Amplifier transistor matching?

Joerg · Jun 6, 2005

Hello Graham,

I'm curious why you'd send missionaries to Russia. They already have the
Orthodox Church there which is making something of a comeback after decades of
communism ( although it never ever truly died out ).

From what I know there was a school principal from Petrozavodsk who
visited a Christian school in the US. This was AFAIK back in the commie
times and he was not so much interested in the bible (I guess his gvt
didn't like that at all) but in why students were better motivated,
among other things, and what he could do about that back in Russia. That
is where the idea came up to send people over from here to teach the
students in Russia English, using the bible as a text book. So, now
there are regular missions to that city, not just from our church but
also lots of others.

Participation is voluntary for students. Usually 100-200 sign up from
what I heard. This year blew us away, a whopping 1000 signed up.

Have your missionaries previously met the Russian Mafia btw ? ;-)

No, but I guess they wouldn't be interested much in missionaries.

Regards, Joerg

http://www.analogconsultants.com

Pooh Bear · Jun 6, 2005

Jim Thompson wrote:

On Sun, 05 Jun 2005 23:58:53 +0100, Pooh Bear
rabbitsfriendsandrelations@hotmail.com> wrote:

Jim Thompson wrote:

On Sun, 05 Jun 2005 23:00:52 +0100, Pooh Bear
rabbitsfriendsandrelations@hotmail.com> wrote:

I'm guessing that budget is low in your case.

Now why would you guess that?

You mentioned Girl Scouts !

Graham

My wife has been a Girl Scout leader for 40 years ;-)

...Jim Thompson

Lol ! Nothing wrong with that Jim !

I just imagined that the Scouts don't normally have a great budget for
tech stuff. I've helped out the UK Scouts in the past btw. They're damn
fine with ropes and had my PA rig mounted on scaffolding towers in no
time at all !

Graham

Pooh Bear · Jun 6, 2005

Johnny Thunder wrote:

Matched pairs were nessary in the early days of transistor amps.

Due to poorer design of driver stages. Parts cost more back then and it wasn't
normal practice to 'overdesign'.

Manufacturing tolerances are tighter today and matching is not as big a
problem.

There's still quite a big spread in hfe. Better design of the driver stage
eliminates the problem.

Graham

Jim Thompson · Jun 6, 2005

On Tue, 07 Jun 2005 01:04:02 +0100, Pooh Bear
<rabbitsfriendsandrelations@hotmail.com> wrote:

Jim Thompson wrote:

On Sun, 05 Jun 2005 23:58:53 +0100, Pooh Bear
rabbitsfriendsandrelations@hotmail.com> wrote:

Jim Thompson wrote:

On Sun, 05 Jun 2005 23:00:52 +0100, Pooh Bear
rabbitsfriendsandrelations@hotmail.com> wrote:

I'm guessing that budget is low in your case.

Now why would you guess that?

You mentioned Girl Scouts !

Graham

My wife has been a Girl Scout leader for 40 years ;-)

...Jim Thompson

Lol ! Nothing wrong with that Jim !

I just imagined that the Scouts don't normally have a great budget for
tech stuff. I've helped out the UK Scouts in the past btw. They're damn
fine with ropes and had my PA rig mounted on scaffolding towers in no
time at all !

Graham

480)460-2350 | |
| E-mail Address at Website Fax

Joerg · Jun 6, 2005

Hello Jim,

They have a great budget when a certain benefactor makes equipment,
and vans, etc., available for free ;-)

A lot of things in our country rely on generosity and that is good. It
shows the next generation where the real values are.

Once you buy a system maybe your wife could train a few girl scouts to
become the "technology team". Then they can set it all up and learn how
stuff like a wireless system works. And what to do when it doesn't. What
was that motto? "Be prepared". So they could check for free UHF channels
at the planned location, figure out the frequency, make sure all the
gear is packed and so on.

