Amplifier transistor matching?

[Ron]

Usually it's current gain that's matched. 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.

I think the original post is misleading you a bit. The 2245 amp
will run just fine with non-matched output transistors, without
'sever distortion problems'. However, to achieve the lowest possible
distortion and stability, matched pairs are advisable. This is true
for *any* symmetrical audio amplifier design.

Some circuits almost don't care. It depends a lot on the driver stage.

True. Especially for low NFB, low gain designs. Still the better
the complimentary paiss match, the better will the amplifier be.

Incidentally I can't really see how a failed output device can be
responsible for severe distortion. Normally it's a works or not
situation with output devices.

Well, when one side of a pair works while the other doesn't, you do
get signal out, but it's seriously (50% and up) distorted. Not just
'poor fidelity' but 'the amp is definitely bad' kind of sound.

I recently had a similar problem with an 2270 output block. I
replaced the bad output device with a matched (measured out of
about a dozen) transistor. The receiver sounds great, but the
repaird channel now measures better than the originally good
one.

-- Ron

 

[Joerg]

Hello Mike,

Try the new G2 series Sennheisers. They use AA batteries and have 9hr
battery life w/alkalines. With the new high capacity NiMh AA's out there
you should get close to that or even longer life. They are not digital, but
they do sound great. ...

Yes, we'll certainly look into those. Especially if they are compatible
with our existing Sennheiser UHF gear.

... Why do you need true digital?

We really don't. But spread spectrum is so much lower maintenance. Most
RF mikes including the Sennheiser SK series operate as UHF secondary
user. That has several disadvantages. One is that UHF channels get
kicked around a lot these days with new temporary DTV channels being
assigned and so on. Another issue is noise. When a large bank of
fluorescents gets turned on this is audible on pretty much any FM based
system. Not so with spread spectrum.

Other problems could have been avoided with clever engineering. Examples
are the pop noise when a mike is turned off, and sometimes even when it
is muted. The answer I got regarding the mute pop was like "pastors use
that switch too often and it wears out". Well, then why is there a mute
switch? And why isn't it properly debounced? After all, these systems
are also advertised for use in churches and smaller congregations do not
have an audio or mixer operator. So I'd expect proper mute switching.

On rare occasions we had mikes "forget" their programmed frequency. In
the middle of a worship service there is not much you can do about that
when it happens. At least not fast enough.

Regards, Joerg

 

[Michael A. Terrell]

Can I interject and ask some advice?

My wife has lots of Girl Scout speaking presentations, but she's a
walker... walks away from the podium and the microphone.

Did it again last week with me frantically waving, "Go back to the
microphone."

These presentations are usually in not-very-well or anciently equipped
locations... last week was in an old Catholic Church Parish Center.

What should I buy in the way of a wireless microphone, with facility
to plug the receiver into almost any PA equipment I might encounter?

Thanks!

Jim Thompson

MCM http://www.mcminone.com has a number of wireless mic systems,
starting at $39 and going up to $499.

 

[Winfield Hill]

Jim Thompson wrote...

My wife has lots of Girl Scout speaking presentations, but she's
a walker... walks away from the podium and the microphone.

Nobody should be bound to the podium. Solve the problem.

 

[Midlant]

He relies on Christ. If He brings you to it, He will see you through it.
John

 

[Pooh Bear]

I think the original post is misleading you a bit. The 2245 amp
will run just fine with non-matched output transistors, without
'sever distortion problems'. However, to achieve the lowest possible
distortion and stability, matched pairs are advisable.

For a design of that era I would tend to agree.

This is true for *any* symmetrical audio amplifier design.

I have to differ with you here. Poor beta match is only an issue when the
output devices load a moderate impedance driver stage. This gives rise to
unequal open loop gain for positive and negative half cycles. NFB will only
*reduce* this assymetry.

The effect can be eliminated by use of a *low* impedance driver stage or a
high gain pre-driver stage, as I have always used. Remember the Quad
'triples' in the 303 ? My amp design in the Vision that spawned the D series
amps actually used quadruples - but the big devices were actually off under
quiescent conditions - so for low level it was essentially similar to a
'triples' design at low power.

High ac gain in the driver - pre-driver stages also improves load related
stability as the load impedance doesn't reflect back in any significant way
to the previous voltage gain stage.

Note *ac* gain. I was briefly misled in the development of that amp by using
DC gain requirements and it was unstable with some loads. Then I ran a gain
/ phase plot and the result became obvious.

True. Especially for low NFB, low gain designs. Still the better
the complimentary paiss match, the better will the amplifier be.

