External clock on MFOS 16 step sequencer

[dummyplugconspiracy]

Sorry… For clarification… Do you want me to remove the resistor completely or put a jumper in place of the resistor?

[eljay]

For the purpose of this test remove it completely and test again with and without an external clock signal.

[dummyplugconspiracy]

I clipped that R12 resistor and tried this again. The sequencer works normally with no external clock when that resistor is removed. When I plug in external clock, the sequencer stops, and I lose my run/stop capabilities, but it does not respond to the new incoming clock either. The generator I am using for this test is capable of 10V maximum output on amplitude. As a test, I turned down the amplitude to see at what point the sequencer would start responding again. It appears the onboard clock switches over to external clock when it reaches around 4 to 5 V of amplitude. At that point The sequencer knows that external clock has been introduced and overrides onboard clock to begin tracking the square wave I’m putting in.

Just to be thorough… I soldered the R12 resistor back in again. When I do that, I get the same results as before. The sequencer will not start/stop with the button or track tempo with the onboard clock. I also noticed I cannot step forward, but I can step backwards, but it skips a step and only hits the odd numbered steps. That’s sort of interesting. The reset button also works. When I introduce external clock at at least 4 V, it works perfectly with whatever frequency I choose on my generator.

I just can’t seem to get everything to work all at once, even though we’ve proven it is recognizing that incoming voltage threshold that defeats the internal clock and defers to external clock. I feel like we’re close here…
What’s next?
Thanks!

[eljay]

From this last test the external clock does require an input with CMOS characteristics so I have modified the Mod board to include the CD40106 that will provide this, in hindsight I should have gone straight to this in the first place, apologies. So here are the new layouts.

Top

Bottom

The BOM is as follows :

D1 Signal diode, 1N914
D2 Signal diode, 1N914
IC2 Hex Inverter, CD40106
C3 Ceramic, 100nF
C2 Ceramic, 100nF
C1 Ceramic, 100nf
SC1 Socket DIL8
SC2 Socket DIL14
IC1 Dual Op Amp, TL072
R12 0.4w Metal Film 1%, 4K7
R11 0.4w Metal Film 1%, 4K7
R9 0.4w Metal Film 1%, 20K
R8 0.4w Metal Film 1%, 130K
R10 0.4w Metal Film 1%, 1K
R7 0.4w Metal Film 1%, 2M
R6 0.4w Metal Film 1%, 100K
R5 0.4w Metal Film 1%, 1K
R4 0.4w Metal Film 1%, 20K
R3 0.4w Metal Film 1%, 130K
R2 0.4w Metal Film 1%, 2M
R1 0.4w Metal Film 1%, 100K

Please note : assuming you are using a +15V/-15V power supply R3 and R8 have been changed to 130K, this lowers the threshold voltage required to generate a clock or start signal to a little over 2 volts. If you are using a 12V supply R3 and R8 can be left at 100k.

You should be able to leave out R5 and R10 (1K resistors) as we are not taking the output straight from the Op amp, I have left them in in case we need to conduct any further tests.

I have added some bypass capacitors to the power rail supplies for the ICs, C1, C2 and C3, they are all 100nF ceramic capacitors. You can leave then out if you wish, they will not affect the functionality of the board, they are there as a belt and braces to remove any potential noise on the power supply.

When you are testing you should use a a signal similar to that being output by your other modular equipment, I would guess this is around 5 volts although it should work at a lower voltage. A 0V-5V square wave would be great.

With the exception of the reset button I don't think the step buttons (Stop/Start, Forward, Back) will work with an external clock being input. They don't on mine, although I do have a Vari Clock version and it is slightly different. I have also noticed on mine that with an external clock driving the sequencer the STOP action at step does not work. I think these differences are to be expected.

