Biasin' GE's..

[petemoore]

  Here's what I have
  PNP POS GND circuit using a couple GE's
  I'm using an input/DC blocking cap
  I have a bias string from +Ground>0k-100k pot>680k to -top rail. Base between each of the two rails via resistor.
  The collector has a 10k resistor to -top rail, and a DC output/blocking cap.
  The emitter has a little resistance [330-460 ohms] to ground.
  That's it...twice...here's the voltages...
  Collector -1.85vdc [1/2v also available with 100k pot base to ground adjustments]
  Base  -.25Vdc
  Emitter -.15Vdc
  Collector is way too low, base isn't far enough above the emitter.
  What would you do to try to spread these voltages out better over the 8.25v available from supply?
  every resistance has be triple/double checked for each stage...
  the second stage seems not to bias just like the first one.
  the reason I don't adjust to 1/2v on the collector using the base to ground 100kpot is that this doesn't really spread the base/emitter voltages into usable bias and the circuit sounds sort ok and boosts with the lower collector voltage I've chosen to show above.
  The thing looks like it should bias, but a string of bad bias luck has plagued me anyway...ANY suggestions welcome here...I think the measured resistances look workable, and I'm sort of reluctant to mess with the 680k's from base to -top rail [tight, and why]..the 10k's from C to -top rail...I thought about messing about with them, but it's really the emitter/base difference I'd like to spread out more.
 

[R.G.]

Germanium's dirty little secret is that the junction leakage is so bad and the junction voltage is so low that to turn a Ge transistor off you have to bias it to BELOW the emitter. In my first job as an engineerling I had to look over some logic circuits done with 2N404's. They all had an auxilliary supply for reverse biasing the base to turn it off.

What is happening to you is that your transistors leak enough that you don't need as much, or perhaps ANY of that bias resistor string from the -9V supply to the base. Many low-current Ge biasing setups are biased only with a resistor from base to ground. What that does is have the resistor suck some of the collector-base leakage out through the base to ground instead of going into the base-emitter and being amplified. This is the so-called leakage bias that you find in really old germanium circuits.

There is a relatively fixed leakage on each Ge transistor that varies from transistor to transistor and also with temperature. If the transistor is hooked up with a collector resistor, the current through the collector is beta times the collector base leakage, and that current appears as a voltage across the collector resistor. Make the collector resistor too big and the voltage on the collector will be very small - with NO biasing on the base at all. If you make the collector resistor smaller, the collector voltage comes up because the resistor drops less of the available voltage even though the collector current remains the same.

If you then put a resistor from base to emitter or base to ground, the collector voltage also comes, but this time because you are removing some of the internal leakage current from the base, so the current from the  collector through the collector  resistor decreases, and that raises the collector voltage.

What happens if you short the base to emitter? There is STILL a leakage current through the collector-base junction, and hence still a collector current. So to really turn a Ge base off you have to back bias it.

That's a long winded way of familiarizing you with biasing Ge. Your devices are too leaky for what you are trying to do, or you have too much base bias on the bases, or your collector resistors are too big, depending on how  you look at the problem.

[petemoore]

  The collector resistors measure 9k8 or ~10k
  The base to ground resistors can measure from 0k - 100k
  The really sorta big resistor is the base to V+, I'd suppose making that smaller would be the way to go, much bigger enough to speak approaches 1meg.
  What's sort of strange is the bias arrangement looks not bad, and tthe transistors are AC128's or many other GE's that came with leakage and Hfe documentation, which bore out in that they all seem to bias otherwise in RM or FF type circuits.

[petemoore]

 '*' means I quoted you.
'>'  means I'm typing this.

Germanium's dirty little secret is that the *"junction" leakage is so bad and the junction voltage is so low that to *"turn a Ge transistor off you have to bias it to BELOW the emitter". In my first job as an engineerling I had to look over some logic circuits done with 2N404's. They all had an auxilliary supply for reverse biasing the base to turn it off.
  >*Base/Emitter Junction.?
  >*'it' below the emitter...hmmm...ok...to turn the base off you need to bias it below the emitter voltage...IOW...it won't make an acceptable ~linear amplifier.
  What is happening to you is that your transistors leak enough that you don't need as much, or perhaps ANY of that bias resistor string from the -9V supply to the base.
  >I tried a 220k instead of the 680k, and that allowed the base to drop a bunch, then adjusting the 100k base to ground it came right back to ~^1/2v, however the emitter still 'follows' the base as it's bias voltage rises...I'm asking if this behaviour is there because base/emitter leakage is great. 
  >I will try opening the base/top rail resistance.
  Many low-current Ge biasing setups are biased only with a resistor from base to ground.
  >like Tonebender MkII
  What that does is have the resistor suck some of the collector-base leakage out through the base to ground instead of going into the base-emitter and being amplified. This is the so-called leakage bias that you find in really old germanium circuits.
  >Most of the transistors used [2SAxxx, 2SBxxx, NTE158's, AC 128's] compare well in other circuits to other transistors [some SB tested GE's] used in this circuit which have been tested for leakage and gain...
  There is a relatively fixed leakage on each Ge transistor that varies from transistor to transistor and also with temperature. If the transistor is hooked up with a collector resistor, the current through the collector is beta times the collector base leakage, and that current appears as a voltage across the collector resistor. Make the collector resistor too big and the voltage on the collector will be very small - with NO biasing on the base at all. If you make the collector resistor smaller, the collector voltage comes up because the resistor drops less of the available voltage even though the collector current remains the same.
  >I haven't messed with the 10k's as the collector resistors on each stage.
  >Now I have this:
  220k base to top rail [was 680k], 0k-100k pot base to ground
  10k collector resistor [hasn't changed]
  4k2 emitter to ground [I've tried values 0k - 4k2]
  The base follows the emitter closely [about .1v or so spread...both bewteen about 0v to 1.25V depending on the...100k and all other resistances tried.
  >>>What would you suggest to try for a larger spread between emitter and base...from what youv'e written I try to surmize...larger collector resistor and adjust base to get collector 'up where I want it....
  Bigger spread from base to emitter is what I'd like to get..
If you then put a resistor from base to emitter or base to ground, the collector voltage also comes, but this time because you are removing some of the internal leakage current from the base, so the current from the  collector through the collector  resistor decreases, and that raises the collector voltage.

What happens if you short the base to emitter? There is STILL a leakage current through the collector-base junction, and hence still a collector current. So to really turn a Ge base off you have to back bias it.

That's a long winded way of familiarizing you with biasing Ge. Your devices are too leaky for what you are trying to do, or you have too much base bias on the bases, or your collector resistors are too big, depending on how  you look at the problem.