TM3030 makes no sound but LEDS work

[CompletIdjot69]

I recently completed a TM3030 issue 1 board.
As stated I have fully populated it and plugged it in.
The LEDS respond to midi signals and I can even hear the clicking from gate edges but no sound otherwise.
I’ve reflowed the board to no avail and triple checked my substitutions of parts to be accurate.
I’m running it off a 15VAC output brick.
I’m using a 3080 currently and have a new Replacement in the mail a preliminary guess to what could be wrong.
I’ve fiddled with trim pots to no avail as well.
Any help would be appreciated! Thank you

[Rigo]

I moved this topic to the Oakley subforum ...

[Synthbuilder]

The LEDS respond to midi signals and I can even hear the clicking from gate edges but no sound otherwise.

The VCO or the VCF may not be working. Do you have an oscilloscope?

Tony

[CompletIdjot69]

I do! Where should I start looking?

[Synthbuilder]

Where should I start looking?

First off is the waveform switch. Check the pins on both edges of the switch, ie. pins 1 and 3. One should show a sawtooth and the other a square(ish) wave. If not the VCO is not working which would explain the silence.

The VCO is built around the three transistors Q3, Q4 and Q5. You should be using BC182L and BC212L in there (or their Japanese equivalents). If there is no L on the end of the BC part numbers then they are the wrong transistors.

If all is well at the switch, the next bit to check is the signal at the negative pin of C46, a 1uF electrolytic capacitor. This should show an constant oscillation at the frequency of the VCO of around 0.5V peak to peak. This is the output of the filter. The cut-off and resonance settings should affect the shape of the signal here. If not the filter is broken somewhere. Common problems are fake 2SC1583s or SSM2210s - now both quite rare.

[CompletIdjot69]

It would appear there is no signal at the switch.I am using BC182Ls and BC212Ls.
My guess is the pair of ssm2210s I got for cheap are the culprit.
I think I’m going to grab some from synthcube. I’ll report back once I’ve given that a shot!

[CompletIdjot69]

I replaced the ssm2210s with new parts from synthcube.
Nothing changed.
I am currently getting nothing on pins 1 &3 on my switch.

[CompletIdjot69]

Another thing perhaps worth mentioning is that C1 and C3 are both parts rated for 1KV. Obviously overkill. All that mouser had I know in the build guide it had mentioned using lower voltage values for certain caps. Could these be affecting it?

[Synthbuilder]

I am currently getting nothing on pins 1 &3 on my switch.

Check the voltages on pins 2, 7, and 8 on the SSM2210. All should be around 5.3V. If pin 2 is not, then suspect the temp co resistor is either broken or not 1K. If pin 8 is not, then suspect U7 is not working or there is a possible solder short around pins 1, 2 and 3.

What transistor have you used for Q1 or Q2? Remember that only one of those has to be fitted and it needs to be in the correct place.

Another thing perhaps worth mentioning is that C1 and C3 are both parts rated for 1KV.

Do you mean C1 and C2? These are the capacitors that create, along with the crystal X1, the clock for the processor. If the LEDs are working correctly then C1 and C2 are functioning as they should be.

Tony

[CompletIdjot69]

The voltages are all floating around 5.5v

Q1 is empty and q2 is a J201

[Synthbuilder]

The voltages are all floating around 5.5v

This probably means that the exponential convertor and tempco resistor are working fine. Since the VCO is relatively simple the only things left are Q2, 3, 4, 5, the diode D2, the timing capacitor C9, and the resistor R6. Inspect very carefully the top and bottom sides of the boards around these parts for any short circuits. Check again the direction of D2.

What is the voltage between pin 1 of U6 and the middle pin of Q2? Use a decent voltmeter to measure this with the red lead on pin 1 and the black lead on the middle pin of the FET. It should be around -1V. Note the minus sign. If it's positive Q2 is probably dead or a fake.

[CompletIdjot69]

There appear to be no shorts between any of the components you had mentioned.

It looks like U1 on this board is a voltage regulator with three legs? Is there a different component I should be checking?

[Synthbuilder]

It looks like U1 on this board is a voltage regulator with three legs?

Sorry, dyslexia strikes again. I meant pin 1 of U6, not pin 6 of U1.

[CompletIdjot69]

Well.
I got no reading, or a very weak reading on the middle pin of the q2. So I swapped it out for one I had kicking around.
Reading then got a little stronger, but not quite -1v.

HOWEVER .
I popped U10 and got Smokey. I bought a cheap batch of LM723CNs. Sounds like I need to source the real deal.
This crossed my mind earlier, but I have gotten ICs from the distributor before with no issues.

This whole experience has taught me I have to just bite the bullet and find the real thing haha.

[CompletIdjot69]

So the reading from pin 1 of u6 and pin 2 of Q2 is about -0.001.

I replaced U10 and in the process the pad on the bottom of the board for pin 9 was damaged, but the top layer seems to still be attached. I powered it on and it is behaving as it did before. So perhaps this isn’t an issue.

Any suggestions of what to check next?

[CompletIdjot69]

I but the bullet and ordered a 2SK30A-O FET to replace this J201 that has a very weak signal. Hopefully this will work!

[Synthbuilder]

It does sound like you have a bad FET there. Pin 2 of the FET should follow the voltage on pin 1 of U6 but at around 1V lower.

Any suggestions of what to check next?

