MK11 Tonebender (fist PNP build !!)

[analog kid]

I ended up settling on trying out some AC188's in my MK tonebender circuit. I do hate to bother you with it but as this is the first time I've tried a PNP pos.grd project There may be a simple answer. BUT everything looks to be correct on the layout and connections. I used the 3P switch diagram on GGG for PNP grounded /True bypass, w/ LED and I've triple checked that.
Problem is, I'm getting bypassed signal and Absolutely NOTHING with effect ON! When I am reading Q1-Q3 voltages I very strangely get almost exactly battery voltage(apprx. 8.22v) on EVERY leg of ALL transistors!  This makes me think I have something wrong with the power -/+     But I do have the battery Black going to the 2Pad on your layout and the batt.Pos Red going to Input Jack RING with the shield going to "ground"  
That's all correct???
   thanks for your time
additional notes :  build is on perfboard, substituted a Xicon Polystyrene .082 for the .01 cap connected to INPUT CAP and in parallel with 1meg resistor . 5k POT for Fuzz control and here is the link to project:
http://fuzzcentral.tripod.com/benderII.html

[petemoore]

Just last night I built a PNP TB from an old NPN TB board, and found a bad transistor.
 Recently I found a bad cap...I think Murphy's laid the law down for a while...lol.
 Sounds like it doesn't work, check out that sticky thread at the top.
 I spent hours OKing stuff on that TB board, then wired a FF in about 40 min.
 Had a 3v 3 Jfet thing going...if it was jeest setting 'so' sounded good too...couldn't find the intermittent failure point in that thing for anything...it's now a 2 stage Jfet thing...Nice too...has a source>ground/cap/boost/switch.
 Check for anticontinuity between the transistor legs.

[analog kid]

So what would you say would be the first to look at if the problem is a Bad component(elec.caps, Trans.,switch?)?? And Does it mean that a transistor ISN"T bad just if you've tested it and got good leakage and gain measurements? if so then I could rule that out.
(there is anti'cont. between the legs of the trans. Even though they ALL hav the same battery vltg.
I also do notice that I can't seem to get an led to light up for nothing in this thing!! Even though the power SEEMS~! to be getting to the Board :?
I am running it Pos from a switch lug (3pdt) middle row , far left lug. And Neg to the -9vlt on the board.   IS It  Possible I've got a bad switch? It's one of the blue ones.  
(I didn't get rather you said you were just going on to building another fuzz with those FETs or if you actually were using them in your tonebender crkt>)

[petemoore]

You' can figure out switching, jacking and LEDing and circuitry all at the same time...
 I don't mess around with all that anymore though, I lose everything except the board, necessary pots and battery clip, and test that [the 'bare' circuit]...THEN, after I get a working circuit I mess with the 'peripheral' stuff...it's very hard for me to tell if one thing's working when the 'next' thing prevents its function from being exhibited.
 Divide and conquer is the theory I tested. Otherwise you'll probably figure something out.
 I caused less wear and tear, fewer shorts and broken wires, and quicker debugging when testing circuits to 'good' status before adding 'wrapper peripherals'.

[petemoore]

You' can figure out switching, jacking and LEDing and circuitry all at the same time...
 I don't mess around with all that anymore though, I lose everything except the board, necessary pots and battery clip, and test that [the 'bare' circuit]...THEN, after I get a working circuit I mess with the 'peripheral' stuff...it's very hard for me to tell if one thing's working when the 'next' thing may prevent its function from being exhibited...thinking of all the things it could be...
 Divide and conquer is the theory I tested. Otherwise you'll probably figure something out.
 I caused less wear and tear, fewer shorts and broken wires, and quicker debugging when testing circuits to 'good' status before adding 'wrapper peripherals'.

