debugging maestro fz-1a

[Harry]

Ok so I'm getting some ugly distortion out of this but it isn't fuzzy or gainy enough as it should be. I'm using small bear's pnp set. They were marked hfe 33/leakage 133 (used for q1), hfe 262/leakage 298 (q2), hfe 110/leakage 351 (q3). I went with the recommended bias resistor values 1k at r8 and 43k at r10.
Voltages
Q1
c -1.65
b -.52
e -.58
Q2
c -1.63
b -.03
Q3
c -1.58
b -1.57
Any ideas?

[LucifersTrip]

There's some problem with Q2 & Q3.  Both of those collectors should be lower (and Q3 B much lower), especially if the leakage reported is correct.
(Supply is 1.65V so if you use Q2/Q3 with very low leakage, it will be close to 1.65V). 300 leakage is alot.

The higher the leakage, the lower the collector voltage.  The larger R8/10,  the lower the collector voltage.

Q3C should be no higher than ~ 1.1 - 1.2V (and as low as ~ .7 or .8V) and Q2C was ~ .3 - .4V on my last one with fuzz maxed.

Look for errors around Q2/Q3, try different  R8/10's to lower voltage and measure the leakages to make sure they're correct.

edit:
Also, I believe for Q1 E should be < B

...and good luck!

[brett]

Hi
Q1 has an issue.
The base should be 0.3 volts more (-more) than the emitter.
It is turned around the wrong way?
The emitter voltage (0.58 volts) tells me that 0.58/10k = 0.058 mA is flowing through the emitter and emitter resistor. That's not much.
Is R1 actually 1 Meg? (so that almost all of the leakage current goes through the base-emitter junction)
cheers

[LucifersTrip]

Funny, I was just editing my last post about Q1, but you're not too far (on those voltages. yes, it's 1M on the base.

edit:
the last one I built: (hfe/leak): 98/330 , 96/330, 140/340

edit2:
just dug out notes on a couple of my 1960's originals (knockoffs w/ same topology)

Q1: .53, .65, 1.6
Q2: 0, .1, .86 (max)
Q3: 0, .03, 1.2*

Q1: .24, .29, 1.6
Q2: 0, .09, .98 (max)
Q3: 0, .06, 1.2*

*I believe resistor from Q3B to ground is 5K, which will make Q3C voltage on the higher end (1.2 instead of closer to .8)

[Harry]

Thanks for the responses!

Replaced r8 with 10k. Rechecked voltages: Q1 c 1.65/b .48/e .52     Q2 c 1.21/b .06/   Q3 c 1.58/b 1.57
1It sounds a bit better now but still not very "fuzzy". Should I keep raising r8 and r10 to bring the collector voltage down?

I don't know if Q1 is turned wrong. It's labeled CV7005 can't find a pinout online, but I'm assuming red dot is collector (cbe config). Maybe it's CEB pinout?

r1 is 100k. r2 is 1M as in the schematic.

[LucifersTrip]

see comments above. you have Q3 BC almost shorted or shorted...Q3B should be near 0

check for soldering errors.  are you sure at this point that all Q's are oriented correctly?


edit:
pinout...read "connections" on the right hand side. it's ebc
http://www.shinjo.info/frank/other/CVspecs/cv4-7a/CV7005.pdf

[Harry]

see comments above. you have Q3 BC almost shorted or shorted...Q3B should be near 0

check for soldering errors.  are you sure at this point that all Q's are oriented correctly?


edit:
pinout...read "connections" on the right hand side. it's ebc
http://www.shinjo.info/frank/other/CVspecs/cv4-7a/CV7005.pdf

You're awesome. It was a bad solder joint on Q3 it's now -.8 at c and -.02 at b. 

[Harry]

ok so just had a chance to play through it again still isn't sounding fuzzy enough.
q1  c1.65/b.89/e1.04
q2  c.82/b.06
q3. c.05/b.02
nothing has been done to it so i dont know why i wouldve lost voltage on q2 and q3's c???

[Harry]

oh also the voltages for the collectors vary widely whenever i check them. last time q3 was -.5 and q2 was -.9. is this normal?