Regards, Joerg

http://www.analogconsultants.com

Jim Thompson · Jun 7, 2005

On Tue, 07 Jun 2005 00:18:21 GMT, Joerg
<notthisjoergsch@removethispacbell.net> wrote:

Hello Jim,

They have a great budget when a certain benefactor makes equipment,
and vans, etc., available for free ;-)

A lot of things in our country rely on generosity and that is good. It
shows the next generation where the real values are.

Once you buy a system maybe your wife could train a few girl scouts to
become the "technology team". Then they can set it all up and learn how
stuff like a wireless system works. And what to do when it doesn't. What
was that motto? "Be prepared". So they could check for free UHF channels
at the planned location, figure out the frequency, make sure all the
gear is packed and so on.

Regards, Joerg

http://www.analogconsultants.com

480)460-2350 | |
| E-mail Address at Website Fax

Jim Thompson · Jun 7, 2005

On Tue, 07 Jun 2005 00:18:21 GMT, Joerg
<notthisjoergsch@removethispacbell.net> wrote:

Hello Jim,

They have a great budget when a certain benefactor makes equipment,
and vans, etc., available for free ;-)

A lot of things in our country rely on generosity and that is good. It
shows the next generation where the real values are.

Once you buy a system maybe your wife could train a few girl scouts to
become the "technology team". Then they can set it all up and learn how
stuff like a wireless system works. And what to do when it doesn't. What
was that motto? "Be prepared". So they could check for free UHF channels
at the planned location, figure out the frequency, make sure all the
gear is packed and so on.

Regards, Joerg

http://www.analogconsultants.com

480)460-2350 | |
| E-mail Address at Website Fax

Ban · Jun 7, 2005

There is simply no mechanism as how a flow of base charge can
instigate a flow of emitter/collector charge. Its applied voltage
that that cause the injection of charge.

The relevent simplified equation is:

Ic ~ Is.exp(Vbe/Vt)

Note the lack of a base current term.

I would suggest obtaining good academic semiconductor physics text
book if you want more details on the correct operation of bipolar
transistors.

Kevin Aylward

But Kevin is right, Ebers-Moll is defined via Vbe and Vbc. And beta_f or
beta_r are only derived from the alpha terms. I think it is just a point of
view, in practical circuits you use hfe and don't care much for Vbe since it
is almost constant in comparison to the current.
--
ciao Ban
Bordighera, Italy

Kevin Aylward · Jun 7, 2005

Ban wrote:

There is simply no mechanism as how a flow of base charge can
instigate a flow of emitter/collector charge. Its applied voltage
that that cause the injection of charge.

The relevent simplified equation is:

Ic ~ Is.exp(Vbe/Vt)

Note the lack of a base current term.

I would suggest obtaining good academic semiconductor physics text
book if you want more details on the correct operation of bipolar
transistors.

Kevin Aylward

But Kevin is right, Ebers-Moll is defined via Vbe and Vbc.

Its not just the Ebers-Moll, its the *device* *physics*. Its how a
transistor actually works from a phyiscal point of view. There is simply
no way that flow of base charge induces a flow of collector charge.

If one don't understand that it is Vbe that determines and injects
carries into the base that are then swept up into the collector, then
one doesn't understand transistor operation at all. Its that simple.

And beta_f
or beta_r are only derived from the alpha terms. I think it is just a
point of view,

No it isnt.

in practical circuits you use hfe and don't care much

This is simply not true. The only practical way to effectivly design
transister amplifiers is to treat the transistor as a voltage controlled
current source. Period. e.g.

di = dv.gm

Av max = Va/Vt etc.

http://www.anasoft.co.uk/EE/bipolardesign1/bipolardesign1.html

Kevin Aylward
informationEXTRACT@anasoft.co.uk
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.

Kevin Aylward · Jun 7, 2005

Pooh Bear wrote:

Jim Thompson wrote:

Kevin has a hair up his butt about voltage control, a really long
hair ;-)

Especially considering he was explaining why beta matching is
important !