[analog kid]

What are the "wrapper peripherals"??
I wondered if I could  play through an effect just leaving the bypass switch OUT and wasn't really sure if it would be ON But I guess it's obvious it'd always be on with out a switch wired in.. since it's only there for Bypassing!!!
But you didn't mention the jacks? (circuit , battery, needed pots) how could you test it out(plaY THROUGH it) w/o the jacks to plug into the inst. and amp to actually hear the crkt working?
So I should just disconnect the wires at the 3P switch end and leave them hanging,,, Trying it that way?

[analog kid]

I came to the determination that even though I PHYSICALLY had the Input jack wired correct(Effect IN to TIP.. -9v to Ring, and -9vlt to board Ground at SLEEVE) I strangely was getting voltage at the Ring where the Red Batt is AND ALSO 9vlt at the TIP instead of SLEEVE.  SO I tried reversing the Tip and Sleeve and the LED immediately Got Power !!!
The problem now is that the power is Cut off when a cable is inserted in the jack. SO does that happen when the Tip and Sleeve ,Ground and Input connection are REVERSED if so then why does the power not get there when the -9v is sent to Ground through the SLEEVE?
Of note: I also used the audio probe before and found that signal was only getting through to the Input cap and only the top of the input resistors and cap , and the collector / emittter of the 1st transistor.

[analog kid]

sorry three posts in a row>>
Well it's true .. ALWAYS trust your first instict!!
I thought something wasn't right the way the -9vlt was grounding to the board through the input jack sleeve.  EVEN THOUGH it "was" wired -9 RING / -9vlt to Ground thru SLEEVE the readings just weren't saying it was there!  So after wasting time with the probe wondering if something was dragging down the vltg or a bad cap... I decided to just yank the jack and use an old one that I'm used to AND BAM , power AND FX Output!!!
I used a brand new TRS input jack  but I guess I got a Bad one from steve @ SmallBear. Unless, since I'm not used to this type Jack(the Tip and Ring contacts are right on top of one another instead of on opposite sides . there is barely daylight between them but they are not touching. there is a small "bump" on the ring contact that looks like should be touching the tip. is there supposed to be a JOINT there?? maybe it just came loose cause you can squeeze them together w/ a jack plugged in and they'll make continuit. between T/R.
Anyway... I 'll set the bias and try out tomorrow. BUT it looks like I had everthing correct right from the start!!! at least I hope

[analog kid]

project: MK11 Tonebender
source for layout /schematic: Fuzz Central

http://fuzzcentral.tripod.com/benderII.html

Layout: http://fuzzcentral.tripod.com/benderII/mkIIlayout.gif

type of build: Perfboard design

Substitutions/Deviations from layout: .082uf polystyrene cap in place of .01 Capacitor off of + of .47 input elec.cap.
and 5k fuzz pot

Symptom: Effect works both in FX on AND Bypass position. But the effected signal is extremely "gated". Only passing signal (Fuzzing) when the strings are attacked a bit harder. If they are just lightly plucked No Sound will come out!! this is Obviously a bias problem but I can't find the cause

voltages:
Battery 7.89 which I know is a bit low but trust me I've tried with the same results with fresher batteries. this just happento be in therenow
Q1:E 0 B 60mV C 6.08v
Q2:E 0 B 48.5mV C 64.2mV
Q3:E 7.29v B 71.4mV C 76mV
these are all showing up positive voltage with the Red meter lead to Ground and Black to Transistor leg, I don't know if they should be Neg in these cases?
and this all shows there to be Obvious trouble with enough Bias getting to BASE but I don't know what is a possible cause. and if there is more to it than that with all the ultra low voltages showing up. (I would rather see NONE Ithink) So I hope maybe one of you will know exactly where to look! Ihave already tried reversing the protection diode ,it's as it is on the layout. and all caps look to be properly oriented

Last edited by analog kid

[analog kid]

OK Although I have a Fuzz going with either all 3 Germaniums in or subst. 3906's in eiter Q2 or Q3 , the voltages are way screwed up and nothing's biasing right.. SO.. Here I am posting the voltages I get from the different Trans. permutations , as Pete asked me to, ALONG WITH the examples given on the Homewrecker/ROG site.  
I hope this helps you Help Me with the T.S. Process! My battery voltage for these tests was 8.5volts and I apologize for being hard to read.

