MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

MFOS Noise Toaster - external gate mod?
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page 1, 2  Next [all]
Author MFOS Noise Toaster - external gate mod?
four_corners
Hey everyone,

I recently built an MFOS Noise Toaster, which has been a blast to mess around with.

I added Ray's CV input mod for the VCO, but I was wondering if anyone could help/point me in the right direction for adding an external Gate control input?

There is already a manual gate built into the Noise Toaster via a momentary switch/button, so there must be a way to add a gate input to control this. It would really take this little guy to the next level to be able to use a keyboard with CV/Gate to play it as a mono synth.

Thanks for the help!
Josh
four_corners
I guess no one has any ideas :(
Permette
Hi

I would try something with a SP3T in place of S8, adding a third input and simply try input resistor (1k ?) before it.

Cheers !
j450nn014n
Ya, I'd be curious about this.
Geo747
Would there be a potential for a problem with that solution though because for a gate input you would want the ground connected to the same ground that is used for the pitch input mod. However, the ARG is referenced to the negative rail, not the ground point (and the trigger button + transistor on the manual input make the input voltage go between the positive and negative rails).

As such, when you apply 0V to the gate input, that would actually be about 4.5V in reference to the negative supply, and therefore the ARG would only go between 4.5V and 9V, thus having shorter attack and release times, and also less depth.

I may have made an incorrect assumption somewhere in this or something im not sure, I dont have a noise toaster to test it on but that was the potential problem I saw
Permette
It sounds you got it further than i, Geo747.
So, i gess, one would need to tweak something in the same way of the manual gate area then ?
Thanks for the knowledge.
Cheers !
four_corners
Geo747 wrote:
Would there be a potential for a problem with that solution though because for a gate input you would want the ground connected to the same ground that is used for the pitch input mod. However, the ARG is referenced to the negative rail, not the ground point (and the trigger button + transistor on the manual input make the input voltage go between the positive and negative rails).

As such, when you apply 0V to the gate input, that would actually be about 4.5V in reference to the negative supply, and therefore the ARG would only go between 4.5V and 9V, thus having shorter attack and release times, and also less depth.

I may have made an incorrect assumption somewhere in this or something im not sure, I dont have a noise toaster to test it on but that was the potential problem I saw


Hmm, yeah I think I follow what you mean. I wish I had a better grasp of this stuff so I could actually come up with a solution myself, but I tend to be more of a 'take a bunch of other peoples suggestions and paste it together' kind of guy.


I wonder if there is any sort of circuit that could be built and patched into the manual gate trigger button itself, so basically you'd set the SPDT to the manual trigger, and then have an external signal that shorted the trigger button? Does that make any sense?

Would this be of any use? This is from the MFOS Weird Sound Generator....



So basically cut out all that stuff from the left of the 1M resistor by the SPDT, and have the Gate signal go into the 1M resistor.

Thoughts?
Geo747
Yeah I was thinking along those lines too at first. My first thought was to simply use a signal diode into X21 to simulate the button, as a signal diode has in effect a very high resistance when in the off state like the button and a low resistance when in the on state like the button, however the diode would be in its on state whenever the voltage drop across it was more than its forward voltage and as it would be going through to the negative supply I think it would always be in its on state (as the 4.5v from ground to the negative supply is greater than its forward voltage).

After thinking about it for a bit the only real solution i could come up with is a little thing with an op amp comparator like this:


I made this with parts I thought it was likely you might have lying around to make it easier to build. The op amp choice doesnt really matter, any common op amp should work as its just acting as a comparator. The 100nF capacitor should be put near the power pins of the op amp as it smooths out any transient current spikes when the comparator changes state. The input voltage is being compared to that formed by the voltage divider made of the two resistors. This voltage is 9*(100000/(100000+22000)=7.37 Volts. When you take into account the fact that the input is referenced to a ground at 4.5V, this means that the gate input needs to be 7.37-4.5=2.87 Volts or higher for the comparator output to go high. If the voltage is below that then the output will instead be at the negative power rail.