Indeed. The explanation was why hfe matching mattered despite the fact
that the bipolar transistor is a voltage controlled device. To wit, its
the voltage drop across rbb' that is the killer.

Why don't you use spice and put two power transistors in parallel. Have
two separate models, with only Bf different. Do runs with and without RB
set to zero. Then put in emitter resisters. Its a good exercise.

Kevin Aylward
informationEXTRACT@anasoft.co.uk
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.

Kevin Aylward · Jun 7, 2005

Jim Thompson wrote:

On Mon, 06 Jun 2005 14:34:24 -0500, BOB URZ
"sound(remove)"@inetnebr.com> wrote:

Kevin Aylward wrote:

Bob Urz wrote:
Kevin Aylward wrote:

Pooh Bear wrote:

Ol' Duffer wrote:

In article <42A0C72A.68DD168@hotmail.com>,
rabbitsfriendsandrelations@hotmail.com says...

I've never specified matched pairs in any design ( pro-audio
btw ) in my entire life despite being responsible for some 10s
of thousands of amplifiers out there. It seems like poor
design to need matched pairs to me.

You should check a batch of power transistors on a curve tracer
sometime. The results may scare you. I routinely see a factor
of three variation in Beta within batches. Of course you can
use big, wasteful swamper resistors, or a bank of 5 unmatched
devices where 2 matched would be sufficient. Or you can let
the amps blow up and they I buy more transistors than I need and
select a good grouping from the middle of the range and fix them
so they don't blow up anymore...

These devices which I currently use for example are pre-graded by
the manufacturer. Worst case match is 2:1 in either gain grade.

http://www.profusionplc.com/cgi-bin/gex/pcatdtl?ipartno=2SC5200-O

A simple low value emitter ballast resistor overcomes the bulk of
beta mismatch anyway and I would never fail to use them. You
can't depend on paralled device temps being identical - in fact
quite the reverse - never mind thermal runaway !

Well, it might be useful to explain just why beta/hfe matching is
important, considering that that the bipolar transistor is a
voltage controlled device!

A bi polar transistor is a current controlled device actually.

Here we go, yet again...

I see that you have succumbed to the Bantam paperback popular
misconceptions Bob. A Bipolar transistor is indeed a voltage
controlled device.

So, the manuals with transistor curve tracers which state
a change a base current and plot collector current and voltage
are wrong? With bipolar beta B=Ic/Ib?

And for FET's its a change in gate voltage vs Drain/source?

Bob

For example, see
http://www.anasoft.co.uk/EE/voltagecontrolledbipolar/voltagecontrolledbipolar.html

and for a simple pictorial description of the actual physics, see:

http://www.mtmi.vu.lt/pfk/funkc_dariniai/transistor/bipolar_transistor.htm

There is simply no mechanism as how a flow of base charge can
instigate a flow of emitter/collector charge. Its applied voltage
that that cause the injection of charge.

The relevent simplified equation is:

Ic ~ Is.exp(Vbe/Vt)

Note the lack of a base current term.

I would suggest obtaining good academic semiconductor physics text
book if you want more details on the correct operation of bipolar
transistors.

Kevin Aylward
informationEXTRACT@anasoft.co.uk
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.

----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet
News==---- http://www.newsfeeds.com The #1 Newsgroup Service in the
World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via
Encryption =----

Kevin has a hair up his butt about voltage control, a really long hair
;-)

Indeed. It is they way they work. Only the illinformed consider that the
transister is current controlled.

What we have here is an enormous amount of erroneous waffle on the
bipolar transistor. Its quite amazing really. There is no point in
perpetuating this daft myth. It only leads to confusion.

Any *actual* *understanding* of the standard 3 junction pictorial
descriptions of basic transistor operation should leave people with no
doubt as to the correct operation of a bipolar transistor. Base current
don't "cause" the flow of collector/emitter current. Its Vbe that
injects charge into the emitter, irrespective of base current. End of
story.