  ROG                                                              
 All 3 Germanium (not matched set)            
   Q1:   Q2:  Q3:                                
C 8.84   .414 4.67                                      
B 70 mV.111 .414                                      
E 0         0    .274                                      
 
MINE
C 6.7v   73mv  8.29v
B 80mv  59mv 72mv
E  0       0      0


NOTE : On the all Ge. setup my Trimpot is responsive BUT will only adjust from between  8.29 to 8.9! seems as long as a 3906 is in Q2 it will bias a  wider range but never below 5.78 or so
 
Here are the comparisons for Q1 *Ge  Q2 and Q3 3906's
                                                                                                   
Q1(ge78)Q2(3906..147)Q3(3906..152)
C 8.67      1.39            4.8                  
B 69 mV    .633         1.39                  
E 0            0            .747                    
           
MINE , the hfe's are very similar
C  6.7v      1.3v       5.8v with trimpot at lowest*highest is 8.8v
B  80mv   .605v      1.3v
E  0          0            .67v




Here's the closest of the examples I could find to my Q2 3906 and Q1/Q3 Germaniums.


Q1 'comensated'
3906 Q2(3906) Q3 Ge  
C 7.03  .902     4.6v      
B 1.11  .614    .902v    
E .511      0      .762v    

MINE
C 6.8v   .742     5.7v*(to 8.21)                                          
B 82mV  .608v   .742
E  0          0       .685

[R.G.]

Well, first let's think a bit.

Looking at the schemo and using our fanstastic recall for schematics of circuits, we recognize instantly that the TBPro MK2 is a fuzz face (Q2, 3) preceeded by a common emitter gain stage.

Everywhere there is a series cap in the signal string (i.e. 5uF at input, 0.1 between Q1 and Q2, and 0.01uf on Q3 collector) the capacitor blocks DC levels unless it is itself shorted or leaky. So we can consider each section in isolation.

One more thing: germanium is NOT silicon. It leaks, is more highly variable, and works at lower voltages. Some things work with germanium that won't work with silicon, and vice versa. That's one way of saying that you cannot simply swap germanium PNPs and silicon 2N3906's and have a reasonable expectation of operation both ways.

Now to the circuit and voltages.
=============================================
Q1
Q1 is a PNP emitter follower. It differs from silicon ones in that it is leakage biased. That means that the leakage through the collector base junction multiplied by the HFE set the operating point. As you can imagine, this is less than a precise condition. The 100K resistor sucks some of the "excess" leakage out the base to keep things in order. That resistor sets the operating point in conjunction with the amount of leakage.  

Your Q1 with its particular amount of leakage and gain and your 100K resistor lead the collector to bias at 6.7V - much better for amplifying than the close-to-off of the ROG one. This stage is working OK.

It *won't* work OK with a silicon 2N3906. The silicon will not leak enough to bias it properly.

Q2/Q3
Now we're debugging a fuzz face. Notice that except for leakage, all the bias for Q2 comes from the emitter of Q3. This is typically sitting at a fraction of a volt and yours is sitting at - zero! The base of Q3 is sitting at 72mv, and its collector is at almost the battery voltage. This is the reason the trimpot won't help - Q3 is conducting little, if any, and the bias for the whole Q2/Q3 thing is off.

I think you have a problem with the emitter of Q3. Remove power, pull out Q3 (you don't say, but from the casual ease you mention swapping transistors, you have them in sockets). Now measure the resistance between the emitter socket hole of the Q3 socket and ground. Should be 1K. If it's not, you have a wiring, solder, or pot hookup problem that needs to be rectified.

