[Project] Turing Machine Mk2

[indicator]

Thank you for the guidance and great work. The module was built and calibrated, but may have been jumbled in shipping or something. I was able to make the adjustment per your included video and everything appears to be operating as expected. Getting "enough random" at 12 o'clock and drift at around 2 o'clock took some finnicky back and forth but the directions were clear and the fine tune accessible. Thanks again!

[emmaker]

Thank you for the guidance and great work. The module was built and calibrated, but may have been jumbled in shipping or something. I was able to make the adjustment per your included video and everything appears to be operating as expected. Getting "enough random" at 12 o'clock and drift at around 2 o'clock took some finnicky back and forth but the directions were clear and the fine tune accessible. Thanks again!

You aren't guaranteed that the previous system has the same "+/-12V" voltages as your system.There is no voltage regulation on board to provide a consistent voltage reference so calibration should be redone when moving from system to system or if modules are added that heavily load the power supply.

[indicator]

Interesting, will keep that in mind.

[citruesilver]

I just finished building the module (bought at Thhonk) and I have great difficulties to calibrate it. I rechecked all the components and they are all in place and soldered correctly (as far as I can tell).

When I feed the module a clock from an osc (square wave) and start calibrating I need to turn the potentiometer all the way CCW in order to get a stable tone. When I try to use the module as a random sequencer it generates a random sequence for approx. half the steps and the rest of the steps are a single tone. The clock input works as it should and the LED confirmes the pulses. Also the LEDs on the top seem to work as well. When I start reducing scale the 'randomness' disappears until gone completely.

Also even when turning the potentiometer a little bit more CW the sequence is not reaching random notes for the full length. Does anybody have an idea what could be causing this? It seems that I am not the only one having the calibration issues but unfortunately nobody has so far offered an explanation or solution.

[Thonk Support]

When I try to use the module as a random sequencer it generates a random sequence for approx. half the steps and the rest of the steps are a single tone.

Does it do this for loop lengths shorter than 16 steps?

[citruesilver]

When I try to use the module as a random sequencer it generates a random sequence for approx. half the steps and the rest of the steps are a single tone.

Does it do this for loop lengths shorter than 16 steps?

Yes it actually does. BTW Steve from Thonk has given me the edvice that my solder points may be the culprits here as they solder may not have reached the other side on all solder points. I'll check and come back...

[KSS]

BTW Steve from Thonk has given me the edvice that my solder points may be the culprits here as they solder may not have reached the other side on all solder points.

huh??
PT holes don't 'need' the solder going all the way through. That's why they're plated. It is still generally preferred and good practice to have a fillet on both top and bottom, but is far from necessary to correct operation. Cehck out any of the new Behringer modules which all seem to have this 'fault'.*

It only becomes necessary when a two sided PCB does NOT have PTH. Then you have to solder all the expected connected tops too. For commercially produced bards this should b a non-issue.

* I agree it's an example of poor soldering, and have said so in Behringer threads. But that's not the same as saying it causes faulty operation.

[Thonk Support]

BTW Steve from Thonk has given me the edvice that my solder points may be the culprits here as they solder may not have reached the other side on all solder points.

huh??
PT holes don't 'need' the solder going all the way through.

Sure they don't need it, but it IS indicative that dry joints are very likely to be present, after putting thousands of Turing Machine kits out there this always sets the alarm bells ringing for us that solder joints need reflowing, It's 8 years of cumulated experience.

[KSS]

Thanks for the reasoning behind the statement.

[citruesilver]

Just go my new solder tip today and will resolder tonight. Hopefully I will have random sequences soon :-)
I'll keep you posted. Thanks for the feedback, guys!

[mwmac]

Has anyone got any recommendations for an alternative to TS78L09CT. It seems to be discontinued in most places (https://www.mouser.co.uk/ProductDetail/ ... %2FfMR5f4=). The only place I found it was at RS but in bundles of 50.

Would the following be a suitable replacement: https://www.mouser.co.uk/ProductDetail/ ... -L78L09ABZ

Thanks in advance.