What's the voltage doing on pin 7 of U7 as you play different notes on your connected keyboard? This should be changing with the usual 1V/octave. That is, play any note an octave below the previous one and the voltage should drop by very close to 1V.

If the voltage on pin 7 U7 is stuck at something like +11V, that may also explain the dead VCO.

[CompletIdjot69]

I removed the j201 from Q2 and replaced it with a 2sk3oa-o on q1.
Measuring the voltage of pin 7 on u7 got anywhere from .447 to .887 volts with my sequencer set to extreme ends of its pots alternating between steps set to a 0-5V output.
Still no sound and no read from the waveform switch referenced to ground.

[Synthbuilder]

Measuring the voltage of pin 7 on u7 got anywhere from .447 to .887 volts with my sequencer set to extreme ends of its pots alternating between steps set to a 0-5V output.

OK, I think this is where the problem is. The VCO may not be being driven correctly from the DAC or slide circuitry.

Try using a standard midi keyboard rather than a sequencer to control the TM3030. You should be getting a voltage on pin 7 that goes from around 0V to around 5V with different notes. For example note A should produce 1.75V, 2.75V, or 3.75V depending on which octave you are pressing.

Common faults around this area of circuitry are using a 74HCT4066 instead of a 4066 for U8. U7 should be a LM358 or AN6562.

One thing that does puzzle me is that you mention your sequencer is outputting 0 to 5V. Does it have both analogue and midi outputs?

[CompletIdjot69]

I misspoke because I was excited.
I’m using the midi out from my sq-1. I unfortunately do not own a midi keyboard.

I was using a 74HCT4066 . I replaced it with a proper 4066.
I am now getting very quiet signal shrouded in noise.
I’m getting response from my pots and I can faintly hear the cutoff changing with knob adjustments as well as the squelchy resonance. Albeit very quietly.

I checked the waveform switch with my oscilloscope again to no avail.

I’m concerned that when I popped my fake 723 that may have damaged other components? I don’t see anything obviously Damaged. Should I check for that?

I feel like this is getting closer and closer, and I’m very excited. Thank you so much for all of your help so far.

[Synthbuilder]

I am now getting very quiet signal shrouded in noise.

When you say signal do you mean you can hear the oscillator changing pitch with different midi notes? If so, check again at the waveform switch, you should be seeing a signal there.

If you really can't see anything at that waveform switch then the VCO is still not working. Check again the voltage on pin 7, U7, which should go from around 0V to around 5V with different notes. For example note A should produce 1.75V, 2.75V, or 3.75V depending on which octave you are pressing. If this is OK and you're still not getting any signal at the waveform switch the VCO must be dead. That means looking again at those BC182L and BC212L transistors, the 2SK30A FET, the timing capacitor C9, and the diode D2.

Using your scope, what's the voltage on pin 1 of the SSM2210 with respect to ground (0V). It should be a sawtooth if the VCO is working, but it might tell us something if there is no oscillation there and it's a steady voltage.

[CompletIdjot69]

So I replaced the two 182L transistors and the 212L transistor with different instances of the appropriate parts.

My oscilloscope showed nothing on pin 1 of U6 in reference to ground. I switched my scope to DC and got a faint but offset sawtooth waveform.

I put my multimeter on pin 1 of u6 in reference to ground (again on dc this time) and got a reading moving from 6.3 V to 9.7 V
On AC I got 1.629V to 1.643V .

I confirmed with my multimeter that D2 is working and that C9 is working. I’m using an axial leaded cap for C9. Could that be a problem?

[Synthbuilder]

My oscilloscope showed nothing on pin 1 of U6 in reference to ground. I switched my scope to DC and got a faint but offset sawtooth waveform.

A working VCO should produce a sawtooth between 12V and 5.5V. With the scope set to DC and 5V/division, you should see a sawtooth on the screen just over one square high in the top half of the screen.

Seeing any sort of sawtooth waveform suggests the VCO is working. What scope and probes do you have?

I put my multimeter on pin 1 of u6 in reference to ground (again on dc this time) and got a reading moving from 6.3 V to 9.7 V
On AC I got 1.629V to 1.643V.

Most multimeters don't do so well at measuring audio signals. However, those are the sorts of figures I would expect to see if the VCO is working.

I’m using an axial leaded cap for C9. Could that be a problem?

No that should work fine. However, axial polystyrene caps are very easy to melt with a soldering iron.

[CompletIdjot69]

I have the very cheap exposed-red-pcb oscilloscope.
The probes are alligator clips

Also: I checked the waveform switch again in reference to 0V.
I found that they were both “stuck” at some kind of positive offset. The lower position (closest to the tune and frequency pot terminals) was at about 7.04V steadily, while the highest position on the waveform switch ( closest to R6, Q3, Q4,Q5) was sitting around 6.84V steadily.

[Synthbuilder]

I'm not sure what to suggest next - it's really difficult to fault find remotely and I could probably find the problem in minutes if it were in front of me. You could try removing Q1 or Q2 completely. With your scope check the voltage on pin 1 of U6. If this is a sawtooth waveform the problem is probably the FET, either Q1 or Q2.

If there is no sawtooth then the VCO is not working at all. But since you did see a sawtooth at some point, you may have an intermittent problem. These are the worst sorts of problems to fault find. But the most common reason for intermittent faults are poor electrical contacts. The problem is locating them.

Do check again pin 7 on U7. Make sure the voltage is changing with midi note number in roughly a 1V/octave fashion.