Question about Atlas DCA gain/leakage calculations for Germaniums

[oddson]

Sorry to resurrect an old thread but it's more-or-less on topic from something I'm trying to understand.

So if you put a reverse bias on the collector-base and open circuit the emitter, a current flows through the collector and base. That is the leakage, and is about constant with voltage changes, but is quite temperature sensitive.

This is different from your web-site article where the base was left open and the leakage was taken to be the current passing between C and E (or are they the same for some reason I don't understand).

I thought it strange that your article tries to make the math easy by careful selection of resistors - wouldn't a simple spreadsheet formula work better since you can't read the real gain off the display anyway? The resistors could be tested and their values corrected in the formula (or just use 1% resistors) and then the two values (V leakage and V apparent gain) entered and 'true' gain and leakage values are spit out. I'd like to build such a thing but I don't think I have sufficient understanding.

Sidebar: Smallbear (in his Fuzzy-One project) notes that at low voltages the apparent leakage falls way off but gain seems to fall even faster. What voltage range provides 'stable' leakage and gain results (knowing full well it's not really stable!)?

[PRR]

>> put a reverse bias on the collector-base and open circuit the emitter, a current flows through the collector and base. That is the leakage

> This is different from your web-site article where the base was left open and the leakage was taken to be the current passing between C and E (or are they the same...

"Leakage" is specified different ways.

The first (open emitter) is "Icbo", leakage from C to B.

The second (open base) is "Iceo", which is Icbo times current gain at that current.

If you do a detailed design, Icbo guides your base bias circuit.

If you are dropping transistors into existing circuits, Iceo is often the better guide for "leaks too darn much".

> What voltage range provides 'stable' leakage and gain results

OK, suppose we knew. Say it was 0.2V. Is that reading useful when the actual pedal runs at 5V across the transistor? Yes there is a fundamental leakage plus a leakage due to C-E voltage. So you read the basic leakage then plot the added leakage due to voltage.

Perhaps the best test is try it in the actual circuit. That avoids multiple layers of measurements and computations. Does it bias-up to several volts on the collector? Or is there so much leakage the collector sags down to emitter voltage?

When contemplating several different "actual circuits" and faced with a bucket of unknown transistors, some sorting may be useful. Even so it is simplest to use a test-rig which aims at voltage and current similar to typical pedals. Otherwise the many variables get to be a headache.

> a simple spreadsheet

Some of R.G.'s writings are from days when spreadsheets were not universal in pedal workshops. Wasn't that long ago that The Computer was upstairs. Wasn't long before that when many musicians didn't have a computer. Now every Tablet and some SmartFones have computation software, but that's new.

[oddson]

thanks Paul... that helps.

My confusion comes mostly from talk of leakage, rated in uA, -- as if it were a single fundamental quantity (at a given temperature) when it seems these things pour current every which way and according to no clear pattern that I can find.

I'm in possession of a fair number of quite leaky Ge transistors that I know full well would make awful FFs (especially with stock resistor values but most no matter what you do with the basic FF circuit). So I'm trying to learn how leakage effects things in different stages of different biasing schemes to see if I can come up with my own designs.  I'm also trying to get a proper handle on just how leaky they really are. (And I'm also trying to understand the mojo factor of Ge and maybe mimic this a bit with Si.)

Sorry if I didn't seem to show the proper respect for RG...  (I'm not worthy! )
I was looking more to see if there was some unstated reason for the alleged simplicity of 'read the gain from the voltage' when you still have math to do anyway.

[PRR]

GE Transistor Manual 1964
http://i.imgur.com/0jt2X.gif