Page 2 of 5

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 02 Jun 2012, 13:43
by DrNomis
grrrunge wrote:
DrNomis wrote:Pretty interesting and cool ideas here, I must do some breadboarding of this someday.... :hmmm:


I really want to get started as well! If it wasn't for all that damn painting i've got to get over with :slap: Moving is a pain in the ass :lol:


I feel your pain, I've had to do at least 9 moves to a new place to live after moving to Darwin in 2000, was a really frustrating and grinding experience.... :thumbsup

And soon I'll be moving to a new house because it's no longer safe where I'm currently living, we've already had two people getting stabbed with a knife, one person died, the other person was more fortunate, mostly caused by drunken Aboriginal itinerants causing ruckuses.... :thumbsup

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 02 Jun 2012, 15:17
by grrrunge
IvIark wrote:Something like this?

Image


Jeez that was fast! I didn't even realize you had posted that until i went back to review the schem i uploaded :D
I've gotta try that!

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 02 Jun 2012, 15:44
by bato001
+1 to this whole thread. I am working on a fuzz using AC187 high hfe germaniums which are about $5.00 a piece. If I can substitute 25 cent 2n5088's and get the same thick, fizzy, creamy fuzz it would be a godsend.

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 02 Jun 2012, 18:23
by sinner
Nothing prospective from me. I just wanted to say - this tread is awesome :thumbsup

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 02 Jun 2012, 19:07
by DrNomis
I think it's good to see if we can find alternatives, if it's going to produce similar results while saving money....... :hmmm:

:thumbsup

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 02 Jun 2012, 19:53
by Duckman
Nothing to say, just... don't wanna miss this thread :lol:

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 04 Jun 2012, 14:19
by IvIark
The layout is now verified for anyone interested.

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 04 Jun 2012, 15:00
by grrrunge
IvIark wrote:The layout is now verified for anyone interested.


Nice! How does it sound?

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 04 Jun 2012, 15:05
by IvIark
I haven't built it myself, but the guy who did wrote
This one sounds really great. Ive always found fuzz faces to be dark and the low end loses all calrity when cranked, but that doesnt happen with this one. The bias knob allows for a huge range of classic fuzz tones too. Dig it.

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 04 Jun 2012, 15:45
by Ihunda
Great thread, I was wondering, couldn't the second transistor with collector removed just be replaced by a simple diode?

I made a quick schematic for it:


1. Take two identical random silicon transistors (for more mojo: your fav fuzz silicons
2. Tie the BASES together,
3. Cut off one COLLECTOR
4. connect a 3k to 6k resistor between the EMITTER of the collectorless devicr and the EMITTER of the other transistor.

By varying the resistance, you can dial in the gain you want...

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 04 Jun 2012, 16:13
by DrNomis
Ihunda wrote:Great thread, I was wondering, couldn't the second transistor with collector removed just be replaced by a simple diode?

I made a quick schematic for it:


1. Take two identical random silicon transistors (for more mojo: your fav fuzz silicons
2. Tie the BASES together,
3. Cut off one COLLECTOR
4. connect a 3k to 6k resistor between the EMITTER of the collectorless devicr and the EMITTER of the other transistor.

By varying the resistance, you can dial in the gain you want...



In theory that should work fine, since the B-E junction of a transistor is effectively a diode, we already know from what's posted on other FSB forum threads that you can use a Ge transistor B-E junction as a diode, so it shouldn't matter whether it's a Si or Ge type... :thumbsup

I reckon give it a go and see if it does in fact work.... :thumbsup

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 04 Jun 2012, 16:25
by grrrunge
DrNomis wrote:
Ihunda wrote:Great thread, I was wondering, couldn't the second transistor with collector removed just be replaced by a simple diode?

I made a quick schematic for it:


1. Take two identical random silicon transistors (for more mojo: your fav fuzz silicons
2. Tie the BASES together,
3. Cut off one COLLECTOR
4. connect a 3k to 6k resistor between the EMITTER of the collectorless devicr and the EMITTER of the other transistor.

By varying the resistance, you can dial in the gain you want...



In theory that should work fine, since the B-E junction of a transistor is effectively a diode, we already know from what's posted on other FSB forum threads that you can use a Ge transistor B-E junction as a diode, so it shouldn't matter whether it's a Si or Ge type... :thumbsup

I reckon give it a go and see if it does in fact work.... :thumbsup


Not according to R.G. Keen - he wrote this over at DIYSB: (I highlighted the part relevant for this question)

I originally posted the idea that you could lower the current gain of a silicon bipolar transistor by putting a second transistor in parallel with it, but with a disconnected collector.

The idea is that the current into the base-emitter junction of a bipolar is a single valued-function of its voltage - that is, one and only one current value results from each value of voltage, although the function may not be linear.