[synthetek]

Would the following be a suitable replacement: https://www.mouser.co.uk/ProductDetail/ ... -L78L09ABZ

yes that will work

https://www.mouser.com/Semiconductors/P ... 09&FS=True

https://www.taydaelectronics.com/l78l09 ... or-ic.html

[mwmac]

Thanks for the speedy response!!

[citruesilver]

No joy,

checked all solder points yesterday and resoldered where to solder had not gone completely through to the other side. Still the same behavior. I have given the module to a friend of mine who is going to double check (he is a soldering pro) but I doubt that the soldering was the real issue here. For arguments sake let's assume it is NOT the soldering, which component could most likely be responsible for the behavior?

I will report back once I get the module back...

[Thonk Support]

Hi, the next thing I would check is the continuity between all the headers between the two boards with a multimeter, with the boards attached together beep test them to make sure the signal can get through ok AND that there isn't a short between any adjacent pins on all the headers.

Also as you have pairs of the 4081, 4050 and 4015 IC's you can swap them round and see if that changes the behaviour at all.

[citruesilver]

Got the module back from my pal. He couldn't find wrong solder points either. Shorts I can also rule out. I will try to test the connections between the PCBs next and swap the ICs of the pairs one by one next.

[citruesilver]

here is a short video showing the module's behaviour...

[bostonguy]

Hi citruesilver, I build a lot of Turings and have done a lot of troubleshooting. I think it’s interesting that the scale is affecting the randomness. If you’re in the continental US and willing to part with the module for a week or two I’d be happy to put it on my bench and see what’s going on. No promises that I can find the issue, though.

[citruesilver]

Hi bostonguy,

thank you for the kind offer. Unfortunately I am located in continental Europe (Germany to be more precise). But maybe you can help me troubleshooting from afar. I will make the prposed checks on the module over the weekend and will then probably post another longer video with the results. If it is still not working after that then I will write this one off as a failure (the first solder project that has failed in ten years time and I have soldered much more complex Synths and Modules than this, well it had to happen someday...) and resell it to someone who has the skills to repair it...

[Thonk Support]

Hey, don't give up so easily relax, have patience and i'm sure we'll get to the bottom of it. You already have a module that is almost working, it's not dead.

[citruesilver]

Hi,

we've tested the connections between PCBs and all the connections of the IC sockets. Then I changed the 3 IC pairs one after another as as suggested. No change in behavior. If set to 8 steps and the sequence locked (full CCW position of knob) the module always plays 5 random notes and then 11 times the same one. If I change pattern length, I get less random notes but the pattern length stays the same (16 notes) If I turn the knob in full random position I get varying notes on the first 5 steps but the rest stays the same. In full CW position it seems that I get the same 16 steps.

The module is calibrated and there is a little bit of travel left on the calibration potentiometer (approx. 10°)

...and it is alwways the same note progression, only varying in pitch in relation to scale setting or tune of the OSC obviously.

Here is the actual video

[Tombola]

Hi, sorry you're having trouble with the module. It's a weird problem that I haven't seen before, which normally suggests either a soldering problem or a short-circuit somewhere giving an intermittent connection. It could always be a faulty chip, but that's relatively unusual.
I wonder if the problem could be in the DAC0800 or the output circuitry. The LEDs seem to be working normally from a brief look at the videos (I don't know if you'd agree) but the CV output is definitely off - like either the bits aren't getting to the DAC, or the output/scaling circuitry is missing an earth connection or something like that.
You could start by re-seating the DAC0800 chip just to make sure it's in the socket correctly + ensuring there aren't any folded over pins. Then check the pins and connections around pins 1,2,3 of the TL074 (both on the front board) + on the output scale pot.
Good luck, Tom

[citruesilver]

Checked the ICs (again), they are all firm in their sockets and no folded over pins. No shorts that I could detect and the TL074 are sitting nicely. I made a ot of close up pictures. If anyone has an idea I am all ears. Like I said I doubt it is the soldering or the connections we've checkes, double-checked and triple-checked...

[citruesilver]

more pics

[citruesilver]

and the last pics