crazy pcb design! oh.. and please help me out with a fuzzface guys?

[darron]

somebody asked me to make them a fuzzface after hearing one i made for someone else. here's the pcb i designed from the tonepad schematic:



the squigles i thought would look cool and not leave large areas to eat away. mmm.. copper. they are carefully designed not to short anything out.

here's the board once i populated it. one capacitor is in backwards... shhh



they sound awesome for low end stuff! BUT... both of them have a problem. once i turn the gain up on the device they start making a popping noise at one point. i can turn the pot to get enough gain before that happens to not worry about it, but it's crappy to give this to someone.

i made a recording! don't we all love pictures and samples? i do
http://members.optushome.com.au/bluespherecreations/forum/fuzz.wav

i made an mp3, but my webspace wouldn't let me host it. so i'm sorry that you'll have to deal with a large .wav

the first one i made was from tonepad's pcb hand drawn then etched. the 2nd one was my own design from tonepad's schematic as seen above. both of them are negative grounded (so someone may yell at me for that already from what i've read) and i've only run both of them off battery so far. i used the higher hFE tranny as Q2. i could just limit the gain a bit.

here's the link to tonepad's pdf http://www.tonepad.com/getFile.asp?id=82

from what i've read in the forum, i suppose that i need to bias them. i don't REALLY know what that means. i'm hoping someone could tell me how to do it by ear? if i repalce the 100k resistor with a 200k trimpot, what should i be listening for?

thanks heappps guys

[bancika]

you can just rename mp3 to wav and upload it...it should work.
by the way, your pcb looks very cool

[darron]

you can just rename mp3 to wav and upload it...it should work.
by the way, your pcb looks very cool

i took your suggestion and just tried that then but my server still detected it to be an mp3, even though i changed the extension. quite a bit of effort there.

thanks too.. i'm sure someone will tell me what i did with the pcb is horribly wrong and that i'm going to hell... hehe

[petemoore]

from what i've read in the forum, i suppose that i need to bias them. i don't REALLY know what that means. i'm hoping someone could tell me how to do it by ear? if i repalce the 100k resistor with a 200k trimpot, what should i be listening for?
 The 100k is most likely the feedback resistor, good at 100k.
 Measuring the voltages on CBE's of transitors 1 and 2  will tell a story worthy of comment. Since you're getting boosted/fuzzed signal I'm assuming, perhaps a simple bias adjustment of the circuit is all that is necessary.
 If the above is true, Q2 Drain resistor adjustment to get 1/2v on Q2's Collector is all that is needed. For that you could....
 Add a parallel resistor if a lower value is needed there.
 Pull the resistor and put in a trimpot.
 Pull the resistor, put a 20k or so pot in it's place, adjust to 1/2v the collector, remove without adjusting and measure the resistance which biased the circuit, choose a fixed resistance to replace the pot with there.
 Trimpot there is cool except It doesnt look like the board would support it.
 Try this if you have room for another surface pot and want to try this.
 attach a 4k7 resistor to a *10k linear pot wiper, put that in as Q2 collector resistor as another external pot...I like this anyway, then I can listen to and adjust the knob indicator for different bias sounds.
 Then I trimmed the 10k...works good with/without, but I wanted the pot to Not go totally offbias when fully CCW/CW.
  Ie...there are anumber of different ways to diddle with the resistance of Q2C, I like the sound of 6k8 - 8k2, but having adjustment between 4k7 and 14k7 [like the 4k7/10k pot does] allows for bias altering, which is sometimes necessary, depending on what it depends on including which transistors are used.
  Anyway, with that arrangement and sockets, trying out transistors and biasing them is made easier.

[any]

I totally dig the "reinnessance" style PCB!!
I do etch artwork on the pcb myself, but this is pretty extreme. 

[Paul Perry (Frostwave)]

I couldn't hear any 'popping' in that sample, just the usual kind of noise that one gets with high gain & inadequate sheilding. The fact that the sound changes with the pickup selection is a hint that the noise is external to the effect, but you could check by connecting a 47K resistor across the input (instead of a guitar lead) and cranking the unit up & see what you hear!!

[darron]

I couldn't hear any 'popping' in that sample, just the usual kind of noise that one gets with high gain & inadequate sheilding. The fact that the sound changes with the pickup selection is a hint that the noise is external to the effect, but you could check by connecting a 47K resistor across the input (instead of a guitar lead) and cranking the unit up & see what you hear!!

there's no fuzz in the beginning. there is quite a bit of noise from being unshielded. the popping isn't that though. the popping actually gets faster as you turn the gain up and it changes in tone. the popping noise only starts to come in at a specific point on the gain control. if i use the neck pickup and turn the tone knob down at the guitar it gets worse still, so that small connection to ground via the capacitor in my guitar is doing something also.