The two diodes are there to clamp the voltage and are probably not that neccessary, but if the input voltage goes above the positive supply by more than the forward voltage of the diode, then current will flow through it so that the input to the op amp can never have a higher voltage than the supply + the forward voltage. The same thing happens with the negative to prevent the input voltage going too low. This is really just a little safety thing to protect the op amp.

Finally this design requires S8 to be switched for an SP3T so that external trigger is another input option to the ARG.

This probably isnt the best solution to this problem, but it should work and I welcome any other improved solutions from other people, or any questions if you dont understand whats going on.

Cheers,
George
four_corners
Geo747 wrote:

This probably isnt the best solution to this problem, but it should work and I welcome any other improved solutions from other people, or any questions if you dont understand whats going on.
George


Wow, thanks so much George! This all makes sense, but I have a couple quick questions if you don't mind.

First, I'm guessing something like an NE5532 or TL072 would work fine for the op amp?

You mentioned the 100nF cap being close to the power pins, so I can literally just place it between the op amp ground and V+ like this...



You mention the gate input just need to be above 2.87 volts, do I need to be careful of too high of voltage, or should the diodes take care of that? For example, I think the gate output of my Keystep is 12V, meaning really 16.5V (12V + 4.5V reference to ground). Will that be okay?

Lastly, since I don't have any SP3T on hand, I'm assuming I can test this out for now by daisy chaining another SPDT off of S8 right? Basically, still have S8 with the Repeat on the first lug, D2/D3 on the middle, but then just send the last lug to another SPDT's middle lug, with the the first lug being the manual gate, and the other being the external gate circuit? I know that is a bit cumbersome, but that way I can at least try it out while I wait for an actual SP3T to come in the mail.

Thanks again for all the help!
Josh
Geo747
Yep those op amps should work fine, and you are correct in how the capacitor should be over the power pins like that. The diodes should take care of any voltages higher than the positive rail (I.e. anything higher than 4.5V) so the keystep should work fine with it. The diodes just shunt the excess voltage to the power rails. The same thing will happen if you apply less than -4.5V so there is no problems if for some reason you run it from something that can go negative.

yeah that solution with the SPDT sounds good and should work fine.

Glad I could help
George
four_corners
Geo747 wrote:
Yep those op amps should work fine, and you are correct in how the capacitor should be over the power pins like that. The diodes should take care of any voltages higher than the positive rail (I.e. anything higher than 4.5V) so the keystep should work fine with it. The diodes just shunt the excess voltage to the power rails. The same thing will happen if you apply less than -4.5V so there is no problems if for some reason you run it from something that can go negative.

yeah that solution with the SPDT sounds good and should work fine.

Glad I could help
George


Thanks again George! I just finished it and it (kinda) works great, haha.

So the gate circuit is working brilliantly, but for some reason, I can only have 1 of the 2 CV inputs working at a time.

Lets say I plug in an external Gate signal, it works perfectly fine. Or say I plug in an external Pitch signal into my other CV input (not the new gate one), that works fine as well. But if I have both plugged in, then i just get no sound at all, the second the second plug goes in the sound cuts out...

Any ideas of things to look for?

Thanks again!
four_corners
If it helps, here is an image of the previous CV pitch mod that I had done that isn't playing nice when used at the same time as the gate input.

Geo747
Are you sure that the ring of both input jacks are going to the virtual ground at the junction of R37, R42, and C15? If you have instead wired one (most likely the gate in) to the negative power instead of the virtual ground, then when both inputs are plugged in, the virtual ground will be short circuited to the negative power supply and therefore the synth will stop working right.

Other than that im not so sure what could have gone wrong im afraid but ill think about it
four_corners
Geo747 wrote:
Are you sure that the ring of both input jacks are going to the virtual ground at the junction of R37, R42, and C15? If you have instead wired one (most likely the gate in) to the negative power instead of the virtual ground, then when both inputs are plugged in, the virtual ground will be short circuited to the negative power supply and therefore the synth will stop working right.