[dummyplugconspiracy]

OK thanks! I will give this a shot. I may have to gather some additional components, but this might be possible using strictly leftovers. I'll let you know how it goes. THANKS

[dummyplugconspiracy]

Hi. I built this board finally last night and installed it today. It still isn't working correctly. Without external clock coming in I cannot get the internal clock and start to function when I have the board connected. I can make the reset button work and I can step backwards skipping a step each time I press the button. When I disconnect the external clock input and external start input from the board I can get step forward and backward and individual steps without skipping any. Reset works also. But stop and run still does not work. Only when I disconnect the boards output also do things return to normal. For the record… I have discovered that I can make everything work without the board if I just turn up my pulse input up to 4.8v. That seems to be the threshold that makes it recognize external clock. That just doesn't seem like very much power. I don't understand why my midi clock converter isn't outputting that already. I think I'm going to bring this back in and test it again. If I understood this correctly… The whole purpose of the board was to increase the voltage at the external clock input, right? Seems so simple ....
Thanks!

[dummyplugconspiracy]

I loaded the sequencer into my cabinet and confirmed that it still wasn't reading clock input from my moon modular midi to clock converter. However, I have a Yusytnth CV standards module that lets you increase the voltage across multiple outputs. So as a test, I increased it one volt at a time until the external clock started to work. Thinking about how I will actually be using my synthesizer, it seems like this might be an acceptable workaround. It's likely I will have to distribute clock a couple places anyway, so this isn't really a waste of a voltage control distribution channel. I would still be interested in doing the Modification if we were certain we had a working design, but for the moment I'm going to proceed like this. Let me know if you want my help troubleshooting the design, but right now the board is disconnected and I probably won't put it back in for a while. Thanks! I truly appreciate the help! I'm certainly not dropping this idea completely.

[eljay]

Please post photos of the new board, top and bottom as before. We seem to be going backwards, the design should work so something else is amiss.

[dummyplugconspiracy]

Happy to! It’s a difficult picture to take, but hopefully you can see enough. Note that I disconnected the power leads for the moment, but I think you can see where they were since a little bit of wire jacket is still visible. Thanks

[eljay]

I can't spot any obvious errors so the next thing is to test the board is functioning as it should. For this, you should apply power to the daughter board and wire up the input sockets only.

I am assuming that when you connect the daughter board to the sequencer that you have removed the connecting wires to the sequencer PCB from the signal pins of the external start and external clock input sockets, the earth connection on these input sockets remain connected to the sequencer ground as before. The signal wires are now connected to the CLK In and Ext Start inputs on the daughter board. At this stage do not connect the daughter board output wires back to the PCB.

You should be able to demonstrate that the internal clock is still working.

Before connecting any inputs to the external start or clock, measure the voltage output on the daughter board at the XCLK out and XSTP out positions, they should both read 0V.

Now connect a 0-5V square wave input to the external start socket, you should be able to see that when the input is high the reading at XSTP is close to the positive voltage rail ~13V, and 0V when the input is low. Repeat this test for the External Clock input, you should observe the same results. You should be able to vary the level of the input signal, it should work from about 2 Volts upwards.

If this all works, connect XSTP back to the sequencer PCB, you should now be able to demonstrate that an external input signal connected to the external start socket (a trigger or gate) will start the sequencer running when it is halted. When there is no input signal there should be no change to the operation of the sequencer. We have not yet tested the external clock connection back to the sequencer.

Let me know how these tests go.

[dummyplugconspiracy]

I didn't have time to actually take the sequencer back out of my cabinet and attach the daughter board, but I'm doing a test that might be useful anyway. I attached the power connector directly to the daughter card and powered it up. I have my signal generator outputting a square wave that my oscilloscope says is about 3V. I connected it to the external clock input of the daughterboard. Below that threshold, the daughterboard outputs 0 V. At or above that threshold, the daughter board shows my square wave at the increased voltage of about 8v. So it is increasing the voltage on output as compared to the voltage on input, and it is respecting a threshold before it makes that change over. I just don't know what implications that part has on the rest of the sequencer signal path when the daughter board is in series. Helpful? I can try some other stuff if you'd like. I just won't be able to involve the sequencer for a few days. Thanks!

