Biasing NPN germanium transistors in Tone Bender Pro MKII

[sirdavy]

TL:DR: Tone Bender Pro MKII - Why can I only achieve a range of 0-3.6V when adjusting the trimpot for Q3?

Hello! I have built the (Fuzzdog) Tone Bender Pro MKII and it works and sounds good on first test with guitar, however, when it comes to biasing the transistors, the collector voltage ranges I see when adjusting the trimpots for Q2 and Q3 do not match those in the instructions.

The instructions suggest adjusting trimpots T1(100K) and then T2 (10K) to achieve voltages around:

Q2 - 1.0 - 1.75V
Q3 - 8.0 - 8.5V

When I adjust T1 I can see a range of voltages between 0 - ~8V, and I can achieve 1.75V although it is with very, very fine movements like cracking a safe. (FYI adjusting to anything above ~3.5v and the sound from the pedal starts to disappear)

BUT, when I adjust T2 the maximum voltage I can achieve is ~3.6V, which is nothing close to 8.0 - 8.5V.

Fwiw Q1 in this circuit has a fixed bias resistor - the collector reads 0.27V.
I built the Tone Bender using the following NPN transistors:

Q1 - NPN OC140 - HFE=65 Leak= 0.050
Q2 - NPN OC140 - HFE=67 Leak= 0.033
Q3 - NPN OC140 - HFE=113 Leak= 0.036

As per instructions I reversed reversed the polarity of the orientation/polarity of C2, C4 and C6, and wired up as per a typical centre-negative power supply. (The caps I used for C2 and C6 were 4.7uf Non-polarised, so, as I understood it, their orientation did not matter).

Is a max collector voltage for ~3.6V expected with my components? Where should I look for issues causing the lower available voltage on Q3's collector?

[mozz]

Adjusting T1 will change the voltage on both Q2,Q3. Adjusting the T2 will only change Q3.  2 meters is helpful. Leakages are kinda low and i think Q2c should be lower then they say. Someone will chime in with voltages. I am thinking .2v for Q2C.

[Rob Strand]

Adjusting T1 will change the voltage on both Q2,Q3. Adjusting the T2 will only change Q3.  2 meters is helpful. Leakages are kinda low and i think Q2c should be lower then they say. Someone will chime in with voltages. I am thinking .2v for Q2C.

Yes something not right with the target voltages.  I don't have my notes and I can no longer quote the voltages off the top of my head.

Perhaps this can be used as a hint,
https://aionfx.com/app/files/docs/deimos_legacy_documentation.pdf
"Q1: Collector -8.5v, Base -0.05v, Emitter 0v
Q2: Collector -0.15v, Base -0.07v, Emitter 0v
Q3: Collector -8.5v, Base -0.15v, Emitter -0.1v"

I agree with Mozz the voltage at Q2.C and Q3.C can affect each other during adjustment, especially when the adjustments are off at the start.

The thing is,  to get 1.75V on Q2.C the emitter of Q3.e must sit at 1.75-0.15 = 1.6V.  That means 1.6V/1k = 1.6mA through the Attack pot.   That same current must flow through the collector of Q3.  To get Q3.C at 8V, you would need Rc = (9-8)/1.6mA = 625 ohm.   That's just a crazy value and not much larger than the 470 ohm.

For Q3.C at 8V normally the collector current would normally be quite low.    We can work the other way:, say T3 is set to half way at 5k.   IC3 = (9-8)/5k = 200uA.   Then the voltage voltage drop across the Attack pot would be 1k*200uA = 0.2V.  So the voltage at Q2.C would be 0.2+0.15 = 0.35V.  With T3 set to 10k that would drop to 0.1+0.15 = 0.25V.

So Mozz's 0.2V looks far more ball-park.

*Setting* voltages on Q2.C is always pretty dodgy as it's subject to a lot of variables.  Trying to prescribe a voltage can easily screw up things elsewhere.