Other than that im not so sure what could have gone wrong im afraid but ill think about it


Ah I think that is it, I didn't wire the new input ground to the virtual ground. Thanks for catching that for me!
four_corners
Geo747 wrote:
Are you sure that the ring of both input jacks are going to the virtual ground at the junction of R37, R42, and C15? If you have instead wired one (most likely the gate in) to the negative power instead of the virtual ground, then when both inputs are plugged in, the virtual ground will be short circuited to the negative power supply and therefore the synth will stop working right.

Other than that im not so sure what could have gone wrong im afraid but ill think about it


Well, now I have a very unhappy sounding gate circuit...

I can tell that it is kind of working, but it sounds like a dying bird. I get about a half second of pitched sound, but it sounds like the gate isn't enjoying it.

Even if I have the gate signal turned off, it messes up the internal gate even if on repeat or manual mode if I send a gate signal, so something is going on with the virtual ground I guess.

Frustrating that when soldered to the power ground, the gate worked perfectly, but would short the pitch CV.
Geo747
Huh thats really weird. I cant work out why thats happening, hmmm. Check that you have wired everything right, and that the diodes are the right way round. also make sure the op amp power is coming from the positive and negative supplies properly i guess. Is the noise toaster running on batteries or using a mains adapter? I'm somewhat stuck for ideas, this is acting weird.

Does it do the dying bird thing when using just the gate and not the pitch cv. If it does it still i would suspect its another part of the circuit thats not working right because with the input soldered to ground or to the negative rail the gate input should still work just on its own.
four_corners
Geo747 wrote:
Huh thats really weird. I cant work out why thats happening, hmmm. Check that you have wired everything right, and that the diodes are the right way round. also make sure the op amp power is coming from the positive and negative supplies properly i guess. Is the noise toaster running on batteries or using a mains adapter? I'm somewhat stuck for ideas, this is acting weird.

Does it do the dying bird thing when using just the gate and not the pitch cv. If it does it still i would suspect its another part of the circuit thats not working right because with the input soldered to ground or to the negative rail the gate input should still work just on its own.


Yeah it is strange... I tried on batteries and DC and it still does the same thing.

Even without the pitch cv the Gate doesnt work correctly. It basically sounds like something is draining the voltage away very quickly, so you get a split second of sound, and then it fades away. If I then give it a few seconds to recover, then I can press another key and get that same short bit of sound that fades away.

I'm gonna use my multimeter to see whats going on with the voltage around the circuit I guess, and maybe even wire it back up to the DC jack ground (where the gate works) to compare the voltage out of the gate circuit.

Thanks again for all your help!
Geo747
Hmm this makes it more interesting. I had two main areas I thought something may have gone wrong in. One of them was that when running from mains power there is a chance that if the CV source is running from the same power we could get the problem of the negative rail being connected to the ground again. But there was no problem using the pitch mod before when using the DC power so that cant have been the problem as that used the virtual ground.

The way it stops working even when just the Gate is being used is significantly more perplexing, as it should work exactly the same and just need a 4.5V lower input to trigger the gate. This leads me to believe there is likely a problem further on in my gate input circuit that just wasnt caught when the input ring was soldered to the negative rail.

The voltage draining away phenomenon makes me suspect that there is a capacitor not working as intended which is causing the problem. In the ARG there is only one capacitor C18. The time C18 takes to charge/discharge is dependent on the attack and release knobs so try changing their values and see if it has an effect on the time it takes for the sound to die away. I doubt it will but it is worth checking.

The other capacitor is the one across the power of the op amp in the gate mod schematic I sent. Try fully removing this capacitor and see how that effects your results.

In terms of points to measure the voltage at, I would try measuring the voltage at the two inputs to the op amp (relative to the virtual ground) and also the voltage at the output of the op amp based on different input voltages.

