Measuring piggybacked Hfe

[lvs]

Good day everyone,

I breadboarded a FF ala Gus Smalley (name hopefully spelled right) - Q1=Ge, Q2=Si - and I like the character of the distortion but I am having a bit of trouble getting definition. I'm guessing I could improve that by piggybacking for "better" Hfe.

I am somewhat in doubt about two things.

- My cheap DMM can measure Hfe. I measured various piggybacked configurations (tran-tran-resistor, tran-diode-resistor) but my question is, does the DMM give me a result as correct as when measuring a single transistor?

- Can one say that Q1 and Q2's Hfe should have a certain ratio (like 7/10) or is it only important they be in the 70-130 ballpark?

Thanks,
Lucas

[petemoore]

The DMM's Hfe meter should tell you what the relative Hfe's are, but not account for leakage...probably gives 'good' readings for SI's since they're not known for excessive leakage.
 I didn't test any piggies, except for how they worked in the FF and Booster circuits I tried them in.
 Build the piggy test socket from an 8 pin IC socket, with wires poking through a piece of perf just big enoung for an 8 pin IC socket, so that it can be plugged into a transistor's [1 Q] socket. That way you can take out the pigs, and stick the pig combo in the DMM and test the relative gains of the various Q's and emitter to emitter resistors configurations in cct and According to meter.

[lvs]

So, one pig could be seen as a black box transistor which can be tested for Hfe and leakage, and replaced, just as a single one... your socketing suggestion looks quite helpful there.

[RedHouse]

(cross-post from thead: fuzzface q2)

(snip)... I just tried a piggy back of 2- 2N3906's (both with an hfe of 180) in my transistor tester, it got a reading of hfe=6

This seems way too low and isn't near the 1/3 Brett was speaking of.

Is piggybacking not testable in a DMM (mine's a decent one BTW) transistor test socket? if not, would one I test them to get accurate hfe readings when piggybacked?

[Joe Davisson]

On my DMM, a value closer to the higher-gain is shown, so with 300 & 600 gain transistors I got like 580. In high-power circuits the parallel transistors would be matched to balance the load, because the higher-gain transistor will try handle more power than the other.

Since the higher-gain transistor will try to conduct "before" the lower-gain transistor, the DMM probably "sees" the higher gain. At least until the power is raised, at which point the lower-gain transistor might take over more of the load. Not really sure, just rough ideas...

Or, the transistors are active devices and therefore have minds of their own =)

[brett]

Hi.

From the "fuzzface q2?" thread;

Hey Redhouse, good on you for trying a piggybacked 2N3906.

Quote:
About that piggyback thing, I just tried a piggy back of 2- 2N3906's (both with an hfe of 180) in my transistor tester, it read hfe=6 which isn't near the 1/3 you were speaking of. Is piggybacking not testable in a DMM (mine's a decent one) transistor test socket? if not, would one I test them to get accurate hfe readings when piggybacked?


You got it. DMMs measure the hFE of transistors at a single point on their hFE vs Ic (collector current) curve. Usually, that doesn't matter too much, coz the curve is kinda flat. e.g. hFE of a 2N3906 might range from 100 (at low Ic) to 200 (high Ic). But piggybacking a 2N3906 makes that range really wide, and start at zero. So the range might be from 0 to 200. Your DMM, like mine, measures hFE at low Ic (say 10uA), which makes piggybacked transistors look like they have almost no hFE. But at the operating Ic in Q2 (or Q1) in a fuzzface, which is more like 500uA, the hFE is going to be much more useful, like 50 or 100.

Sorry if that's as clear as mud. Main thing is that it works.

One of these days I should build a transistor tester (like RG's tester at geofex) and plot hFE vs Ic for a range of regular and piggybacked transistors, which will help explain what's going on.

For those interested at the electronics micro-scale, what I *think* is happening is that two similar transistors have fairly much the same hFE vs Ic curve, but the transistor without any collector current has a much higher Ib (and of course, hFE=0).  

Ib is higher because in an active transistor, most of the electrons flow through the base, into the collector and out the collector leg.  Without the collector leg, and a "full" collector, the electrons pile up in the base, combining with holes, and making for a high Ib.

The mental "model" of this that I use is that the "normal" transistor has a dynamic hFE and Ic, which is set by the circuit to operate at 1 point (e.g. 500uA and 200).  That hFE=200 means that Ib=2.5uA (because hFE=Ic/Ib).  The piggybacked junction has Ic = 0 and a fixed amount of Ib (say 2.5uA).  You can therefore imaging that hFE would be 50% of the normal value.  Increasing Ic (at fixed hFE, for a first approximation) would increase the transistor's Ib, but the piggybacked junction would stay the same (I think).  So the piggyback effect would decrease with increasing Ic, and vice versa.

I'm kinda guessing at this, but it seems to make sense.  (Heck, I'm a biologist, not an EE)

Sorry about the long, messy explanation.  I have the feeling that RG could have said this in ten words. :roll:

[petemoore]

Very interesting, and well presented concepts.
 I got super low Hfe readouts on Piggybacked transistors, but in the circuit, the very low meter readings seemed incosistant with actual circuit output.
 I can see enough to re-iterate in laymans terms how the electrons would not have an escape rout with a disconnected collector, and how at idle the Hfe readings could thoeretically be different than with a load, whatever the case there as to why is quite interesting...
 For now, [and if past FF tunings are any indication of piggybacked FF tunings...in the future] I think disregarding meter Hfe readings, and going by ear is the way to go.
 I just put some wires on a 10k pot and used it as a Variable Resistor, so I could insert the V/R between the emitters and tune away, trying different Q's and tunings..then pulled the pots wires from the resistor socket pins without disturbing the shaft position, and read the resistance of that V/R pot, the put a fixed resistor there...however, of course, you could put that V/R pot surface mount for even more tweaking fun as you use the box...
 Going by the DMM's numbers for Hfe on a regular FF is good for starters, but different types may throw the validity of numbers, because of things like frequency response etc.
 Finding the sound by ear IMO should be the final say on FF tunings especially, but all circuits should be subjected to the central scrutinizers..Your ears, and will sound different than they did yesterday.

[RedHouse]

Thanks Brett and Pete.

[lvs]

Ditto.