Jen Fuzz voltages?

[LucifersTrip]

I pride myself on having zero unfinished pedals and zero pedals in the graveyard...I don't have any in the graveyard but I do have 2 incomplete, one of which I built a year ago and didn't have a box for...so I never finished tweaking it:



Ironically, I made one of my first posts here a year ago asking for voltage readings for this pedal and a data sheet for the 2N1149 which I was using (I actually didn't have a DMM at the time with hfe measuring and chose the 2N1149 since it crossed). I couldn't understand why I was only getting a mild distortion. Unfortunately, no one was able to help me with any info...It turns out the 2N1149 was an ancient grown junction transistor with a gain of 7 !!!  

For the voltages (using all 2n5088 with hfe ~ 350-400), before tweaking, I have:

EBC
Q1 > 6.9, 7.01, 9.08
Q2 > 0, .58, 1.25
Q3 > .64, 1.25, 4.97

I'm getting a good fuzz (almost boxworthy), with the only minor problem some light feedback (not hum, hiss, or oscillation) when the fuzz is maxxed and no strings are being hit.

Q1 voltages are suspect, especially Q1E

So, 1 year later, does anyone have any ballbark voltages for this one or a similar circuit?

thanx again


---------
Btw, here's an actual datasheet for the ancient 2N1149 from a killer Russian site:
http://transistor-spravochnik.ru/description/2n1149/527
...but they have E & B reversed

[Gus]

I am not a fan of the biasing used for Q1.  " Read the The Art of Electronics" section about that bias or other textbooks.  Q1 is a EF gain of about 1

Look at the first stage of this it has an  EF first stage
 

For a gain of about X10 look at this first stage

[LucifersTrip]

I am not a fan of the biasing used for Q1.  " Read the The Art of Electronics" section about that bias or other textbooks.  Q1 is a EF gain of about 1

Thanx much...I'll look into that.

I guess my questions are:

1) Is Q1 supposed to work like a standard linear amplifier or is the small difference between E-B voltages normal for an EF stage?

"For linear amplifying, the collector must be more positive than the base; the base must be more positive than the emitter by about 0.4 to 0.7V for silicon"

2) If I get ~ 4.5V (I am) on Q2's collector, does it matter what Q1's voltages are?

[LucifersTrip]

I replaced Q1 with PN2222, hfe 220

Q1 > 5.96, 6.12, 9.04
Q2 > 0, .58, 1.28
Q3 > .67, 1.28, 4.71

Almost exactly the same, with the biggest changes Q1 EB & Q3C
I believe it may actually sound a little better (smoother, which would be expected with Q3C closer to 4.5v).

Still (Q1 > 5.96, 6.12, 9.04) remains suspect with 5.96 too close to 6.12?

For kicks, I will de-solder one final time and lower Q1 to 100 just to see if it affects Q3C further.

---------
correction to my earlier post regarding the 2N1149 Russian datasheet
http://transistor-spravochnik.ru/description/2n1149/527
E & C are reversed not E & B

[Gus]

Again  search for emitter follower design.  IMO the transistor bias using a resistor from base to + like used in the first post has issues, both transistor hfe and TEMP.

An emitter follower often has an output level a little smaller than the input.  So think about this what is you input level and what is the emitter DC voltage?

Maybe build the 3 transistor EF test circuit and install different hfre transistors in the circuit and note the emitter voltage with each transistor(emitter voltage offset from 1/2 by design), then use the same transistors in a test circuit with the EF biasing used in the the Jen circuit.

[LucifersTrip]

Again  search for emitter follower design.  IMO the transistor bias using a resistor from base to + like used in the first post has issues, both transistor hfe and TEMP.

An emitter follower often has an output level a little smaller than the input.  So think about this what is you input level and what is the emitter DC voltage?

Maybe build the 3 transistor EF test circuit and install different hfre transistors in the circuit and note the emitter voltage with each transistor(emitter voltage offset from 1/2 by design), then use the same transistors in a test circuit with the EF biasing used in the the Jen circuit.

Yes, I did a bunch of EF searches after your last post.  R.G. wrote a nice quick one:
http://www.diystompboxes.com/smfforum/index.php?topic=34551.msg241769#msg241769

I'm not trying to re-design the circuit/stage, especially since I'm getting a really good sound now. I'm just trying to find  some notes on what type of voltages I should be getting for Q1.