Too much leakage in Q1 will cause the Q1.C voltage to drop.  You can only decrease Q1's base resistor or collector resistor to compensate.

[mozz]

Actually working on sort of this right now. Guy has a original FZ1 and needed unobtainable 2n270. So after searching i found 2n408's were sort of recommended. I had boards made up for FZ1 and FZ1a. I have a bunch of 2n408's and a pile of RCA house marked numbers. In the corner of my workbench, now i spy a old project, perf tonebender MKII with sockets, so i start messing with it. I'm thinking i can make the FZ1 boards into a tonebender boards with a few jumpers or cuts. So i got the tonebender working with the house marked RCA (less leakage than the FZ1 needs) and while searching for voltages i found: Q1C 9.0v Q2C .17v Q3C 8.4v.(Elec warrior/ Lucifers trip most likely).

Mine are running at 8.7v, .34v, 7.5v so i am close. Soundwise has yet to be determined due to wife sleeping. Gains are slightly over 100 and leakages are ~150ua. It's been hot (then cold) here in the garage and i can't believe any voltages or leakages i am currently measuring.

I don't understand more than a few of these pcb websites have a lot of voltages wrong, or sort of out of the ballpark. A little reading here does wonders on first hand experiences.

[Rob Strand]

ine are running at 8.7v, .34v, 7.5v so i am close. Soundwise has yet to be determined due to wife sleeping. Gains are slightly over 100 and leakages are ~150ua. It's been hot (then cold) here in the garage and i can't believe any voltages or leakages i am currently measuring.

I believe those values are from ElectricWarrior.  I just checked some easy to get notes and he had:
Vcc -9.67V, Q1.C -9.02V, Q2.C -0.17V, Q3.C -8.44V.

Another issue with some of the quoted voltages is they don't specify Vcc.  For example, Q1.C at 9V doesn't make sense for a 9V supply rail.  What it's really saying is 9.67-9.02 = 0.65V below whatever Vcc is.  Also for Q3.C.  8.44V really means 9.67-8.44 = 1.23V below whatever Vcc is.  You can see that if you used 8.44V with Vcc=9V you will get an incorrect figure of 0.54V below Vcc.  It might sound totally fine but it's not the correct way to set-up the voltage based on ElectricWarrior measurements.

In my detailed notes I re-jigged a whole lot of results to be Vcc-Vc, not only my results but from other people as well.   For Q2.C it makes more sense to keep the raw value.

[m_charles]

welcome to the wild world of germanium tone benders (and the reason it's easy to make a tone bender, but very difficult to make a great sounding tone bender).
I like q3 collector around 7v, but that's just me. Indeed as others have said the transistors are very interactive with one another. All of that said, if t2 isn't causing a significant sweep in q3's collector voltage, then something is definitely wrong. Are you positive you didn't mix up the trim pots? positive you didn't mixup any of the other pots? orientated your transistors correctly? my guess is that it isn't a subtle transistor thing, my guess is you've got something mixed up.

[mozz]

Leakages seem to low , he posted them in the first post.

"In my detailed notes I re-jigged a whole lot of results to be Vcc-Vc, not only my results but from other people as well. "
There was also a percentage value, which is better to use i don't know.
q1c 92% of vcc
q2c 1.82%
q3c 89%

[sirdavy]

My voltage at Q1c (0.28v) seems drastically lower than the measurements you've all suggested here (~ -8v). Where in the circuit could I have screwed up so badly?

The instructions for this Tone Bender is for PNP transistors, but it includes the simple changes required for NPN transistors, which I am using. Do the collector voltages change when one uses NPNs?

[mozz]

Measure from ground to the point.  Your reading should be positive. 

[sirdavy]

Measure from ground to the point.  Your reading should be positive. 

I'm using my dmm (set to 9v range) and measuring with my black negative probe on the ground of the PCB and the red positive probe on the collector (red dot) of the transistor. I am getting a very consistent +0.28V.