Kevin Aylward
informationEXTRACT@anasoft.co.uk
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.

Kevin Aylward · Jun 7, 2005

BOB URZ" <"sound(remove) wrote:

Kevin Aylward wrote:

Bob Urz wrote:
Kevin Aylward wrote:

Pooh Bear wrote:

Ol' Duffer wrote:

In article <42A0C72A.68DD168@hotmail.com>,
rabbitsfriendsandrelations@hotmail.com says...

I've never specified matched pairs in any design ( pro-audio
btw ) in my entire life despite being responsible for some 10s
of thousands of amplifiers out there. It seems like poor design
to need matched pairs to me.

You should check a batch of power transistors on a curve tracer
sometime. The results may scare you. I routinely see a factor
of three variation in Beta within batches. Of course you can
use big, wasteful swamper resistors, or a bank of 5 unmatched
devices where 2 matched would be sufficient. Or you can let
the amps blow up and they I buy more transistors than I need and
select a good grouping from the middle of the range and fix them
so they don't blow up anymore...

These devices which I currently use for example are pre-graded by
the manufacturer. Worst case match is 2:1 in either gain grade.

http://www.profusionplc.com/cgi-bin/gex/pcatdtl?ipartno=2SC5200-O

A simple low value emitter ballast resistor overcomes the bulk of
beta mismatch anyway and I would never fail to use them. You can't
depend on paralled device temps being identical - in fact quite
the reverse - never mind thermal runaway !

Well, it might be useful to explain just why beta/hfe matching is
important, considering that that the bipolar transistor is a
voltage controlled device!

A bi polar transistor is a current controlled device actually.

Here we go, yet again...

I see that you have succumbed to the Bantam paperback popular
misconceptions Bob. A Bipolar transistor is indeed a voltage
controlled device.

So, the manuals with transistor curve tracers which state
a change a base current and plot collector current and voltage
are wrong? With bipolar beta B=Ic/Ib?

This is a functional relation not a causal one.

We can use a watch and produce a relation (graph) between the positions
of the planets and the positions of the watch hands. Does this mean that
the watch hands are controlling the planets motion, or that the planets
motion is controlling the watch hands?

And for FET's its a change in gate voltage vs Drain/source?

Yes.

Its voltage that makes charge move.

Kevin Aylward
informationEXTRACT@anasoft.co.uk
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.

Mr.T · Jun 7, 2005

"Jim Thompson" <thegreatone@example.com> wrote in message
news:m3u9a1dl28qjegf4qeshodsjc49mfhnekg@4ax.com...

Be prepared!
That's the Boy Scout marching song
Be prepared!
If you should meet a Girl Scout who is suitably inclined
Be prepared!

What, no attribution for Tom Lehrer?

MrT.

Anno Siegel · Jun 7, 2005

Mr.T <MrT@home> wrote in sci.electronics.design:

"Jim Thompson" <thegreatone@example.com> wrote in message
news:m3u9a1dl28qjegf4qeshodsjc49mfhnekg@4ax.com...
Be prepared!
That's the Boy Scout marching song
Be prepared!
If you should meet a Girl Scout who is suitably inclined
Be prepared!

What, no attribution for Tom Lehrer?

Perhaps because Tom Lehrer's text is different:

...
If you're looking for adventure of a new and different kind
And you come across a girl scout who is similarly inclined,
Don't be nervous, don't be flustered(?), don't be scaaaared!
Be prepared!

Anno

Jim Thompson · Jun 7, 2005

On Tue, 7 Jun 2005 18:16:58 +1000, "Mr.T" <MrT@home> wrote:

"Jim Thompson" <thegreatone@example.com> wrote in message
news:m3u9a1dl28qjegf4qeshodsjc49mfhnekg@4ax.com...
Be prepared!
That's the Boy Scout marching song
Be prepared!
If you should meet a Girl Scout who is suitably inclined
Be prepared!