By putting another identical junction in parallel with the "working" one, the parallel (I use the term "dummy") junction will have the same voltage as the working one, and should in consequence, eat as much current. That, I reasoned, should leave only half the base current for the working junction, and so the overall transistor would have equivalently half the current gain.

That turned out to be correct in concept, wrong in detail. I forgot the internal rbe of the transistor junction. What a bipolar transistor base-emitter junction does that a diode junction does not is to have an equivalent nonlinear resistor arising from the transistor action. This internal resistor causes the working junction to have a higher voltage at the same current as the dummy junction; looked at the other way, the dummy junction eats more current than the working one at the same voltage.

That forces the dummy junction to eat more than half the base current, so the current gain is lowered by more than half - sometimes a lot more than half.

Brett's work showed that putting an external resistor from the dummy junction to the emitter of the working junction could make the dummy junction eat a variable amount of the input base current, effectively letting you tune the current gain of a piggybacked transistor.

And that's how that works.

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 04 Jun 2012, 20:50
by Ihunda
Thanks grrrunge and DrNomis, I love this site :)
It sounds like adding the resistor between diode cathode and transistor emitter might be close to emulating the internal rbe in the transistor.

What a bipolar transistor base-emitter junction does that a diode junction does not is to have an equivalent nonlinear resistor arising from the transistor action.


But

Brett's work showed that putting an external resistor from the dummy junction to the emitter of the working junction could make the dummy junction eat a variable amount of the input base current, effectively letting you tune the current gain of a piggybacked transistor.


So the internal transistor resistance is not big enough anyway (which results in too more gain loss without the added 2K-200K).

Anyway, only one way to found, a looper and a breadboard, I commit to do the test on Wednesday, If I don't reply back by then please harass me until I do, public crowd motivation :)

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 04 Jun 2012, 21:41
by grrrunge
I've got a wannabe-GE LPB-1 coming up on my breadboard ;) I'll let y'all know...

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 06 Jun 2012, 12:39
by Ihunda
OK, I am back so I tried 2 circuits: transistor with diode and transistor with collector less second transistor. 100K trimpot between cathode / emitters.
Transistors are 2N3904 and the diodes are 1N4148 and 1N34A germanium.
A simple booster, 10K collector and 1K emitter resistor, 220n input cap, 470K resistor between base and 4.5V. I used a 9V battery.

So my experience:
  • Both circuit lower the gain of the booster altougth it actually becomes more overdriven (less clean) as the gain goes down. I didn't expect that.
  • The 1N4148 diode alone was impossible to work with, even with a 100K trimpot, vdrop is around 0.6 for the led, 0.7 for VBE for the transistor so I figured I needed another diode in series
  • Adding the 1N34A diode (1N4148+1N34A=1V drop) actually worked better, reducing the trimpot value reduces gains and with the entire 100K, I am getting close to no effect although a 200K pot would have been better
  • The transistor version sounds much better to my ears but I have no objective way to quantify that!!
  • I don't have a germanium transistor so I can't compare but the transistor version with around 3K for the trimpot sounds mushy and a bit farty so maybe that's it :)

Although different the two diodes version is actually interesting, maybe a smaller input cap to get rid of fartiness might help.

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 10 Jul 2012, 09:14
by sinner
I'm hoping to find the time to try this today

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 10 Jul 2012, 10:17
by beedotman
sinner wrote:I'm hoping to find the time to try this today


Do You have all required ingredients? :lol:

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 10 Jul 2012, 10:30
by sinner
beedotman wrote:
sinner wrote:I'm hoping to find the time to try this today


Do You have all required ingredients? :lol:


almost... I'll need to swap two piggies for an hamster and a sticky tape ;)

BTW, we don't need that Ge diode in there with this application, do we? Si's wont be sensitive to heat, ain't that kind of function diode serves, stabilize Ge Q3?

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 10 Jul 2012, 10:58
by beedotman
sinner wrote:BTW, we don't need that Ge diode in there with this application, do we? Si's wont be sensitive to heat, ain't that kind of function diode serves, stabilize Ge Q3?


IMO Ge diode helps here with transistor bias and prevents temperature fluctuations. I think with Si trannies could be omitted, but maybe higher value resistor between B and C could be usefull here to simulate leakage and lower value resistor between ground and emitter to lower gain, but maybe piggybacking does it all without additional components... :hmmm:

Re: Piggyback Si transistors to simulate Ger ???

PostPosted: 10 Jul 2012, 18:11
by sinner
Resistor between collector and base of Q3 stage is on my new schematic. I did this after I played on my soul-bender again today, this really make huge impact on sound, and I love this ability

I did shit today, just get back home

Also, I kicked that between pots volume resistor, as only solasound back in 70 used that :mrgreen:

I may try other sooulbender mods as well, who knows.