2SC859 Magic in Univox Unidrive?

[seedlings]

I've been researching the Unidrive and some folks are using 2N5089s and MPSA18 for Q3.  Is there any 'magic' in the 2SC859 transistors for the Unidrive circuit, or is the magic in being able to say it has NOS original transistors?

http://www.univox.org/pics/schematics/unidrive.gif

This datasheet has an Hfe of 240:


For reference I read this thread:
http://www.diystompboxes.com/smfforum/index.php?topic=36445.0

?~magic~?

CHAD

[R.G.]

I've been researching the Unidrive and some folks are using 2N5089s and MPSA18 for Q3.  Is there any 'magic' in the 2SC859 transistors for the Unidrive circuit, or is the magic in being able to say it has NOS original transistors?
?~magic~?

No magic. You need to read some of my polemics on transistor gain.

First of all, bipolar transistor gain is neither constant for a type number (i.e., all 2N5088s are not the same), nor are they the same at different temperatures for the *same* device. They drift. In fact, a transistor's gain varies with its collector current.

That being the case, any number on a datasheet has to quote a range or be viewed as "average" or "typical" or some such fuzzy description.

It's a general principle in engineering that whatever cannot be controlled precisely must be made irrelevant. In the 60s, a lot of work was put in to designing transistor circuits where the exact gain, frequency response, etc. of a transistor did not matter much within wide bounds. Otherwise, every unit off the manufacturing line would have to be hand tuned.

While there are some special cases, in general, if it's an NPN, the exact part number doesn't necessarily mean much one way or another. This is one way of looking at Keen's Second Law: When in doubt, use a 2N5088. This works an amazing number of times.

But as you note, there is HUGE value to the seller to advertise "NOS transistors".

[seedlings]

No magic. You need to read some of my polemics on transistor gain.

That is what I suspected, and I like reading.  Link please?

I don't mind paying a little extra, but shipping increases the component value by 5x.

CHAD

[R.G.]

The polemics are here in the forum. Search on my name, bipolar gain, and datasheet, maybe hfe.

[seedlings]

The polemics are here in the forum. Search on my name, bipolar gain, and datasheet, maybe hfe.

After significant searching, I found this:

http://www.diystompboxes.com/smfforum/index.php?topic=83739.msg696323#msg696323

Very thorough. I sure hope it's the post you tell everyone to search for 

CHAD

[kaycee]

I've built a couple of these and used several 'run of the mill' NPN silicons, and all sound good in it. I settled on on a set of 2N2222, but BC108, and 3904 were fine as well. Its got a ton of output and enough gain to get into fuzz territory, I havent sourced any of the original trans to compare, but I'd be surprised if there was a discernable difference - just socket the positions and try a few its a great build!

[seedlings]

I've built a couple of these and used several 'run of the mill' NPN silicons, and all sound good in it. I settled on on a set of 2N2222, but BC108, and 3904 were fine as well. Its got a ton of output and enough gain to get into fuzz territory, I havent sourced any of the original trans to compare, but I'd be surprised if there was a discernable difference - just socket the positions and try a few its a great build!

That circuit breadboarded with 2n2222 or 2n3904 or some other 2nxxx lying around sounded... meh... like a fairly clean boost to me.  'They' say it's designed to push an already overdriven preamp, which that it does, but it doesn't add a whole lot punched into a cleanish tube amp.  Granted, each transistor probably needs attention to get it biased well, and I just plugged the components in there.... except the resistor input network.  I left that out completely.

CHAD

[seedlings]

I ordered some of the originals from smallbear because I succumbed to the hype, and I was scared.

And the guy who wants the pedal told me to.
CHAD

[kaycee]

Odd, mine were all built stock, with the input impedance on a rotary switch, and even with that dialled fully on its still a hairy beastie, let alone at just 33k?

[seedlings]

Odd, mine were all built stock, with the input impedance on a rotary switch, and even with that dialled fully on its still a hairy beastie, let alone at just 33k?

Well I found another flaky socket on the breadboard.  Now it growls nicely with any NPNs I have on hand.  Smallbear shipped the *magic* transistors, so we'll see how much mojo they add.  I was A/B-ing back and forth between a standard fuzz (Ge with Q1~80, Q2~90) and... the Unidrive is louder, with more 'bark' and believe it or not, the fuzz sounded thinner (1.49uf input, .44uF output, 10uF bypass on Q2).  Probably the 10uF output cap on the Unidrive brings up the low end.

CHAD

[kaycee]

Heres my latest build of this.





I have the input impedance on a rotary switch in the middle, acts like a pre-gain/smooth control. Using metal can 2N2222's in there, sounds great, but it will run on just about any si NPN I should think.

There is also an op-amp input buffer before the booster circuit. I tried the Millenium C bypass, but the LED doesn't completely go off when in bypass, I'll deal with that issue later, more interested in using it right now!

[seedlings]

^ Sweet!  How's it sound?  I just finished mine the other day.  I put a blend cap on the input (.00047uF and 100uF with a 500K pot).  The 100uF is overkill, but you can get a square-wave kind of buzz.  Don't like it?  Back the pot down a wee bit.  The 2SC859 trannies seem less compressed than the SS (2n3904 or 2n2222 like yours).  Sorry, no gut pic on my phone.  Take one later before the owner comes to collect.



Sound sample:

CHAD

[seedlings]

CHAD

[teleK]

I desperately want the rest of Keen's laws