Rangemaster bias question

[Frequent Fog]

Hi, I just registered for this forum. I'm a long term lurker.
Anyway, I have a question about a rangemaster I have breadboarded.
I biased it so the collector reads -7v. BUT, when I start playing through it the collector voltage drops quite a bit, to around -6.4v or even less, it varies.
Does this mean I have an "unstable" transistor?

[Rob Strand]

7V is a static measurement with no signal.

When you put a signal through the unit the average (/DC) voltage on the collector can shift.  All normal behaviour.   It's not that the transistor is unstable it's related to the positive AC swing on the collector not being symmetrical with the negative AC swing.    As a result the average (/DC) voltage can change.

(Imagine a stage clipping on only one of the peaks.  The positive AC signal doesn't equal the negative AC signal, so the average (/DC)  level must shift.)

[PRR]

That's your "sweet even-order distortion". (Not always sweet but that's the cliché.)

You could actually learn some by watching the "DC" vary with signal, but nobody does. Perhaps because needle-meters are obsolete and digital meters don't show transients and trends well.

[Frequent Fog]

7V is a static measurement with no signal.

When you put a signal through the unit the average (/DC) voltage on the collector can shift.  All normal behaviour.   It's not that the transistor is unstable it's related to the positive AC swing on the collector not being symmetrical with the negative AC swing.    As a result the average (/DC) voltage can change.

(Imagine a stage clipping on only one of the peaks.  The positive AC signal doesn't equal the negative AC signal, so the average (/DC)  level must shift.)

Thanks!
I played it for a while, and checked the voltage again w/o signal and it was down to around -5.5v. I unplugged the power and tested the transistor and the leakage was over 500 microamps. After taking it out of the circuit an letting it rest the leakage is back down to 60 microamps, which is what it was to begin with. It is a Tung Sol 2N321 from 1962.

[Frequent Fog]

That's your "sweet even-order distortion". (Not always sweet but that's the cliché.)

Does that mean the way the rangemaster affects the amps distortion? I was always puzzled by that in the RG Keen article because it just sounds clean on it's own.

[Rob Strand]

Thanks!
I played it for a while, and checked the voltage again w/o signal and it was down to around -5.5v. I unplugged the power and tested the transistor and the leakage was over 500 microamps. After taking it out of the circuit an letting it rest the leakage is back down to 60 microamps, which is what it was to begin with. It is a Tung Sol 2N321 from 1962.

Shifts without signal are going to be due to leakage, bad connections, bad electrolytic caps.

The act of removing the transistor can affect things and I've found the leakage moves around in the jig.

If you had to desolder the transistor then the leakage will skyrocket and die down over about 15min to 30min after that.  Leakage depends on temperature and desoldering puts a lot of heat into the package.   Even touching the transistor can heat it up and affect the leakage.

You can measure the collector and/or emitter voltage in circuit then switch out the base so the base is open circuit then measure the collector and/or emitter voltage with the base open.   Don't touch the transistor or the leads when you disconnect the base.  Measure the voltages immediately after switching and at a few point in time afterwards, say at 30 sec to 1 minute intervals.    Measuring this way at least avoids touching the transistor.    If the leakage is very high the collector voltage and emitter voltage will get stuck at around 2.5V until the leakage drops.   Maybe in this test you won't see any change in leakage.

Another test is to place the transistor in the leakage test jig with the power off for 1/2 hr then power it up and see how the leakage moves over time after power up.  Again, this way you aren't handling the transistor.

Working out what's going on always takes a bit of effort.

[Frequent Fog]

It's on a breadboard, and it DID have a loose connection: the wire from the level pot to the collector kept coming loose. I didn't realize that could affect leakage. I did have to re-bias from a 48k resistor to a 39k, but it seems stable now.
I don't have a proper leakage test jig yet, but intend to make one. I'm using a $20 component tester.
Thanks!

[Rob Strand]

It's on a breadboard, and it DID have a loose connection: the wire from the level pot to the collector kept coming loose. I didn't realize that could affect leakage. I did have to re-bias from a 48k resistor to a 39k, but it seems stable now.
I don't have a proper leakage test jig yet, but intend to make one. I'm using a $20 component tester.
Thanks!

Bad connections cause high resistance joints so it's like the 10k pot value is fluctuating instead of the leakage.   Both will affect your collector voltage.

The power down trick should work on the component tester.   I guess you don't know what signals are put through the device under test at power up while the tester is trying to work out what parts is present  - it could make the leakage go up a bit, then it will settle down.