[darron]

from what i've read in the forum, i suppose that i need to bias them. i don't REALLY know what that means. i'm hoping someone could tell me how to do it by ear? if i repalce the 100k resistor with a 200k trimpot, what should i be listening for?
 The 100k is most likely the feedback resistor, good at 100k.
 Measuring the voltages on CBE's of transitors 1 and 2  will tell a story worthy of comment. Since you're getting boosted/fuzzed signal I'm assuming, perhaps a simple bias adjustment of the circuit is all that is necessary.
 If the above is true, Q2 Drain resistor adjustment to get 1/2v on Q2's Collector is all that is needed. For that you could....
 Add a parallel resistor if a lower value is needed there.
 Pull the resistor and put in a trimpot.
 Pull the resistor, put a 20k or so pot in it's place, adjust to 1/2v the collector, remove without adjusting and measure the resistance which biased the circuit, choose a fixed resistance to replace the pot with there.
 Trimpot there is cool except It doesnt look like the board would support it.
 Try this if you have room for another surface pot and want to try this.
 attach a 4k7 resistor to a *10k linear pot wiper, put that in as Q2 collector resistor as another external pot...I like this anyway, then I can listen to and adjust the knob indicator for different bias sounds.
 Then I trimmed the 10k...works good with/without, but I wanted the pot to Not go totally offbias when fully CCW/CW.
  Ie...there are anumber of different ways to diddle with the resistance of Q2C, I like the sound of 6k8 - 8k2, but having adjustment between 4k7 and 14k7 [like the 4k7/10k pot does] allows for bias altering, which is sometimes necessary, depending on what it depends on including which transistors are used.
  Anyway, with that arrangement and sockets, trying out transistors and biasing them is made easier.

thanks pete. i replaced the 100k with a 200k trimpot for tonight. i'll get stuck into taking measurements with my dmm in the morning. it's getting late. thanks

[ubersam]

you can just rename mp3 to wav and upload it...it should work.
by the way, your pcb looks very cool

i took your suggestion and just tried that then but my server still detected it to be an mp3, even though i changed the extension. quite a bit of effort there.

thanks too.. i'm sure someone will tell me what i did with the pcb is horribly wrong and that i'm going to hell... hehe

You need to do more than changing the extension to .wav, you'll need to decode it from mp3 back to wav. You'll need something like RazorLame to do that.

[darron]

you can just rename mp3 to wav and upload it...it should work.
by the way, your pcb looks very cool

i took your suggestion and just tried that then but my server still detected it to be an mp3, even though i changed the extension. quite a bit of effort there.

thanks too.. i'm sure someone will tell me what i did with the pcb is horribly wrong and that i'm going to hell... hehe

You need to do more than changing the extension to .wav, you'll need to decode it from mp3 back to wav. You'll need something like RazorLame to do that.

i already have a wave file posted! hehe. we were trying to put an mp3 up to save download time but my server wouldn't allow it.

i'm going to take some measurements and try petemore's advice now.

[darron]

i took some readings:

q1
e=8.22
b=8.16
c=7.90

q2
e=7.96
b=7.88
c=2.33

other factors, my brand new roland 9v battery that i got with a tuner pedal only read 8.34v which strikes me a s being a tiny bit cheap. my tranny's are modern ones that are reportedly higher gain and less prone to leakage. i replaced the 100k resistor with a 200k trimpot and put it in the middle for the sake of these readings. last night i was even picking up radio with it when turned all the way down hehe.

what do you mean i should adjust q2's drain resistor to equal half of the collector? i suppose that is R3, (100k). but that connects to the ...
erm..

anyway. thanks for your help petemore. could you please talk me through this one last step?

[vanessa]

I'm no rocket scientist but I noticed that you have left all those little copper designs in the PCB as artwork. Even though they may not touch one another in your drawing sometimes these things can be accidentally made to touch in the Press-N-Peel process. You may have a short as a result and it would be my guess that it might be hard to track down. If all else fails I would go around the main traces with a small Dremel engraver tip to make sure you do not have a short.