Cheers,
George
four_corners
Geo747 wrote:
Hmm this makes it more interesting. I had two main areas I thought something may have gone wrong in. One of them was that when running from mains power there is a chance that if the CV source is running from the same power we could get the problem of the negative rail being connected to the ground again. But there was no problem using the pitch mod before when using the DC power so that cant have been the problem as that used the virtual ground.

The way it stops working even when just the Gate is being used is significantly more perplexing, as it should work exactly the same and just need a 4.5V lower input to trigger the gate. This leads me to believe there is likely a problem further on in my gate input circuit that just wasnt caught when the input ring was soldered to the negative rail.

The voltage draining away phenomenon makes me suspect that there is a capacitor not working as intended which is causing the problem. In the ARG there is only one capacitor C18. The time C18 takes to charge/discharge is dependent on the attack and release knobs so try changing their values and see if it has an effect on the time it takes for the sound to die away. I doubt it will but it is worth checking.

The other capacitor is the one across the power of the op amp in the gate mod schematic I sent. Try fully removing this capacitor and see how that effects your results.

In terms of points to measure the voltage at, I would try measuring the voltage at the two inputs to the op amp (relative to the virtual ground) and also the voltage at the output of the op amp based on different input voltages.

Cheers,
George


Hi George.

I don't want to take up too much of your time with such a small issue, but again, I really appreciate all your help!

I'll try removing that 100nF cap to see what that does first, and depending on the results, I'll try messing with the C18 cap and/or checking the voltage at the spots you mentioned.
cygmu
What op amp are you using?

The input voltage of a TL07x op amp is meant to be at least 4V above the negative supply (BN here) or bad things happen -- basically the output can swing to the opposite power rail from what you were expecting. So in this setup a TL07x might be a problem. I had a circuit fail badly with TL07x for this reason, but an LM324 worked perfectly in its place.

I can't quite figure out why the change to where you soldered the sleeve of the input jack would expose this problem though.

The op amp is still powered from BP and BN, right?

The other thing about the manual gate circuit is that I don't believe Ray's description -- I can't see any reason why the 2N3906 would ever switch on. If you increase the 200R resistor a bit then I would believe it -- to me it looks like it needs to be enough that the voltage divider 10k vs 200R gives a 0.6V drop to turn the transistor on.

If my suspicion is correct then the op amp gate circuit will work a bit differently from the manual button, because the op amp *will* conduct current away from C18 when its output is low. So if you just "pulse" the op amp, you'll send the gate high, then suddenly low.
four_corners
cygmu wrote:
What op amp are you using?

The input voltage of a TL07x op amp is meant to be at least 4V above the negative supply (BN here) or bad things happen -- basically the output can swing to the opposite power rail from what you were expecting. So in this setup a TL07x might be a problem. I had a circuit fail badly with TL07x for this reason, but an LM324 worked perfectly in its place.

I can't quite figure out why the change to where you soldered the sleeve of the input jack would expose this problem though.

The op amp is still powered from BP and BN, right?

The other thing about the manual gate circuit is that I don't believe Ray's description -- I can't see any reason why the 2N3906 would ever switch on. If you increase the 200R resistor a bit then I would believe it -- to me it looks like it needs to be enough that the voltage divider 10k vs 200R gives a 0.6V drop to turn the transistor on.

If my suspicion is correct then the op amp gate circuit will work a bit differently from the manual button, because the op amp *will* conduct current away from C18 when its output is low. So if you just "pulse" the op amp, you'll send the gate high, then suddenly low.


Thanks for the info Cygmu.

Right now there is a NE5532 in the circuit, but I tried a TL072 with the same results. Of course, I happen to be out of LM324's, but I'll look and see if I have anything close. If not, I'll order a few and try everything out again in a few days.

Yes, the op amp is still powered from BP and BN, and it is indeed very weird how the sleeve grounded to the DC ground works fine, but to the virtual ground it gets all screwed up.

I'm not sure if this is of any help, but when the sleeve is wired to the DC ground (when the gate works), I just noticed there is a very slight pitch ramp up over about a second or two before it stabilizes, versus the manual gate tigger which instantly opens. This is all with the attack set to open instantly.

