x16 pin spot (4015 to be specific).

Are there any chances I can desolder that somehow?
All my attempts failed...

Don't know what tools you have or your skill level so can't say.

But!

When ever you want to remove something do not try and save it. Clip off the part leaving some of the leads (in this case the plastic case).

Melt the solder and use pliers, tweezers or the iron tip with something holding the lead next to it to pull the metal out (I use a dental pick). Use wick or a sucker to then remove the solder from the hole. Use additional solder and flux as needed.

[mmmcc666]

I didnt build my turing machine mk2 but I thought this would still be the correct place to ask.

It’s calibrated in that it locks very reliably CW and CCW.
However the noise output is very quiet.
Also when it is set at noon, fully random, it seems to eventually tend towards all zeros but never to all ones.
This is annoying because it creates long periods where nothing happens on the cv output or the pulses expander outputs.

My mk1 didn’t behave like this. Has anyone got ideas what’s wrong with it? Thanks!

[arthurdent]

I didnt build my turing machine mk2 but I thought this would still be the correct place to ask.

It’s calibrated in that it locks very reliably CW and CCW.
However the noise output is very quiet.
Also when it is set at noon, fully random, it seems to eventually tend towards all zeros but never to all ones.
This is annoying because it creates long periods where nothing happens on the cv output or the pulses expander outputs.

My mk1 didn’t behave like this. Has anyone got ideas what’s wrong with it? Thanks!

The noise output is controlled by the trimmer that is used for the calibration; as you turn the trimmer down to calibrate, the noise output is reduced. It's possible that the trimmer is turned down too much; that may also explain why the TM seems to "lock in" at the NOON setting. You may want to try a re-calibration, see if that improves the performance.

[mmmcc666]

Ok, thanks, I’ll try it again later tonight

[Quiet Wilds]

Im having the problem that when I power my case on, the Turing Machine lights up, but then the LEDs quickly fade and I get no response. This is my second build ever so Im sure I fucked up soldering somewhere, based on this any ideas where to start troubleshooting? I have all the proper equipment (soldering iron, hot air rework station, multimeter), but I'm very new at this.

EDIT: The 4081 chip wasn't seated properly. Now it is, and all lights go on except one (I assume I fucked up that soldering joint) but the module still doesnt function properly. Now all LEDs are full on all the time even with clock input.

[arthurdent]

Im having the problem that when I power my case on, the Turing Machine lights up, but then the LEDs quickly fade and I get no response. This is my second build ever so Im sure I fucked up soldering somewhere, based on this any ideas where to start troubleshooting? I have all the proper equipment (soldering iron, hot air rework station, multimeter), but I'm very new at this.

EDIT: The 4081 chip wasn't seated properly. Now it is, and all lights go on except one (I assume I fucked up that soldering joint) but the module still doesnt function properly. Now all LEDs are full on all the time even with clock input.

When you say the 4081 "was not seated properly", was it just not inserted fully or was one of the pins bent. And the usual stuff. Do a close inspection of all your solder joints. Are any diodes, capacitors installed in reverse. Are the rest of the chips installed correctly. Any damaged traces on the PCB. Some clear photos would be nice if you'd like input/feedback. Have you tried to do the calibration yet; does anything change if you go through that procedure.

[Zijnzijn Zijnzijn!]

Hey all, I bought a TM v2 a while back on Reverb, and it's never been able to latch onto a pattern (the guy I bought it from built it and didn't know it was even supposed to latch on at all ¯\_(ツ)_/¯). I tried the calibration method, but the only thing that comes out is a white-ish noise signal, at every trimmer position EXCEPT at a complete counter clockwise turn, which gives out a single tone. Any nudge clockwise brings it back to white noise. I have a soldering iron but I don't know much about what to look for. Any thoughts?

[arthurdent]

Hey all, I bought a TM v2 a while back on Reverb, and it's never been able to latch onto a pattern (the guy I bought it from built it and didn't know it was even supposed to latch on at all ¯\_(ツ)_/¯). I tried the calibration method, but the only thing that comes out is a white-ish noise signal, at every trimmer position EXCEPT at a complete counter clockwise turn, which gives out a single tone. Any nudge clockwise brings it back to white noise. I have a soldering iron but I don't know much about what to look for. Any thoughts?

From my previous post:
"...the usual stuff. Do a close inspection of all the solder joints. Are any diodes, capacitors installed in reverse. Are the chips installed correctly. Any damaged traces on the PCB. Some clear photos would be nice if you'd like input/feedback."

Since you didn't build it yourself, You probably don't have a copy of the build manual to use as a guide; if it was a kit from Thonk you can download the manual there or you may find something at GitHub or the MTM website. If the builder just bought the PCB & panel and sourced his own parts, there may be an issue with the specific components that he chose. Unfortunately there are a lot of possibilities...

