Page 3 of 4

Re: Transistor Vbe Matching Using A Commutating Chopper-Based Circuit

Posted: Sun Dec 22, 2019 8:38 am
by JR.
If you add a 50k or so resistor from the junction of the two emitters to ground that would prevent potential reversed junctions from zenering when switched for the wrong polarity.

JR

Re: Transistor Vbe Matching Using A Commutating Chopper-Based Circuit

Posted: Sun Dec 22, 2019 8:56 am
by mediatechnology
Not sure if we're talking about the device backwards or the NPN/PNP switch being wrong.
Under normal circumstances (device inserted correctly) wouldn't that add a voltage divider?

Re: Transistor Vbe Matching Using A Commutating Chopper-Based Circuit

Posted: Sun Dec 22, 2019 9:42 am
by JR.
mediatechnology wrote: Sun Dec 22, 2019 8:56 am Not sure if we're talking about the device backwards or the NPN/PNP switch being wrong.
Under normal circumstances (device inserted correctly) wouldn't that add a voltage divider?
Yes to mitigate polarity switch in incorrect mode.

In the correct forward direction the R to ground only has roughly a half volt across it so not much current loss. Reversed it divides down the voltage to <4V and below zener threshold. We do not want low noise devices to ever zener.

You could also switch in clamp diodes across the b-e junctions but that is far more complexity than just adding one resistor.

JR

Re: Transistor Vbe Matching Using A Commutating Chopper-Based Circuit

Posted: Sun Dec 22, 2019 10:07 am
by mediatechnology
JR. wrote: Sun Dec 22, 2019 9:42 am We do not want low noise devices to ever zener.
JR
I agree with that completely.

With the ZTX851, a car horn and florescent ballast driver, I'm really curious to see if it would not be degraded.
This is one bad-ass rugged high-current transistor for its small size.
Destroying one, by Zenering the B-E junction may become a personal challenge. :lol:
I had great fun blowing up 1510s and got paid for it.

If someone wanted to use a TL071 in place of the OP07 and use it as a follower a clamp would be a good idea for a number of reasons the second of which would be avoiding latch-up of the TL071.

Re: Transistor Vbe Matching Using A Commutating Chopper-Based Circuit

Posted: Sun Dec 22, 2019 11:03 am
by JR.
mediatechnology wrote: Sun Dec 22, 2019 10:07 am
JR. wrote: Sun Dec 22, 2019 9:42 am We do not want low noise devices to ever zener.
JR
I agree with that completely.

With the ZTX851, a car horn and florescent ballast driver, I'm really curious to see if it would not be degraded.
This is one bad-ass rugged high-current transistor for its small size.
Destroying one, by Zenering the B-E junction may become a personal challenge. :lol:
Reportedly zenered junctions can be rehabilitated by self-annealing (operating with high current in forward direction).

This mechanism could explain reports that some circuits sound better after burning in for hours/days (or not). :lol:

JR

I had great fun blowing up 1510s and got paid for it.

If someone wanted to use a TL071 in place of the OP07 and use it as a follower a clamp would be a good idea for a number of reasons the second of which would be avoiding latch-up of the TL071.

Re: Transistor Vbe Matching Using A Commutating Chopper-Based Circuit

Posted: Sun Dec 22, 2019 11:52 am
by mediatechnology
I may see if I can find a value just below 144K in place of the 150K with a second resistor to ground that keeps Rev-Vbe in bounds and also pulls out just enough current to make Ie closer to 100 µA.

Ie isn't that critical though but having that second R would keep a TL071 within its' CM range with no DUT in socket or one of the wrong polarity for the switch setting.

I think having an R limit the voltage at that node is better than using a regulated Ve <|5V|. It's fun sweeping Vcc and Vee over a 3:1 range and not watching delta-Vbe change. (That was before Vc was limited to |600mV|.)

Thanks for the suggestion.

Since the ZTX851 are so quiet any bets if they make a good noise source when the B-E junction is Zenered? :lol:
Would be fun if it did both well: Quiet and noisy.

Re: Transistor Vbe Matching Using A Commutating Chopper-Based Circuit

Posted: Sun Dec 22, 2019 3:22 pm
by JR.
mediatechnology wrote: Sun Dec 22, 2019 11:52 am
Thanks for the suggestion.
glad I could help
Since the ZTX851 are so quiet any bets if they make a good noise source when the B-E junction is Zenered? :lol:
Would be fun if it did both well: Quiet and noisy.
coincidentally over on GDIY some kid is complaining about an intentional zener connected transistor (in a synth) not sounding good enough...

Have fun... I don't know if low Rbb impacts zener behavior that is some kind of junction avalanche breakdown mode. Good (clean) process could matter, or not. :lol:

JR

Re: Transistor Vbe Matching Using A Commutating Chopper-Based Circuit

Posted: Mon Dec 23, 2019 1:23 pm
by mediatechnology
I posted this thread at electro-music.com and forum member elektrouwe shared a link for a similar Vbe tester he did a number of years ago and posted at muffwiggler: https://www.muffwiggler.com/forum/viewt ... p?t=153845

elektrouwe's YATMA (Yet Another Transistor MAtcher) can scan up to 8 devices at a time. Because his tester operates in "2 terminal" mode it can simultaneously match NPN and PNP devices.

Image
elektrouwe's YATMA (Yet Another Transistor MAtcher) https://www.muffwiggler.com/forum/viewt ... p?t=153845

elektrouwe's post is very informative and I encourage our readers to check it out at muffwiggler.

Just noticed that elektrouwe and I drew our schematics on the very same day, December 22, exactly 4 years apart.

Re: Transistor Vbe Matching Using A Commutating Chopper-Based Circuit

Posted: Sun Dec 29, 2019 10:00 am
by terkio
YATMA is an approach that takes care nicely of the RdsON mismatch of the switches.
The measurement is done directly at the DUT, so the voltage drop, RdsON * Ie, does not come in the measurement.

It comes in indirectly though, because the current to the DUT depends a little bit of RdsON.
Applied to a pair of BJTs like ZTX851, I figured the measurement error is 6μV. ( Assuming Diff RdsON 3.5 ohm, Vee -15V ). This error doesn't depend of Ie the chosen current to the transistors.
This result is based on Re = 0.026 / Ie.
Re the dynamic resistor value of the emitter.
26 Ohm at 1mA and room temperature.

Furthermore, in case this 6μV error is not good enough, it can be reduced to oblivion using a constant Ie current. Using a CCS or bootstrapping.

Re: Transistor Vbe Matching Using A Commutating Chopper-Based Circuit

Posted: Wed Jul 15, 2020 12:55 pm
by terkio
DG451 is better than DG431 with
Ron 5 ohm and Ron match 0.5 ohm.