While you have Q3 out, hook your ohmmeter between Q3 collector socket hole and the battery lead. Verify that the trimpot will run from 1K to 21K.

Write back with what you find. And get a good battery in there.

[analog kid]

Thank you RG, but Let me first tell you.. I understand you can't help me properly if I don't give you the CORRECT information!! And when you said that my Q3 emitter is at 0, I knew that wasn't right So I checked them all again That Q3 emitter WAS a mistype. !! It's around 72mV
This time for Q3 I got
C  7.9v all the way down And 8.3v all the way up
B  77mV
E  72mV
Q2
C 78.5mV
B 54.6mV
E 0
Q1
C 7.22V
B 59mV
E 0

The other measurements WERE what I took. These do tend to drift a bit from read to read and I know I should always use the same meter but as I have 2 of them in use while taking hfe/leakage measurements I may have used  different ones on the fist vltg readings
I'm sorry about the Emitter mistake... Now for the other info you asked for (IF it's still relevant!!!
My Q3 Emitter socket is at   4.66k WITHOUT POWER
                                   and  4.75 WITH POWER
Resistance of Q3 Collector socket and Battery runs from .488k to 20.53
Hope this helps the process.

[petemoore]

Thanks RG, I kind of was off there about Si pretesting on a TB...I've gotten spatty Fuzz that way...and could tell the circuit was basically firing tho gated... :oops:
 Anyway...around the Q3 emitter seems like it might be a good place to look.
 The only way to ground from there is through the gain pot...1k.

[R.G.]

Pete - no embarassment needed. We're all (most especially me) learning here.

analog -

OK, that helps.
Back at Q3:

This time for Q3 I got
C 7.9v all the way down And 8.3v all the way up
B 77mV
E 72mV  

Weird. Only 5mv from base to emitter. Not normally enough to turn a PNP on. Well, let's use the tools and think.

With 72mv on the emitter and the emitter measuring 4.7K to ground, you have a current through the emitter of .072/4.7K = 15.3uA. That means the collector can only be from 1K*15uA to 21K*15uA, or 15mv to 0.321V below the battery. And that seems to match what you measure.

So we have uncovered one problem: the emitter pot is measuring 4.7K, not 1K. You need to get that fixed, although that may not be the whole problem. For now, solder a 1.2K resistor across the lugs of the pot to correct the bias condition as that pot affects bias.

So with Q3 not acting too weird, what's the problem??
Let's look at Q2.

Q2
C 78.5mV
B 54.6mV
E 0

For germanium the base-emitter junction turns on at a tiny fraction of a volt and gets fully massively saturated at about 0.2 to 0.3V. The 54.6mv is not unreasonable for a germanium, but the 78.5mv on the collector indicates that Q2 is almost fully saturated on, which doesn't match the base condition. Hmmm...

Germanium is tougher than silicon, as it is less definite about what's "on" and "off" for a forward junction.

We have to guess here. Q2 may be leaky, or it may have a soldering problem. A bum coupling cap is possible, but not too likely, and bad R's are very unusual given only that the builder put in the correct value.

So let's do some more resistance sleuthing. Take all the transistors out of socket. Measure resistance to ground for all bases and emitters and resistance to power supply for all collectors.

Sorry for not being able to just say "Aha! It's your base resistor!" or the equivalent. Germanium is less definite than silicon.

As a note to observers - the general procedure here is to have a model in your head of how transistors work, whether they are silicon, germanium, JFET or MOSFET, and to read the voltages in the circuit. Using those voltages, you can often simply say "that can't be working right". You also get to use Ohm's law on the voltages and resistors to figure currents, and that adds another layer to your picture of whether a device is operating correctly or not. There is usually one or more blindingly obvious fault that pops out of this. Germanium is sloppier, hence we're into second level sleuthing.