[LucifersTrip]

nothing has been done to it so i dont know why i wouldve lost voltage on q2 and q3's c???

right after you touch or solder, you're heating them up and the leakage will be higher = lower C voltage.

oh also the voltages for the collectors vary widely whenever i check them. last time q3 was -.5 and q2 was -.9. is this normal?

that's what happens, especially with leaky ge's. they take time to settle. unfortunately, that's one of the things that make this circuit more difficult to get right.  I usually turn it on and take measurements after 15 minutes or so, then take measurements a short time later to make sure the readings have settled.

remember, you have readings from a working one:

EBC
Q1: .53, .65, 1.6
Q2: 0, .1, .86 (max fuzz...as low as .3 or .4)
Q3: 0, .03, 1.2 (or as low as .7 or .8)

so you still need to sort out Q1. remember E voltage is < B < C

edit:
Q3C should never get that low...typo?

q3. c.05

[Harry]

yeah no typo. q3 fluctuates widely. left it on for fifty minutes q3c goes from 1.08 to .61 to .16 to .4 etc. it never stabilizes. none of the voltages stabilize in fact.

[brett]

Hi
CV7005 = OC71 (possibly a militiary  spec) in a TO-1 case
Looking from underneath the sequence, left to right, is emitter-base-collector.
Hope that helps.
cheers

[Harry]

hi Brett
lucifer actually posted the spec sheet for cv7005. it is wired correctly, but thanks.

i cant figure out how to bias this because the voltage on its collector is totally different everytime i check it.

also for q1 should i raise r3 to get its emitter less than its base?

[LucifersTrip]

i cant figure out how to bias this because the voltage on its collector is totally different everytime i check it.

are you touching the transistors when you measure them?  temperature change (even in the room) will change the hfe/leakage and the voltage in the end.
though, you have a pretty big range...too much

that would suggest a poor solder joint / intermittent connection, bad transistor*, or even a bad power supply (are you using a AA battery or another method)

*are they already soldered...can you measure hfe/leakage?

[Harry]

not touching transistors they are soldered in. using as battery. temperature has been pretty wild here but the voltages are changing within minutes. ill check the solder joints again...

[Harry]

...all joints appear to have continuity. i guess ill reflow all the joints just to be sure its not some intermittent problem.

oh yeah and the battery is still reading 1.65 volts but ill pop a different one in for the hell of it.

[brett]

Hi again
let's ignore Q1 for now.
Q2 might be working.
Q3 looks like either 1. a defective device or 2. a short-circuit in the physical layout, or 3. the collector resistor is way too large* (? not sure about this).
The collector is within a whisker of ground. So is the base. The emitter must be at ground. Maybe there's a base-emitter short either internally or externally.

*way too large includes an open circuit (is there -1.6 V on the power end of the 10k resistor? What about the other end?)
good luck - you'll get there

[Harry]

ok so i changed q3 bias resistor to 10k. voltages appear to stabilize now with the following:
q1 b .64  e .71
q2 c 1.04  b .06
q3 c 1.34  b .02

EDIT: just test drove it... sounds pretty great! nice fuzzy sound with short decay. im actually quite satisfied with it at this point but still would like to get these voltages more on par, hopefully itll milk out a little more gain which would be a plus.

[brett]

Hi
the large voltage drop between the collector and base of Q2 and Q3 makes me think that there isn't much leakage current flowing. With the base at almost ground, only part of the signal is being amplified (less than 0.25V, so that the signal and the bias voltage add up to 0.3V and the transistor 'tuns on'). That's a major reason why the signal decays so fast.
The easiest way to simulate leakage is to put a resistor of about 100 k between the collector and base of Q2 and Q3. A smaller value (22k ??) will create 'better' bias, more gain and more sustain, while a larger value (1M ??) will leave it 'starved' and 'blatty' with a quick decay. (with a collector-base difference of about 1V, a 100k resistor gives about 10uA of extra 'leakage' current. Note that these leakage numbers are different to those measured with RG's tester, which works at 9V)
A more authentic way to simulate leakage is to find 2 leaky germanium transistors and put the c-b junctions in parallel with your Q2 and Q3. (of course, you could just use them as Q2 and Q3, too).
For me, I'd just use a resistor, and change the value until it sounds moderately misbiased like the originals.
cheers

[LucifersTrip]

the large voltage drop between the collector and base of Q2 and Q3 makes me think that there isn't much leakage current flowing.

that's the confusing thing...he got those from Small Bear with hfe 262/leakage 298 (q2), hfe 110/leakage 351 (q3). With those numbers, Q2 and Q3C voltage should be lower...

anyway, I would just slowly increase the B to ground resistors until I got closer to ~ .6 - .8V for Q2 and .8 - 1.2V for Q3