inverted transitor beta question

[jmasciswannabe]

So I am having fun building one of Deadendfx's latest offerings which happens to feature a pair of transistors that are inverted, or I have seen it call reverse beta, basically the emitter is going where the collector goes and vice versa. Anyway, I have been tossing all kinds of transistors in there and noticed that I have some bc550c (fairchild marking) that is registering 130mv and others (not sure what brand) around 2v at the collector. Each transistor is in same general gain range on the dca55. I was curious and have messed with some others even though I knew their gain range was too high, MPSA18 from unknown and some from small bear. Unknown again doing the 130ma and the smallbear 3v or so. I've also got 2n5089 that measure in the low mv range. Could the ones giving the low ma readings compared to volts be some type of rebranded fakes? What would cause this?

http://www.muzique.com/lab/reverse.htm

The bit of circuit I am talking about in particular:



As always, I appreciate the time and discussion!

[Rob Strand]

When you use a transistor in reverse it acts just like a normal transistor except the gain is much reduced.
A ball-park for the reverse gain is about 4 but it could easily vary from 1 to 10 or more.

The reverse gain is much different to the normal forward gain.    I'd go so far is to say there is little correlation between
the two values.    When you test the gain with a hFE meter you need to put the transistor in with the C and E reversed
just like to do in the circuit.   That should give you a low gain measurement  corresponding to the reverse beta.
If you tester tries to be smart and work out the pins then it will likely produce the forward gain,  the tester would need
to have an override feature to get around that.  (I normally measure reverse beta with my own test circuit from a few resistors.)

Like forward gain the reverse gain is a function of the collector current.  That means the gain produced by the tester
might not correspond to the gain in the actual circuit.   You can get the same problems with forward gain;  look up a datasheet
and you will see a graph of hFE varying with collector current.  The reverse gain is rarely if ever given in the datasheet you
have to measure it.

The voltages on the schematic amount to reverse gains of about 4 to 5.   The circuit voltages are sensitive to the gain
so if your transistors have reverse gains much different to 4 you will need to tweak the collector resistors to get
the correct biasing, or, play around with different transistors.   There's no easy way out.

Could the ones giving the low ma readings compared to volts be some type of rebranded fakes? What would cause this?

It's certainly possible.  However, the reverse beta isn't spec'd by the manufacturer so it's hard to know from that.  Small transistors are so similar it could take hundreds of (normal forward mode) measurements before you could tell you have out of spec/re-branded transistors, some measurements are not easy to do accurately.  You could be lucky though if one particular parameter is off.

[jmasciswannabe]

Thanks a lot Rob!

I usually test transistors with known good ones before adding them in the bag so they at least performed and sounded the same in some fuzz circuit or another, but the drastic voltage difference threw me for a loop. When I said the gain readings were the same, I meant the forward gain for both brands of transistor. I'm essentially trying to plug transistors into there and get within the indicated voltage range, Dino was kind enough to narrow that down to bc550c 520-560hfe. I had great luck finding bc549c in that range but to my dismay ended up with the very low ma at the collector, the base and emitter were pretty close. So you are saying it might just be that instead of reverse gaining 4-5 it could be reverse gaining a whole lot more and that and that toying with the collector resistor I might be able to dial it into the 4-5 range?

[Rob Strand]

When I said the gain readings were the same, I meant the forward gain for both brands of transistor.

Yep, that's what I understood.   To know something is off you can only compare against the datasheet values.    However,  there is a trend these days for gains of given transistors to bunch-up more than they did in the past.    In the old days I remember low gain 2N3904's were common but in the last 20 years or so the gains have become progressively higher and more consistent.    They all agree with the datasheets but if I saw a new low gain 2N3904 I would be suspicious it was faulty whereas years ago I wouldn't batter an eyelid.

So you are saying it might just be that instead of reverse gaining 4-5 it could be reverse gaining a whole lot more and that and that toying with the collector resistor I might be able to dial it into the 4-5 range?

Yes, exactly that.    You aren't dialing in the gain you are dialing in the collector voltage. No different to playing with transistors used in the forward mode.   If the collector voltage is off you have to change something in the circuit to compensate and that's often the collector resistor.   At the end of the day the circuit is reflecting what the reverse gain is, in fact what the reverse gain is under the circuit conditions.

[jmasciswannabe]

You aren't dialing in the gain you are dialing in the collector voltage. No different to playing with transistors used in the forward mode.   If the collector voltage is off you have to change something in the circuit to compensate and that's often the collector resistor.   At the end of the day the circuit is reflecting what the reverse gain is, in fact what the reverse gain is under the circuit conditions.

This is tripping me up a little. By adjusting the collector voltage via the mentioned resistor, doesn't that end up dialing in the gain or am I missing something?

[Rob Strand]

doesn't that end up dialing in the gain or am I missing something?

It's a bit confusing because there's different gains.

The transistor gain is more or less fixed.  It varies a bit with current.  That's a current gain I_collector / I_base.

The signal gain (voltage gain) depends on the circuit values and the bias point.
(Even if the transistor gain didn't vary with current that would still be the case.)

[antonis]

Just beware of R23/R26 values 'cause by making a value mistake (e.g. wrong multiplier band reading) you might destroy the transistors..

[jmasciswannabe]

Gotcha, well kind of. Makes sense about the different gains. I am using a minimum resistor so no worries about frying the transistor. I've got 200 of them just in case, LOL. I was hoping that I could zone in on the correct transistor by identifying the hfe of a suitable match and then finding other matching hfes within a batch and hoping they would behave the same. I have found this is not the case, at least with the bc549c I have. Perhaps 550c has something special going on (although I read the only difference is a higher maximum voltage.) If the transistors used in those spots dent replicate the suggested voltages then the you get a misbiased sound. I did try and dial in the collector voltage but no luck. Maybe the wiser approach is a redesign of that section.

[Rob Strand]

I was hoping that I could zone in on the correct transistor by identifying the hfe of a suitable match and then finding other matching hfes within a batch and hoping they would behave the same. I have found this is not the case, at least with the bc549c I have. Perhaps 550c has something special going on (although I read the only difference is a higher maximum voltage.) If the transistors used in those spots dent replicate the suggested voltages then the you get a misbiased sound. I did try and dial in the collector voltage but no luck. Maybe the wiser approach is a redesign of that section.

What often happens with batches of transistors is the transistors have similar characteristics.   So you might find you can't find a magic match and need to tweak the collector resistors anyway.  (Not sure why the collector voltage adjustment didn't work.)

FWIW,  I have a BC237 sitting around my desk.   The forward gain is around hFE=360.   It's a similar transistor to the BC54x stuff.    I wired it up according to the first stage on your schematic, the one with the 1k emitter resistor.  To get 1.7V on the collector I had to tweak the base resistor to 593k in place of the 470k.    From that we can see my transistor isn't far off the ones recommended on that schematic.   My base current was decreased by a factor of roughly  470k/593k = 0.79, so only 20% off.    The equivalent tweak on the collector resistor would be to decrease it by 0.79 ie. the 100k goes to 79k.  (I tweaked the base resistor since I had those resistors sitting around.)

The main point is, even when it's fairly close you still end-up having to tweak something.    Most DIY circuits which have similar issues simply replace the collector resistors with trim pots.

One thing to check is the  pinout of your transistors.  Make sure you have E and C reversed.  The BC54x and BC55x transistors have C and E flipped compared to the 2N3904 pinouts.    If your E and C aren't reversed you will be getting the large forward gain and the circuit won't work.