What, no attribution for Tom Lehrer?

MrT.

Actually I had forgotten that it was Tom Lehrer. I should have
remembered since he was a performer at the junior prom at MIT in 1961,
along with, IIRC, my memory is fuzzy here... Jenny Richie ??

...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona Voice

480)460-2350 | |
| E-mail Address at Website Fax

Rich The Newsgroup Wacko · Jun 7, 2005

On Sun, 05 Jun 2005 10:56:30 -0500, Bob Urz wrote:

Kevin Aylward wrote:

Well, it might be useful to explain just why beta/hfe matching is
important, considering that that the bipolar transistor is a voltage
controlled device!

Tastes Great!

A bi polar transistor is a current controlled device actually.
A small base current is used to control a larger collector current.
A FET is a voltage controlled device.

Less Filling!
--
Cheers!
Rich
------
"There was a young lady of fashion
Who had oodles and oodles of passion.
To her lover she said,
As they climbed into bed,
"Here's one thing the bastards can't ration!""

John Larkin · Jun 7, 2005

Kevin has a hair up his butt about voltage control, a really long hair
;-)

Indeed. It is they way they work. Only the illinformed consider that the
transister is current controlled.

No, being 'informed' doesn't make an engineer pick a more difficult
analytical approach to quantitative circuit design; quite the
opposite. 90% of the time, a simple 0.6Vbe-plus-Beta model is the
easiest and *most reliable* way to design transistor circuits.

This is sci.electronics.design, not sci.device.physics. If I had to
analyze the physics of every component I use, I wouldn't get anything
done. What matters is how they behave.

What we have here is an enormous amount of erroneous waffle on the
bipolar transistor. Its quite amazing really. There is no point in
perpetuating this daft myth. It only leads to confusion.

Any *actual* *understanding* of the standard 3 junction pictorial
descriptions of basic transistor operation should leave people with no
doubt as to the correct operation of a bipolar transistor. Base current
don't "cause" the flow of collector/emitter current.

That's like saying that poison doesn't cause death because only heart
failure *really* causes death.

John

Joerg · Jun 7, 2005

Hello Jim,

My inclination is to a multiple-microphone version for later
expansion, since they do a lot of Flag Ceremonies with multiple
speaking parts.

The Sennheiser system does that nicely. Currently we have four but I can
see it going to 6-8. However, we'll only do that with AA battery powered
gear, no more 9V.

On our current gear the transmitters show their frequency on a little
LCD, same for the receivers. Receivers also show RF level and audio
level. The latter makes it really easy to see if a mike is active
without the mixer pot turned on.

Regards, Joerg

http://www.analogconsultants.com

Kevin Aylward · Jun 7, 2005

John Larkin wrote:

Kevin has a hair up his butt about voltage control, a really long
hair ;-)

Indeed. It is they way they work. Only the illinformed consider that
the transister is current controlled.

No, being 'informed' doesn't make an engineer pick a more difficult
analytical approach to quantitative circuit design; quite the
opposite.

The voltage controlled model is the simple model, and the correct one.

90% of the time, a simple 0.6Vbe-plus-Beta model is the
easiest and *most reliable* way to design transistor circuits.

Not at all. This is not true in the slightest.

Please explain, as discussed in this thread, from a beta perspective,
why a hfe mismatch of 2 can result in say, 10 times mismatch in current
(or more) for || devices. Why is it not just a factor of 2?

Hint: Rbb' drops a voltage to Vibe, Ic = Io.exp(Vbi/Vt)

This is sci.electronics.design, not sci.device.physics. If I had to
analyze the physics of every component I use, I wouldn't get anything
done. What matters is how they behave.

