Wonderful oscillation issue - how to keep it?

[moid]

Hi everyone

I finished the below pedal today



which made me happy (I had this old empty tin of hair gel lying around and wondered if I could squeeze a fuzz into it - yes, just!). I made one modification to the original pedal design (it's a Devi Ever OK) which was to place a 22k trimmer pot inside the enclosure so that the board ground and the ground from the output jack both went to lug 2 of the trimmer, and then connected lug 3 of the trimmer with the DC power jack ground to complete the circuit. When the trimmer is set on to full I get a really lovely tremelo/ocillation effect stuttering through the fuzz which is way more impressive than I'd hoped, so much so that I think I'll rebuild this circuit again in a normal box so I can get three pots on the outside. If I turn the trimer off, the pedal makes the expected gated fuzz sound (also good). This was really great until I plugged this pedal into any other pedal that was sharing the same power supply (I've only got those cheap multi power socket connecters that all go to one plug - five plugs on a cable as it were) and at that point I lost the tremelo effect and the fuzz just behaved in its normal original design. If I set both pedals up so that my fuzz has a completely different power supply (different cables and plug) then I get the tremelo again. Whatever I've done obviously has something to do with grounding and I'm pretty weak in understanding this stuff and I'd be really greatful if anyone could offer any thoughts about what I did and is their something I could do to make the pedal share the same power supply as another pedal and keep the tremelo sound?

I have checked applying the trimmer independently to the board ground and the then the output jack ground and I get odd noises but I only get tremelo when both grounds connect to the trimpot. I'd also like to put this pedal inside a large enclosure with a couple of other fuzzes and connect them together but share the same power supply but that won't be nearly as cool if I lose the tremelo effect. Any thoughts wuld be great to hear, happy 2016 by the way

[crochambeau]

Ground is ground.

So... when you connect the signal carrying cable from one circuit to the next that ties grounds electrically to the same plane. (shunting your useful variable resistance into a non-op state)

So.. if you have a circuit that is benefiting from a lifted ground, that will go away if the connected circuit has a true ground reference.

So, in order to attain the lift you'll either need to lift the ground locally (as in, at the circuit (within the larger circuit) that gives you love) or fully isolate the signal path.

Honestly, I would recommend the first option if available, as it will ultimately be cheaper to implement and allow you control, but slapping a transformer to decouple grounds will also probably solve your problem.

Look into analog ground versus digital ground plane to cull some ideas for implementation.

[moid]

Thanks very much, I will look into this as soon as I've had some sleep I'm just glad to hear there is a way of fixing this!

[moid]

OK so I've tried to read a lot of rather complex pages around this area and haven't found anything that seesm to deal with exactly what I want, so my best guess from inferring from the ideas of others is that I should create something called a ground plane which I think I can make from either a strip of copper sheet or possibly a piece of vero board with copper foil on it and no holes (I have an odd piece of this that came as part of a large sheet so maybe that's what it's for; if so I'm glad I kept it!) and then attach all the grounds to this ground plane instead of making them form a star ground (as I would usually do) that goes to the dc jack ground connection. If this is true would you have any idea how large this ground plane neeeds to be?

This seems remarkably simple and easy to do which makes me really suspicious that I've not understood what I've read... especially because there won't be any cable connecting the dc jack ground to anything - and I thought to make a circuit work I had to have power coming in through the positive terminal of the dc jack and the ground going out via the negative terminal of the jack... feel free to laugh at my lack of electrical knowledge here

[crochambeau]

Physical size of the ground plane hinges on current, it's not going to be a concern here. I think you might be overthinking it a bit, though I might be wrong.



I swiped this from viewtopic.php?f=257&t=38374&start=69 hopefully it's the same.

Anyway, can you get the oscillation if you plant your variable resistor between ground and the emitters of the conjoined NPN/PNP pair? While leaving your in/out & DC supply ground tied together?

[moid]

Thanks for your replies, I really apprecitate them I don't think I can do what you suggested with the current build due to the way the vero design that I worked from is laid out, but I think I could try to redraw the vero so that there was space to stick a pot in between the two emitters and then connect the scond lug of the pot to the ground on the... dc jack? or should it go to the ground plane? I will try to draw something tomorrow, work started up again today so things are hectic again until the weekend I think I will try to make another version fo this circuit that I can put in a proper enclosure and that will be easier to make adjustments to - that tin pot I put it in is tiny and rather fragile and I don't want to make any major changes to it in case I break it. Thanks.

Edit: Hmmm I think I can squeeze the pot in, if I make the connection the emitters have to the ground of the board by placing a cut to the left of Q2 and then a wire to connect both emitters to a 22K pot (lug 3) and then send lug 2 as a wire to the dc jack ground that looks possible... I might try it on a new version of the vero though... or maybe get brave and try it on breadboard (I've just bought one and never used it before but this might be a good circuit to try because it has so few components and I really need to force my brain to learn new things - that and read schematics - I think I can almost understand the one you posted)

[crochambeau]

I could try to redraw the vero so that there was space to stick a pot in between the two emitters and then connect the scond lug of the pot to the ground on the... dc jack? or should it go to the ground plane?

I'm trying to offer ideas that maintain DC jack (negative) and ground plane as the same thing, but please understand this is SHEER SPECULATION on my part, and it very well may not work.