[analog kid]

Oh man, by Emitter Pot do you mean the Fuzz control pot?? Because I DID use a 5k pot there instead of 1-2k and i put it on the 'deviation from..' line on the "what to do when it doesn't.." post I made.  I thought that It wouldn't affect the way the circuit works But just the taper of the FUZZ control.? ? SO it DOES affect bias... hmm.......
Well I DID actually , believe it or not have a 1k resistor on there which brought it down to around  .8 k which I thought may be to low , then I put either a 4.7k or 2.2 k which brought it down to 1.5k BUT since I didn't get any "Remedy" to my problems from this adjustment, I have just left the 5k pot As is.   and although I take it you dont' expect it FIX the problem I'll adjust it back to 1.5k for the time being.

Now I'll go on to your Q2 problem ideas...
QUOTE:
We have to guess here. Q2 may be leaky, or it may have a soldering problem. A bum coupling cap is possible, but not too likely, and bad R's are very unusual given only that the builder put in the correct value.
IF it helps the amount of guessing, I have replaced the .1uf which I guess is the coupling cap(from a 100v orange drop to a smaller poly)  AND I've tried OTHER Germaniums with lower leakage measurements
        Now here are the resistances
               E        B
Q1      0 to .2   /  99.8k
Q2       0 to .2  /   104.6k
Q3      1.57k    / Ok this one's different first I got NOTHING(INFINITE) and had to edit this post when I took it again I get 11.87Meg Ohms!!Then since they're connected I touch the lead to Q2 collector and come back to it and get  anywhere from 2M up to 20+ Meg ohms!!! Is this a problem!?
     
C to power
Q1 9.88k
Q2 104.7k
Q3 is .4k to 20.5k depending on trimpot

[analog kid]

OK, I did some more messing with the Q3 Base reading....
It is strange and I do get an 0.F(INFINITE)on that socket sometimes when I let it sit and go back to touch it.  And I notice it will stay infinite for a couple of seconds and then either either Rise or Jump to around 30Meg OHMS And I swear at other times (if I take another E/B resistance to ground reading and go back to that one) It will be stuck on INFINITE and I'll have to "wake it up " by taking the resistance to Power first.
NOW this COULD just be my DMM "ranging" very easily but I don't know, It's strange

[R.G.]

Q3 1.57k / Ok this one's different first I got NOTHING(INFINITE) and had to edit this post when I took it again I get 11.87Meg Ohms!!Then since they're connected I touch the lead to Q2 collector and come back to it and get anywhere from 2M up to 20+ Meg ohms!!! Is this a problem!?

...ooops... Q3 base should be infinite from base to ground; my mistake, it connects only to Q2 collector. Measure from Q3 base to power.

[analog kid]

Good , that had me confused.
In that case then It's the same as Q2's Collector to power, 104.2k (which I assume at least that it should be since they're connected.)
 
ARE ALL THE READINGS THAT I'VE GIVEN CORRECT FOR WHAT YOU NEED FOR "DIAGNOSING" R.G.
THANKS

[analog kid]

Here I took the voltages again, this time with my range selectable DMM so they MAY be more accurate.. since I don't seem to be able figure much else productive to do with this thing til I hear back about the Tale those resistances tell.
Q1
C 9.90v
B 76.5mV
E  0
Q2
C  122.8mV
B 75.9mV
E 0
Q3
C 10.04v
B  122.8mV  
E  78.9mV
(both Q1 C and Q2 B  seem NOT to stabilize very well. Meaning one time they'll start at 100mV and the next time 120mv but will steadily keep rising very slowly whatever the initial reading.)

[analog kid]

Should I just give up on this one for now then : If you don't think the 'problem' can be isolated just let me know so I can decide , although it seemed it was getting narrowed down to either Q2 or Q3 Base and collector ( even though there are a few things involved there)
I have several NOS germaniums to work with and I'll spend tonight trying to select some more candidates for this, and other , fuzz circuits.
AC125 , AC128, OC75