[eljay]

The output level isn't what I would expect, it should be a lot closer to the value of your positive power supply rail, if you are using +15V I would expect an output peaking at something around 12 - 13 V. Just confirm when you are measuring that you are using a common ground on the input and output i.e the 0V from your power supply. Also check that Pin 14 of the CD40106 reads +15V and Pin 7 0V.

You should repeat the same test on the External start / XSTP, the circuitry should be identical and therefore the results should be the same.

The input threshold of 3V looks right assuming you are using +/- 15V power supply and have made the suggested resistor changes to 130K. If after checking your measurements again you get the results expected then the board looks to be working as intended.

Assuming success at this stage, you can move on to the tests outlined above connecting the board back into the sequencer. If these pass then the final test is to re-connect the external clock out to the sequencer, for this I'd like you, as a final attempt to get this working, to put a 1K resistor in series with the output signal i.e connect one end of the resistor to XCLK out and then a wire from the other side of the resistor to the sequencer PCB.

[eljay]

I've just spotted that in your implementation C2 isn't positioned correctly, this may be the cause of the problem so please correct this first.

[dummyplugconspiracy]

Nice job noticing that capacitor! I definitely missed that mistake when I was triple checking. I corrected it, but I'm still getting the same results. With external clock plugged in, the sequencer acts as it should, but with external clock removed, the standalone functions aren't working. I cannot get the run stop button to work, and the step forward button doesn't work at all. Step back button works but only hits the odd numbered steps. So it seems like the same problem even though we corrected the capacitor thing. I did confirm that my grounding scheme is correct. I am getting 15 V at 10:14 and 0 V at pin seven.
Still working with the board installed in the sequencer… I confirmed that my Clock input is a square wave at 5 Hz and 3 V. The output from the board is pretty clearly reading 7.5 V. Definitely lower than you were expecting. But the external clock input does seem to be functioning even without low-voltage.
I guess I'm a little unclear how the sequencer is interpreting that threshold input. It seems like even if the board was out putting insufficient voltage, that's not a reason why the sequencer shouldn't function when nothing is connected. When there is no voltage on the external clock input, the rest of the sequence or circuitry should essentially disregard the daughterboard, right? There's some voltage threshold at which it takes over. So it seems like something must be infecting the rest of the circuitry either by drawing voltage away from it or confusing that threshold somehow. This is where my technical skill understanding the fine details falls off and I have to rely on guys like you.
Sort of frustrating how the daughter board is doing exactly what it's supposed to do, but also introducing a new problem in the process. What shall we try next?

[dummyplugconspiracy]

I also ran the test on the external start input and results were the same as you anticipated. Same 3 V in and same 7.5 V out. Lastly… I did go in and put a 1K resistor in line with the output for both external clock and external start. No change to behavior.
I also tested output voltage with the inputs running to the daughterboard, but the outputs disconnected. Same results as before. About 7.5 V output and the sequencer works as normal because obviously it does not know the external clock is there.
As a point of interest, I also tested the leads continuing into the sequencer for Clock and external start. Clock is feeding back a triangle wave of roughly 2 V. External start is not feeding back anything worth mentioning. Barely measurable. But I thought you might want to know that. Is it possible the problem is in my sequencer, not in the daughter board? As I mentioned earlier, the sequencer works perfectly if I just boost that incoming clock using one of my utility modules, so it seems a little unlikely that something further downstream is the problem.
In case it matters… I am using a very basic Chip I bought on Amazon. The Marking is CD40106BE if that means anything to you. In the past I've been told those extra characters usually aren't very important. The other Chip is a TL072 that I have left over from another project. That came from soundtronics so I'm fairly confident in it.