Your suspicion does seem correct about the opamp. I was just doing some voltage tests, and that exact thing would happen, as I send a gate signal, the voltage would spike and then drop quickly.


PS to George - removing the 100nF cap seemed to not effect anything...
cygmu
If my theory about phase reversal is correct (that's the fault mode when you exceed the common mode input range of the op amp) then perhaps it all makes sense: when you changed the sleeve of the input jack to virtual ground instead of BN, the voltage that the op amp sees on that input goes up 4.5V, so a high gate will take it closer to the positive supply. The threshold is also quite close to the positive supply so there is a chance that you end up with an output that swings negative when it shouldn't.

According to this page
https://www.eetimes.com/document.asp?doc_id=1231129
op amps with BJT input stages are susceptible to phase reversal; the NE5532 is one of those. A TL072 is not, but it definitely exhibits this behaviour.

So that could be it. Annoying!
four_corners
cygmu wrote:
If my theory about phase reversal is correct (that's the fault mode when you exceed the common mode input range of the op amp) then perhaps it all makes sense: when you changed the sleeve of the input jack to virtual ground instead of BN, the voltage that the op amp sees on that input goes up 4.5V, so a high gate will take it closer to the positive supply. The threshold is also quite close to the positive supply so there is a chance that you end up with an output that swings negative when it shouldn't.

According to this page
https://www.eetimes.com/document.asp?doc_id=1231129
op amps with BJT input stages are susceptible to phase reversal; the NE5532 is one of those. A TL072 is not, but it definitely exhibits this behaviour.

So that could be it. Annoying!


What a weird phenomenon, lets hope that is the case!

I just noticed I have some LM358N's, which should work fine right? Just a dual version of the quad LM324?
Dave Kendall
Quote:
I just noticed I have some LM358N's, which should work fine right? Just a dual version of the quad LM324?


Yes they are - go ahead and try them.
four_corners
cygmu wrote:
If my theory about phase reversal is correct (that's the fault mode when you exceed the common mode input range of the op amp) then perhaps it all makes sense: when you changed the sleeve of the input jack to virtual ground instead of BN, the voltage that the op amp sees on that input goes up 4.5V, so a high gate will take it closer to the positive supply. The threshold is also quite close to the positive supply so there is a chance that you end up with an output that swings negative when it shouldn't.

According to this page
https://www.eetimes.com/document.asp?doc_id=1231129
op amps with BJT input stages are susceptible to phase reversal; the NE5532 is one of those. A TL072 is not, but it definitely exhibits this behaviour.

So that could be it. Annoying!


Well... still no luck. I tried rebuilding the gate circuit as well with new components, just in case I was blind to some goof I had made, and still the same thing. Maybe I have just incorrectly translated the schematic?



This has the pinout based on an LM358/TL072/NE5532, all which have the V- on pin 4, and V+ on pin 8.

-------

Just a bit of a recap for anyone joining that doesn't want to read everything prior...

With the Gate sleeve soldered to the virtual ground, all 3 op amps make the dying bird sound, which sounds like a dissipating signal, that recharges a few seconds later and can dissipate again with sending another gate signal.

Without the CV signal plugged in, and the gate sleeve soldered to the DC ground (or even just touching my finger), I pretty much get the same thing as the manual trigger button (besides a slightly lower pitch).

-------

I know George mentioned earlier that it most likely wouldn't work based on what I was suggesting, but can anyone potentially think of any other way to effect the gate directly at the manual gate trigger, versus the switch before it?

I obviously don't know enough about electronics, but basically a transistor based circuit that takes an external voltage to the gate of a JFET, which shorts the 2 lugs of the N.O. pushbutton trigger?

EDIT: I guess I mean more like a relay? Can't a transistor be used as a relay in a way?
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page 1, 2  Next [all]
Page 1 of 2
Powered by phpBB © phpBB Group