[darron]

I'm no rocket scientist but I noticed that you have left all those little copper designs in the PCB as artwork. Even though they may not touch one another in your drawing sometimes these things can be accidentally made to touch in the Press-N-Peel process. You may have a short as a result and it would be my guess that it might be hard to track down. If all else fails I would go around the main traces with a small Dremel engraver tip to make sure you do not have a short.

i went over it with my dmm before i populated the board. i did have one part short that i cleaned up. there were are few other areas that didn't have continuity but i scraped a bigger gap in there just in case for the future. maybe next time i won't push it quite so far or make sure that there's a border between the squiglies and the lines. the problem that i'm having though is the exact same problem that i have with a different fuzz face that i made from tonepad's design also. i'm pretty sure it's a common biasing thing that people can help me clean up...

[darron]

i've read up on it a bit more. im going to put the 100k resistor back in and tell the guy not to turn the gain up too much! lol. the problem that i have is an oscillation called motorboating. will correct/better biasing fix this guys? geofex suggest chucking a massive capacitor across the power. any suggestions as to what i should use?

[vanessa]

I do like the way the "squiggles" look but seeing that most people will never see them I'm not sure if I would add them to a design.

Another thing that's burnt up a lot of time for me troubleshooting is putting the wrong resistor values on the board or the right ones in the wrong places. Here's what happened to me once. I put a 470k resistor where a 470 ohm was to go. I spent a whole day trying to get that circuit up and running and it drove me nuts. When I figured out I had placed the wrong value in there I made a mental note to always double check my resistor values. It can happen with other components as well.

[petemoore]

what do you mean i should adjust q2's drain resistor to equal half of the collector? i suppose that is R3, (100k). but that connects to the ...
  Basically Q2's *COllector reisistor does well to be a trimpot [~20k], I prefer the sound of a FF with a drain resistor biasing it to 1/2v at a value between 4k7 and 15k...the transistors chosen let you determine what value Q2 will bias them.
  I believe we're talking bipolar transistors for the Rocket, the pins are emitter/base/collector...somehow 'drain' entered in here, that is the output pin of a field effect transistor.
  will correct/better biasing fix this guys?
  Correctly biased circuits always sound different.
  correcting the biasing will improve it's performance markedly.
  See: 'Mosfet Boost Building Problem' thread for the dogmatic approach, measureing and testing every component and node connections on the board.

[darron]

I do like the way the "squiggles" look but seeing that most people will never see them I'm not sure if I would add them to a design.

Another thing that's burnt up a lot of time for me troubleshooting is putting the wrong resistor values on the board or the right ones in the wrong places. Here's what happened to me once. I put a 470k resistor where a 470 ohm was to go. I spent a whole day trying to get that circuit up and running and it drove me nuts. When I figured out I had placed the wrong value in there I made a mental note to always double check my resistor values. It can happen with other components as well.

the copper side will be side side facing the user if they ever open the box to replace a battery. but you are still very correct that it won't be seen. anyone know of a place to buy blue pcb in australia? that looks cool for something that's not seen also!

these resistors came straight out of the packets so i was safe in this case

[darron]

what do you mean i should adjust q2's drain resistor to equal half of the collector? i suppose that is R3, (100k). but that connects to the ...
  Basically Q2's *COllector reisistor does well to be a trimpot [~20k], I prefer the sound of a FF with a drain resistor biasing it to 1/2v at a value between 4k7 and 15k...the transistors chosen let you determine what value Q2 will bias them.
  I believe we're talking bipolar transistors for the Rocket, the pins are emitter/base/collector...somehow 'drain' entered in here, that is the output pin of a field effect transistor.
  will correct/better biasing fix this guys?
  Correctly biased circuits always sound different.
  correcting the biasing will improve it's performance markedly.
  See: 'Mosfet Boost Building Problem' thread for the dogmatic approach, measureing and testing every component and node connections on the board.

thanks pete. that's actually pretty helpful. the unit sounds good, but if biasing it will make it sound better ill do that to this one before i let it go and to my own one which used the nos ac128s compared to these new higher quality ones. i think the older one sounds better, but that may just be the biasing making the difference.

as for this motorboting, according to this http://www.muzique.com/lab/fuzzface.htm a 1000uf (i don't think it specified a value actually) capacitor across the power terminals is certain to fix my problem completely. i doubt this, but i'm going to give it a shot because i'm hoping it actually will at least help a lot.

[Dirk_Hendrik]

Am I the only one who is amazed that PCB works at all??? Looks like one giant shortcut to me

[Chuck]

Couldn't the board actually have strange capacitance effects?