[blester]

I didnt build my turing machine mk2 but I thought this would still be the correct place to ask.

It’s calibrated in that it locks very reliably CW and CCW.
However the noise output is very quiet.
Also when it is set at noon, fully random, it seems to eventually tend towards all zeros but never to all ones.
This is annoying because it creates long periods where nothing happens on the cv output or the pulses expander outputs.

My mk1 didn’t behave like this. Has anyone got ideas what’s wrong with it? Thanks!

I have this same issue. We’re you able to fix it by recalibrating?

[Zijnzijn Zijnzijn!]

From my previous post:
"...the usual stuff. Do a close inspection of all the solder joints. Are any diodes, capacitors installed in reverse. Are the chips installed correctly. Any damaged traces on the PCB. Some clear photos would be nice if you'd like input/feedback."

Since you didn't build it yourself, You probably don't have a copy of the build manual to use as a guide; if it was a kit from Thonk you can download the manual there or you may find something at GitHub or the MTM website. If the builder just bought the PCB & panel and sourced his own parts, there may be an issue with the specific components that he chose. Unfortunately there are a lot of possibilities...

Sorry about that! Here's pics of the innards:

[arthurdent]

From my previous post:
"...the usual stuff. Do a close inspection of all the solder joints. Are any diodes, capacitors installed in reverse. Are the chips installed correctly. Any damaged traces on the PCB. Some clear photos would be nice if you'd like input/feedback."

Since you didn't build it yourself, You probably don't have a copy of the build manual to use as a guide; if it was a kit from Thonk you can download the manual there or you may find something at GitHub or the MTM website. If the builder just bought the PCB & panel and sourced his own parts, there may be an issue with the specific components that he chose. Unfortunately there are a lot of possibilities...

Sorry about that! Here's pics of the innards:

As a start, I'm not terribly impressed with the soldering. In particular I see several joints that look very thin and could use some touch-up:

In the first picture, there are two 2x8 pin headers on the left hand side of the board. On the upper header the top left pin looks thin. For the lower header, the top left pin, then the two pins in the second row could use touch-up. On the top left of that board is 2x5 pin header between the diodes and the fuses; in the photo it looks like 2 of the pins on the left-hand side are bridged.

In the second photo, along the top edge just right of center there is a row of six pins; the second from the right looks starved. And most of the pins showing here are just poorly done, in MY opinion, they could all use touch-up/clean-up.

On the third photo, there is one pin in the top left-hand corner just above the power header that appears to have little or no solder.


I'd start by pulling all of the IC's, getting some isopropyl alcohol and giving the boards a good cleaning, then going over all of the solder joints one by one, and then a final clean-up and re-assembly, power up and calibration try. SORRY, wish there were something easier but I don't see anything else that's "intuitively obvious to the casual observer".

[J_Behan]

I'm having the same issue with calibration. Can anyone confirm a trimmer value that actually works? Had I known this was such a problematic build, I wouldn't have bothered with it.

[HeWhoWantsJeans]

I just bought a secondhand DIY Turing Machine MK2 and I'm experiencing the same calibration issues (ie: the Trim Pot doesn't seem to do anything until nearly entirely CCW) and when I do get a clear tone from the Output during this point in calibration, the Lock knob doesn't seem to do anything. I just took it to a repair tech who went over it and nothing seemed off about the soldering - though the Trim Pot physically looked like it had been on the receiving end of a chainsaw and we thought maybe the IC's could be suspect, so I have all new IC's and a Trim Pot coming in the mail from DigiKey.

Bummed to hear that the calibration seems to be an issue for so many. I'm a huge fan of the Radio Music (and Chord Organ) and it's sad to see this design appear to be problematic.

[patched]

Same issue here with calibration. The window for the correct spot is so incredibly small that I have no way of hitting it. Either it has some randomness at CW, or no randomness at noon.

[emmaker]

I'm wondering if there might be a problem with socket and chip height between to the back and front boards when using leaf pin sockets verses round pin sockets.

Anyone use leaf sockets and have any issues?

Thanks in advance.
Jay S.

[patched]

I'm wondering if there might be a problem with socket and chip height between to the back and front boards when using leaf pin sockets verses round pin sockets.

Anyone use leaf sockets and have any issues?

Thanks in advance.
Jay S.

I use leaf socket for all my ICs

[emmaker]

I use leaf socket for all my ICs

Thanks
Jay S.

[djthopa]

Hi! So yesterday i made a bad move and plugged the gates expander (volts expander wire from main pcb) into power.
Magic smoke and burned module.
Where shoud i start debugging? Ics related to the gates expander. Guess ill have to fire up eagle and follow traces....at first sigth i cant see a ything burned but its smells like it....shite