I read jmusser's build report & he actually describes EXACTLY what I'm getting, so I guess I'm almost satisfied:
http://www.diystompboxes.com/smfforum/index.php?topic=43036.0

thanx again for the info

[John Lyons]

Take a look at Gus's first schematic there.
The pull up resistor (1M) on Q1 is the "problem" I think.
An Emitter Follower does not contribute and gain so rebiasing it
isn't going to change the sound for the worse. You can audio probe
the output to hear how it sounds now if you are curious.
It's used as a buffer here. I would either tie a 1M
to from from the base or 470k to +v and 1M to ground
and see what you get.

By the way. If you have not built an audio probe (or used one) then
I would HIGHLY recommend doing so. It is a key tool to understanding how
circuit blocks sound individually and for troubleshooting.

[glops]

By the way. If you have not built an audio probe (or used one) then
I would HIGHLY recommend doing so. It is a key tool to understanding how
circuit blocks sound individually and for troubleshooting.

I'm a little fried right now.  I built an audio probe a couple of years ago but never really used it.  Tonight I breadboarded a Roland Double beat fuzz and got the undertone of the fuzz barely coming through sound, which usually means there's at least one thing wrong in the circuit.  On my breadboard apparatus, I have in and outs on a terminal that go to the switch.  If I place the out to a pad and use another wire as my "probe" from the same pad, does that work the same as an audio probe?

[LucifersTrip]

Take a look at Gus's first schematic there.
The pull up resistor (1M) on Q1 is the "problem" I think.

An Emitter Follower does not contribute and gain so rebiasing it
isn't going to change the sound for the worse. You can audio probe
the output to hear how it sounds now if you are curious.
It's used as a buffer here. I would either tie a 1M
to from from the base or 470k to +v and 1M to ground
and see what you get.

yes, "problem".  For me, it's actually not a problem unless I know my Q1 voltage readings are actually off from what they should be, using that schematic.

So unless those voltage readings are off, there's no reason to make changes to the circuit, especially since I am getting a good fuzz.

After I first got those readings, I did try various pull up and pull down resistors to either no or worse effect.

As it stands now, I'm really close to boxing, but with 2 minor things nagging at me 1) are my Q1 voltages indeed correct for this schematic, regardless of the original "poor" design" and 2) I get feedback when the fuzz pot is in it's final 5%, which I have not been able to tame yet with B-C caps, etc....

Oh yeah, I did probe after the 1st stage and the sound was very similar...but there was heavy oscillation with the fuzz up.

thanx much for all the info

[Gus]

lets look up a BC238B transistor
http://www.datasheetcatalog.org/datasheet/motorola/BC239.pdf

I would guess the current gain would be somewhere around 200 (greater than .10ua and less than 2ma(9VDC/10K =.9ma,  5VDC/10K = .5ma etc))for the original transistor NOTE it is not well defined and changes with Ic.  Look at the Typ column numbers

IIIRC pages 71 and 72 in "The Art of Electronics" book has something about this type biasing.

[LucifersTrip]

lets look up a BC238B transistor
http://www.datasheetcatalog.org/datasheet/motorola/BC239.pdf

I would guess the current gain would be somewhere around 200 (greater than .10ua and less than 2ma(9VDC/10K =.9ma,  5VDC/10K = .5ma etc))for the original transistor NOTE it is not well defined and changes with Ic.  Look at the Typ column numbers

yes, the BC238B (Q1, Q2) is listed as typical 180. The datasheet I used said min 200:
http://transistor-spravochnik.ru/description/bc238b/23994

on the other hand, the BC239C (Q3) has min 420 (ironically, the 5088 I used has hfe 422)
http://transistor-spravochnik.ru/description/bc239c/24006

de-soldering Q2 and swapping for a lower hfe one was my next move....which I did last night.

recap: when I built this over a year ago, I just blindly followed the schematic ("2N5088 should work well"). The 2N5088's did work, but were probably not the gains (much higher) originally used.

In my current tweaking, I've replaced the 1st (2N5088) with a PN2222A (hfe 222). That lowered Q1 EBC a little but did not alter the sound (since the 1st is just an impedance matching stage)

I just  replaced the 2nd (2N5088, hfe 400+) with a PN2222A (hfe 224). Q3 collector voltage dropped from 4.71 to 4.6

That was the largest change.