If this turns into a lost cause or it's just taking more time then you have to dedicate to it, please let me know. I have proven that I can get the results I want using that utility module, so it wouldn't be the end of the world if I just continue doing it that way. I would rather have the sequencers responding to the same voltage as everything else in my synthesizer, but I need to be realistic about your time. And mine, I suppose. But I'm still up for solving this if you are.

Thanks!

[dummyplugconspiracy]

I decided to do another test using my actual synthesizer. My midi to clock converter is out putting 2.5 V. I had never measured that before today. It seems to work with all of my other devices that support clock input. The threshold to make the sequencer recognize clock input is 5.6 V. And sure enough… When I looked at the settings on my CV standards module I had it set to add exactly 3.1 V to that input. So the math all makes sense. What I never really noticed before, is that anything less than 5.6 V makes the sequencer behave exactly the same way it does when I attach the daughter board without ext clock connected. Start stop does not work, and I can only step backwards by odd steps. So now I am back at my bench. I have the daughter board powered, but it is accepting no input at the moment. Instead I just wanted to test what it outputs without me connecting the external clock. The outputs are at 0 V, which seems to be ideal. So this does not support my theory that the daughter board is outputting voltage even when no external clock is connected. I really thought I might be on to something here. Oh well!
Then I connected the daughterboard outputs and inputs so it is fully connected and powered like it would be a normal use. No ext clock connected. When I test the output side of the daughterboard then, my oscilloscope shows a flat line at 15v for both the output to external clock on the circuit board and the outlet for external start. Is this normal? Oddly… Right now without external clock input, the run stop still won't work but my steps forward and backward seem to be working normally. That's unusual. Maybe it's just the sequence of how I connected and disconnected the oscilloscope triggered some thing that doesn't normally get triggered, but I thought it was worth mentioning.
If there is some sort of voltage leak, can it be solved by just putting a diode across those outputs or something? I'm sure my experience is showing in that statement… But it seems like diodes can fix things like that, right?

[eljay]

The 15V output from the daughter board when there are no inputs is the problem, it should be 0v. I have no idea why this is happening, particularly when you have confirmed 0V output when testing the daughter board in isolation. I can only assume some sort of error in your off board wiring. I'm not sure how to move forward from here but I will give it some thought.

[eljay]

First test. Disconnect the input wires to the daughter board, apply power, leave the outputs connected to the sequencer and measure the voltage on the output of the daughter board (pins 4 and 8 of the CD40106). It should be 0V for both outputs. If it is 0V the sequencer should operate correctly using the internal clock.

[dummyplugconspiracy]

OK - I can try that when I get back out there. In a previous email I mentioned that with the daughterboard disconnected totally, I was measuring a little voltage on the wire leading to the ext clock input on the sequencer PCB. But there is definitely not 15V coming up those wires from the PCB without the daughterboard connected. I will also test them when connected normally to the ext clock and ext start jacks. This is the second MFOS sequencer I've built and they both appear to work perfectly (and the same) other than this clocking problem that is apparently normal behavior without a mod. And when I boost the external clock voltage externally, even that works perfectly on both sequencers. So it seems unlikely that I wired something that wrong in the sequencer. I'd say it's more likely that I have another hidden problem on the daughterboard. I'll check it again. Maybe I didn't perform the test correctly or just described it poorly. I'll do that again too and see if I get the same results.
THANKS

[dummyplugconspiracy]

I just ran that test. The daughter board is powered up, inputs are disconnected, but outputs are connected into the sequencer. Testing that Chip I am reading 15.1 V on pin four and 15.1 V on pin eight. I took a bunch of other readings today with the board fully installed that I will report separately since it's a lot of stuff and I'd rather type it.
I did notice that if I sequentially disconnect things, only the external clock output from the daughterboard into the sequencer is preventing the sequencer from normal operation. If I had to, I suppose I could just install a switch to break that, right?