[mozz]

Well there is a 10k collector resistor on q1, other side of it should be 9v.  If you have .28v on the collector, your transistor is shorted, it's in wrong, resistor is wrong.  That part of the circuit is DC isolated from the rest of the transistors.  Check for solder shorts, c3 shorts,

[sirdavy]

Not sure if this is significant:

With dmm negative probe on PCB ground:-

Touching either lead of R2 = +0.28 & +9.8V  (power supply is 9.8V)
Touching either lead of R1 = 0 & 0


R1 = 100k
R2 = 10k

[mozz]

R1 is going to the base and the voltage there is going to be small.  Since you are using low leakage transistors, might not be able to measure much.  Still, .28v on the collector means the transistor is drawing a lot of current, which points to a bad transistor or installed wrong.  Can you swap it out? Check c3 for shorts. 

Don't despair, will get it working soon.

[sirdavy]

I have swapped it out for another oc140 and the voltage is the same. In fact, without a transistor in the socket, if I touch the dmm probe to the empty collector socket I get the same +0.28V

By C3 shorts, you mean a short on the cap with that assignation?

[mozz]

Ok, the 10k collector resistor. 1 end is going to your supply, 9v. The other end, with nothing connected, would still read 9v, if it is reading lower, that means there is current flowing through it. So, the only other thing connected to it is C3, it would have to be shorted. Only other thing i can think of is a bad ground somewhere. Hoping someone else is checking my troubleshooting here. Some pictures would help, of top and bottom of pcb.

[sirdavy]

Hi @mozz, thank you for sticking with me here.

OK, so I desoldered R2 and C3 from the circuit, and unplugged Q1 & Q2, as per the images (hope you can see them OK). (Btw it looks like there are four sockets for the transistors but the middle one between the emitter and collector is false/unsoldered and just there to provide more support).

I put a new 10k resistor through the holes in the same spot as R2 and it still measures ~ +9V on one side and 0.28V on the other. I might be showing up my naivety here but for a test I put a 10k resistor across my 9V supply and it's also +9V on one side and 0.28V on the other. So, wouldn't those measurements then be expected in the circuit?

(Also, as I said in my first post, even with the way off biasing settings of Q2 - 1.75V & Q3 - 3.6V, the Tone Bender seemed to work, and sounded good, although I don't really know what to expect)

[sirdavy]

[mozz]

" I put a 10k resistor across my 9V supply and it's also +9V on one side and 0.28V on the other"
 I don't understand that, you must have a bad meter or dead battery in it.

If you put a 10 k resistor across a 9v battery, you are gonna measure 9v. Slowly it will go down as the battery depletes.

Can't really tell anything from those pictures.

[Rob Strand]

There was also a percentage value, which is better to use i don't know.
q1c 92% of vcc
q2c 1.82%
q3c 89%

What is best depends on the circuit.

It doesn't make a lot of difference provided the target percentages have enough digits.   92% of 9V is 8.28V, or 9-8.28 = 0.72V from the supply, whereas 92% of 9.5V is 8.74V is 9.5-8.74= 0.76V from the supply.   The difference is small in the scheme of things.   What you might not want is to round 1.82% to 2%.

The main point about measuring to the supply when then collector voltage is close to the supply voltage is it provides a much more reliable number as it essentially factors out the supply voltage.   On the other end of the scale is measuring to ground can end up with target voltages higher than the supply on the test unit if the supply voltage is low - completely unusable target voltages.

All the above only applies to measuring targets close to the supply voltage.  There's little advantage using that method for mid supply or less than mid supply voltages.   Nonetheless if you quote test/target voltages it doesn't hurt to say what the supply was.

[amptramp]

I don't think anyone has asked yet, but are you sure the transistor pinouts you are using are correct?  I notice that you had transistors with long leads, all insulated with red sleeving, soldered into the circuit and strange things happening to the pinout can occur with this setup.