There's very little going on in that circuit, so you haven't many pivot points to apply electrical leverage. The point I drove at regarding ground is that it's more or less all shared, and it's an electrical plane that spans most anything you have plugged into one another (and it will find and stay at lowest common point, electrically speaking). So my advice is to stop thinking of ground plane, as it'll be a huge can of worms - and start thinking about *virtual* grounds, points that are isolated from the over all system/ground plane except at one point - from which you can introduce a variable that may achieve your aims without tanking everything else.

So, slapping a resistor at the common emitters of that stage was as close to that as the schematic offered at first glance (to me), but that is no where close to my saying it will work as your existing conditions, and it's certainly not a statement that there are not better or more elegant ways to go about it.

[moid]

OK thanks I'll bear that in mind, hopefully I can try this at the weekend.

[moid]

Hi Crochambeau

Ok some successes and some failures! I built the circuit on breadboard (first time I've done this) and I amazed myself by doing this with the schematic you found and it actually worked as a 'normal' OK Fuzz! Then I added in the potentiometer where you suggested and connected its lug 2 back to the ground on the breadboard... I lost the tremolo sound alas... what happened instead is that the fuzz became extrmely soft, muffled and wooly sounding - like listening to music being played through a blanket. Which was not an awful thing in itself and actually sounds quite interesting, and I found if I changed the pot from the 25K lin I was using to a 1K lin I was able to get a good range from original gated OK Fuzz to muffled gated fuzz to very muffled... but as the sound gets more muffled I lost the higher pitches to the point where the high E string becomes silent and the next highest is pretty quiet - is there any way to still allow some sound through for these strings? I tried a range of pot types from 1K to 100K but the 1K was the best. Maybe the ground potentiometer is acting like a kind of low pass filter - only passing the bass notes?

Any ideas about somewhere else to try the ground potentiometer for the tremolo effect? I've taken a photo of the spaghetti clamp disaster that is the breadboard in case it helps (or just provides you with a laugh). It did occur to me that I have no idea (what I'm doing) whether everything is grounding correctly anyway as I am frequently getting a very loud buzzing hum sound which sometimes goes away if I touch various components... and then fades back in later. Maybe I should build another vero board that I could put into a proper enclosure (it did occur to me that maybe the paper thin aluminium of the tin I used for the pedal might cause ground problems) and then try the same suggestion you made on that in case it works better in a proper case?

I've tried this on a daisy chain power supply with it attached to another pedal turned off in true bypass to test for grounding on the same daisy chain but it didn't make any major changes to the sound if I went direct to the amp with no other pedal in the chain (completely removed the pedal from the chain and power) or whether I went through another pedal plugged into power but turned off.

Even if this can't be fixed I want to say thanks anyway because the experience has forced me to read a schematic for the first time and actually breadboard it successfully and that's no mean feat for my non engineering brain



the bits of masking tape have the numbers of the resistors / capacitors on them... I'm colour blind which makes electrical components a bit more tricky than it should be... and the crocodile clips are holding jumper wires that fits into the breadboard, they aren't just scattered randomly across the table top! And some of the crocodile clips have switches because they were left over from an earlier failed project where I was trying to make a fuzz with various feedback locations controlled by switches... they are turned on (I checked!).

LInk to hi res version of the above image
http://i.imgur.com/Xd4Ca94.jpg

[crochambeau]

Sticking a bypass capacitor across the emitter resistor will return your highs.

Have you built the breadboard to the exact configuration of your other build yet? Can you recreate the tremolo oscillation type effect?

[moid]

Thanks! Just to check - the emitter resistor is the potentiometer that I connected the emitters of both the MPSA18 and 2N2907A to isn't it? And by across it I assume that means have a capacitor span lugs 3 and 2 (the ones I'm using to connect to the board and to ground). I guess the capacitor value is something I should try a bunch and see what happens? I think I'll solder a socket across the lugs and that way I can try different capacitors but if you have any guesses as to whether I should start at pf, uf or nf please say and I'll audition from my selection until something sounds good. If htis works that'll be great because it might fix a problem I have in another fuzz pedal where I tried to get creative and killed the high pitches... I 'restored' them by adding an EHX Screaming Bird circuit, but that is really hard to control without causing huge amounts of piercing feedback (I've nothing against a good burst of feedback, but I prefer to be able to hear some guitar as well )

I've built the breadboard to the circuit you posted which looks (as far as I can tell exactly the same as the vero board layout I used:



Sadly I haven't been able to recreate the tremelo effect on the breadboard. The breadboard pedal sounds identical to the boxed pedal I made, just minus the tremelo so I'm pretty sure I built it correctly.

I might just build another veroboard of this because that would be easier than the breadboard version to play with - I keep knocking bits of that apart and so far it seesm that any fix is likely to require some off board work? It'll have to wait until next weekend or so though, my workload for the rest of this month is horrible . Thanks for the advice.

[crochambeau]

Sadly I haven't been able to recreate the tremelo effect on the breadboard. The breadboard pedal sounds identical to the boxed pedal I made, just minus the tremelo so I'm pretty sure I built it correctly.

To answer the cap question, yeah, that's what I meant. But I fear I might be distracting from the goal by throwing out these suggestions.

Honestly, at this point I would consider it more important to recreate the effect you like, than once that is working you can explore what changes you can do/make to get it to play nice with others.

[moid]

Hey distractions are great ways to learn new things I will try this out at the weekend (hopefully). Thanks again.