Its not about the physics is about understanding how to design
transistor circuits correctly. To do this, one must understand that the
transistor is a voltage controlled device. The why is irrelevant. Sure,
for a simply switch one might just work out the base current needed to
saturate the device, but for any design that isnt trivial, the beta
model is, essentially, useless. The first order model for gain doesn't
even have a beta term.

http://www.anasoft.co.uk/EE/bipolardesign1/bipolardesign1.html

To wit,

Av = RL/re = 40Vdc

Av max = Va/Vt.

What we have here is an enormous amount of erroneous waffle on the
bipolar transistor. Its quite amazing really. There is no point in
perpetuating this daft myth. It only leads to confusion.

Any *actual* *understanding* of the standard 3 junction pictorial
descriptions of basic transistor operation should leave people with
no doubt as to the correct operation of a bipolar transistor. Base
current don't "cause" the flow of collector/emitter current.

That's like saying that poison doesn't cause death because only heart
failure *really* causes death.

No it isn't. Base current is not equivalent to poison, its equivalent to
the body stinking after the heart failure. i.e. its just a nuisance.

This is really all getting a bit tiresome. I have explained this many
times. Base current is an effect caused by an application of voltage.
Without understanding this basic fact, it is impossible to *design* an
amplifier correctly. All one can do is piss about under the illusion
that there was a "design" performed. Period.

Kevin Aylward
informationEXTRACT@anasoft.co.uk
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.

John Larkin · Jun 7, 2005

On Tue, 07 Jun 2005 18:37:56 GMT, "Kevin Aylward"
<see_website@anasoft.co.uk> wrote:

John Larkin wrote:
Kevin has a hair up his butt about voltage control, a really long
hair ;-)

Indeed. It is they way they work. Only the illinformed consider that
the transister is current controlled.

No, being 'informed' doesn't make an engineer pick a more difficult
analytical approach to quantitative circuit design; quite the
opposite.

The voltage controlled model is the simple model, and the correct one.

90% of the time, a simple 0.6Vbe-plus-Beta model is the
easiest and *most reliable* way to design transistor circuits.

Not at all. This is not true in the slightest.

Then I owe somebody about $200 million.

Please explain, as discussed in this thread, from a beta perspective,
why a hfe mismatch of 2 can result in say, 10 times mismatch in current
(or more) for || devices. Why is it not just a factor of 2?

Because the beta model doesn't address current sharing of base
currents when transistors are paralleled. Discrete circuit designers
rarely (as in, never) hard-parallel bipolar transistors anyhow. If
they ever elected to do so, they would have to analyze the situation
properly.

Nobody can afford to design without regard to min/max beta specs. Not
even you.

Its not about the physics is about understanding how to design
transistor circuits correctly. To do this, one must understand that the
transistor is a voltage controlled device. The why is irrelevant. Sure,
for a simply switch one might just work out the base current needed to
saturate the device, but for any design that isnt trivial, the beta
model is, essentially, useless. The first order model for gain doesn't
even have a beta term.

Well, maybe yours doesn't, but that's just the equations you prefer,
to make your point.

This is really all getting a bit tiresome. I have explained this many
times. Base current is an effect caused by an application of voltage.
Without understanding this basic fact, it is impossible to *design* an
amplifier correctly. All one can do is piss about under the illusion
that there was a "design" performed. Period.

So, if A causes B, and B causes C, you conclude that A does *not*
cause C. OK, can't argue any more about that.

John

Amplifier transistor matching?

Joerg

Guest

Pooh Bear

Guest

Pooh Bear

Guest

Jim Thompson

Guest

Joerg

Guest

Jim Thompson

Guest

Jim Thompson

Guest

Ban

Guest

Kevin Aylward

Guest

Kevin Aylward

Guest

Kevin Aylward

Guest

Kevin Aylward

Guest

Mr.T

Guest

Anno Siegel

Guest

Jim Thompson

Guest

Rich The Newsgroup Wacko

Guest

John Larkin

Guest

Joerg

Guest

Kevin Aylward

Guest

John Larkin

Guest

Log in

Welcome to EDABoard.com

Sponsor