Transistors

[estch71]

Hi,

Is it possible to use two different transistors in effects?  What I mean is, can I use two different NPNs, two different PNPs, or two different Germanium types.

Thanks

[armdnrdy]

That's a pretty broad question.

I believe that it would depend on the circuit...or certain sections of the circuit.

Anything specific in mind?

[estch71]

Thanks for replying.  I've been looking at some fuzz schematics that call for different kinds of transistors - silicon, germanium, and they list types to use.  I guess what I was asking is do I have to use matching sets?  For example, there's a silicon based fuzz that I wanted to build and I only have singles of transistors - BC types.  Can I mix and match?

[armdnrdy]

Post a link to the schematic of the circuit you want to build.

[estch71]

[Mike Burgundy]

Yes you can, and this has been done. Do realize different transistor types may require different bias schemes/settings, and PNP and NPN usually require different supply setups - that's where the whole positive ground fuzzface thing comes from. Then again, there's stuff out there using PNP and NPN in the same circuit.
So yes, but it might not work (well) without further tweaking.

[estch71]

Thanks!  Really appreciate it.  I've been a guitar technician for years but have just started getting into pedal building.  So far, I've built a SHO clone, Valvecaster, a ground loop controller, and a feedback loop.  I just need to learn more about circuitry and voltage, among other things.

[armdnrdy]

And here, for your reading pleasure, is an informative tutorial at Geofex on HFE testing courtesy of R.G. Keen:

http://www.geofex.com/Article_Folders/ffselect.htm

[Kipper4]

I recently made both an npn and a pop fuzz face.
I only started building last Christmas.
I ordered ac128 especially for the pop but some how it didn't work well so I replaced the q1 transistor with a silicon pop ans its wonderful.
But I guess it depends on the circuit and circumstances.
I wish you luck in your quest I hope you enjoy this as much as I do
Rich

[Kipper4]

Here's the thread

http://www.diystompboxes.com/smfforum/index.php?topic=102487.msg909829#msg909829
I doubt it will help just me rambling

[LucifersTrip]

you'll realize with time, that in the end, for basic (and even many more complicated) bjt fuzzes & distortions, the part number means very little and the hfe and leakage (for ge's) is the most important. I wouldn't worry about mixing, unless the circuit specifically calls for matching.

there's tons of fuzzes that mix npn/pnp/ge/si:

http://www.mosritecelebrity.com/_wp_generated/wpac7c8b52.gif
http://4.bp.blogspot.com/-VHWoNuANSrk/T2ZonvSd4tI/AAAAAAAABIs/SdEPbILxvl8/s1600/SOLASOUNDSLTDMKI.jpg
http://home.centurytel.net/flanneldrawers/Jordan_Bosstone_schematic.jpg
http://www.rawbw.com/~emiller/Euthymia/DIY/HeathkitT28Fuzz.jpg
etc

[Thecomedian]

Op, this is what you're looking for

http://en.wikipedia.org/wiki/Sziklai_pair

You may find a discrepancy when you view the darlington pair and notice it is Emitter Q1 to Base Q2, while Fuzz is Collector Q1 to Base Q2.

The Sizzleachoo pair you will see is Collector Q1 to Base Q2. taking the above inversion, it may be meant to be Emitter Q1 to base Q2.

The other possibility is that you must change where the resistors go in the circuit to accommodate the emitter of the PNP which will be connected to positive just like the collector of the NPN, and then take the voltage Out from the collector of the PNP

I now have a new project to breadboard.



http://www.diystompboxes.com/smfforum/index.php?;topic=38394.0
http://www.diystompboxes.com/smfforum/index.php?;topic=57715.0
found some necro threads

A decent example circuit for a complimentary pair
http://www.radio-electronics.com/info/circuits/transistor_2_device_amp/two_transistor_amplifier.php

[Mark Hammer]

The original question has an assumption underneath it, regarding what part numbers mean, and whether they reflect wholly different parameters versus something more similar.  Research into how children acquire words during their first few years suggests that they operate under the priciple that "new words must refer to something I don't already have a word for".  If I point to a ball and say "purple", the child who knows "ball" will assume that "purple" refers to something other than a ball, and perhaps a quality of the ball.  Similarly, we tend to assume that because there is a different number for a given transistor that it must be unique in its functioning and unlike other transistors with other numbers.

The different part numbers reflect devices that are not completely overlapping in their characteristics, but certainly far from completely unalike.  I guess it's a bit like asking whether or not a Dachshund can mate with a German Shepherd.  It may be a little clumsy (as I once witnessed in a neighbours back yard, the Dachshund was the female of the pair), but they are both dogs and it CAN happen.  And the result can be productive, even if it required some tweaking to make it work.  Neither a Dachshund of a German Shepherd could mate with a guinea pig or a goat, however.

Similarly, A 2N3392 and a 2N3904 or 2N5089 can "mate" with  BC549.  As small-signal NPNs, they are all capable of working with each other, even if some adjustment to biasing may be required to do so.  They are all "dogs" and able to mate.  Much like trying to mate a dog and a cat, however, subbing a PNP for an NPN or a power transistor for a small-signal is unwise.

[pinkjimiphoton]

fwiw, yes, you can pretty much mix and match any npn or pnp type transistors  in a circuit. not npn AND pnp (some cases you can) but npn transistor will often work with any othe npn transistor, etc.

socket. experiment. you'll find all kinds of interesting possibilities.

[chromesphere]

The original question has an assumption underneath it, regarding what part numbers mean, and whether they reflect wholly different parameters versus something more similar.  Research into how children acquire words during their first few years suggests that they operate under the priciple that "new words must refer to something I don't already have a word for".  If I point to a ball and say "purple", the child who knows "ball" will assume that "purple" refers to something other than a ball, and perhaps a quality of the ball.  Similarly, we tend to assume that because there is a different number for a given transistor that it must be unique in its functioning and unlike other transistors with other numbers.

The different part numbers reflect devices that are not completely overlapping in their characteristics, but certainly far from completely unalike.  I guess it's a bit like asking whether or not a Dachshund can mate with a German Shepherd.  It may be a little clumsy (as I once witnessed in a neighbours back yard, the Dachshund was the female of the pair), but they are both dogs and it CAN happen.  And the result can be productive, even if it required some tweaking to make it work.  Neither a Dachshund of a German Shepherd could mate with a guinea pig or a goat, however.

Similarly, A 2N3392 and a 2N3904 or 2N5089 can "mate" with  BC549.  As small-signal NPNs, they are all capable of working with each other, even if some adjustment to biasing may be required to do so.  They are all "dogs" and able to mate.  Much like trying to mate a dog and a cat, however, subbing a PNP for an NPN or a power transistor for a small-signal is unwise.

Thanks Mark, i will never think of darlington (or sziklai) pairs the same again.