[eljay]

A switch is not the solution, the board is not operating correctly, the readings you have taken should be 0V. We need to track the signal through the board to establish where the error is occurring. We will just track the external clock signal. Please measure the following : on the TL072 measure the input voltage at pin 3, it should be 0V, measure the output at pin 1, it should be 0V. If these are correct measure the following on the CD40106, Pin1, should be 0V, Pin 2, should be 15V, pin 3, should be 15V, pin 4, should be 0V. Please also check that pins 11 and 13 on the CD40106 read 0V, from your photo it looks as though the wire grounding pin 11 isn't soldered correctly.

[dummyplugconspiracy]

You want that test done with inputs disconnected but outputs connected as before?
Thanks

[eljay]

Yes please.

[dummyplugconspiracy]

I went over the daughter card in even more detail, cleaned up a couple points that I was worried about, soldered a few things better, and fixed that pin 11 ground wire even though I'm pretty sure it was fine if not a little sloppy. I put the daughter card back in and no it still doesn't work completely, I have noticed that now my step forward and step backward work normally where before I couldn't step forward and stepping backward was only by odd-numbered steps. So that's sort of interesting.
Then I test of the points on the chips you asked about.
TL072
Pin 3 is -11.2v
Pin 1 is -13.6v
Definitely not what you wanted to see there. But I went ahead and tested the other Chip also.
CD40106
Pin 1 is 0v
Pin 2 is 15.2v
Pin 3 is 15.2v
Pin 4 is 0v
Pin 11 is 0v
Pin 13 is 0v
That appears to be exactly what we want. So this is getting interesting…
Just for fun and because I had some laying around, I swapped the TL 072. That did not have any effect.

This seems like constructive information… What's next?
I realized today that my workaround of increasing voltage using that standards module isn't really a great solution because it's not possible to take that amplified clock signal and plug it into modules like my clock divider. It overdrives them and they don't work. So my brilliant idea about splitting that signal up across my three sequencers using the Clock divider is a nonstarter. So I have to make this work. And I feel like we're getting closer Thanks!

[eljay]

I am at a loss. The TL072 readings are correct, my mistake in incorrectly specifying the expected output, the output at pin 1 will either be close to the negative power rail or the positive power rail, in this test it should be close to the negative power rail approx -13.5 volts, which is what you are seeing. This goes through the diode (D1) and becomes a positive only voltage of between 0V and approx 13.5V, in this test it will be 0V which is what you are seeing at the input of the CD40106 pin 1.

This time you are seeing the correct outputs from the CD40106, previously you said you were seeing 15V on the XCLK output line when connected to the sequencer, this time it is 0V (Pin 4), so you have managed to change something when you 'cleaned up' and possibly fixed a poor ground wire to pin 11. With 0V on the XCLK output I am at a loss as to why the sequencer internal clock isn't working, I think it should be but my knowledge is limited. You could try putting a 1N914 diode on the XCLK output before the 1k resistor (XCLK out ----> diode, orientation ' >| ', maybe even with the CD40106 in place the timing capacitor is still discharging via the daughter board, a diode should stop that but it may prevent correct operation of the external clock when it is plugged in.

Some other things to try if that fails, but as I say I am out of ideas, repeat the tests for the external start half of the circuitry by measuring the following, for the TL072 pin 7 should be approx -13.5V, pin 6 approx -11.5V. For the CD40106 the input at Pin 5 should be 0V, Pin 6 15V, Pin 9 15V, Pin 8 (output to XSTP) 0V. If Pin 8 is not 0V, i'e it is being held high at 15V this might be stopping the internal clock from behaving normally, in which case you could remove the XSTP connection to test whether the internal clock then works. If Pin 8 is 0V perhaps the only solution is to put a switch in series with the XCLK out so that the link to the daughter board can be completely severed when the external clock is not in use (I think now this is possibly the best solution), I'm sure someone with better knowledge than me could explain why the original solution hasn't worked. You would only need to do this for the external clock connection, the external start connection doesn't have the same issues.