See above... I do tend to use quite a bit of local feedback and close the
loop gently as it happens. 'Low NFB' as commonly used by the audiotwats as a
generalisation is misleading. There'll be NFB somewhere ! It doesn't just
vanish.

Well, when one side of a pair works while the other doesn't, you do
get signal out, but it's seriously (50% and up) distorted. Not just
'poor fidelity' but 'the amp is definitely bad' kind of sound.

As I mentioned elsewhere - when an output device fails, I've rarely known it
to be anything other than a short. That normally takes out the other output
device(s) and fuses etc and there's no sound.

The exception to that rule is the Hitachi style lateral mosfets. They
usually fail open. You see the effect when a big Mosfet amp amp doesn't clip
symmetrically any more !

I recently had a similar problem with an 2270 output block. I
replaced the bad output device with a matched (measured out of
about a dozen) transistor. The receiver sounds great, but the
repaird channel now measures better than the originally good
one.

That'll teach you !

Graham

 

[Jim Thompson]

You mentioned Girl Scouts !

Graham

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

...Jim Thompson

 

[Kevin Aylward]

For a design of that era I would tend to agree.


I have to differ with you here. Poor beta match is only an issue when
the output devices load a moderate impedance driver stage. This gives
rise to unequal open loop gain for positive and negative half cycles.
NFB will only *reduce* this assymetry.

The effect can be eliminated by use of a *low* impedance driver stage
or a high gain pre-driver stage, as I have always used. Remember the
Quad 'triples' in the 303 ? My amp design in the Vision that spawned
the D series amps actually used quadruples - but the big devices were
actually off under quiescent conditions - so for low level it was
essentially similar to a 'triples' design at low power.

High ac gain in the driver - pre-driver stages also improves load
related stability as the load impedance doesn't reflect back in any
significant way to the previous voltage gain stage.

Note *ac* gain. I was briefly misled in the development of that amp
by using DC gain requirements and it was unstable with some loads.
Then I ran a gain / phase plot and the result became obvious.

Could you explain what you actually mean here by "misled"? DC gain has
never had any relevance to amplifier stability. No one is going to
design an amp feedback loop based on DC conditions. Its only the phase
at the unity gain frequency that matters.

Kevin Aylward
[email protected]
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]

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.

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
[email protected]
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.

 

[Johnny Thunder]

For most gigs, the price of a battery is negligible compared to the show
going 'down.' That said, the new Shures seem to do that long on a 9v. The
really nice ones have battery meters on the actual receiver, so that you can
monitor the battery condition remotely. I've seen a bunch of these lately.
They've performed flawlessly IME, but the included mic is a little large.
There's an ultra small mic option which is less noticable, but the (big)
stock mic sounds better than most lav's I've worked with....
<http://www.musiciansfriend.com/srs7/search/detail/base_pid/270296?src=3WBZ4
DS>

jak

Matched pairs were nessary in the early days of transistor amps.
Manufacturing tolerances are tighter today and matching is not as big a
problem.

JAM

 

[Bob Urz]

Midlant wrote:




I've heard US contributors to the audio groups mention 'Caig' as good for
switches - maybe pots too.

Caig stuff works great. Cramolin D5 or pro gold.
Its not cheap, but it will work better than anything else
i have tried.

One problem in the old Marantz is the tape monitor switches.
Its common to have them get oxidized internally and cause a channel
to cut out. These switches are sealed. My patented way to revive them
is to carefully burn a hole in the back with a thin solder pencil.
You have to be careful to work the switch in and out when you do
this so the plastic does not jam up the switch internally. You then
use some D5 or such and work the switch in and our for a minute or
so. Then seal it back up.

Bob

 

[jakdedert]

Joerg wrote:

We really don't. But spread spectrum is so much lower maintenance.
Most RF mikes including the Sennheiser SK series operate as UHF
secondary user. That has several disadvantages. One is that UHF
channels get kicked around a lot these days with new temporary DTV
channels being assigned and so on. Another issue is noise. When a
large bank of fluorescents gets turned on this is audible on pretty
much any FM based system. Not so with spread spectrum.

Other problems could have been avoided with clever engineering.
Examples are the pop noise when a mike is turned off, and sometimes
even when it is muted. The answer I got regarding the mute pop was
like "pastors use that switch too often and it wears out". Well, then
why is there a mute switch? And why isn't it properly debounced?
After all, these systems are also advertised for use in churches and
smaller congregations do not have an audio or mixer operator. So I'd
expect proper mute switching.

You'll like the Shures mentioned before. No such problems in my experience.
No pops, no noise when the receiver is operating without a transmitter,
extremely wide choice of frequencies...auto frequency search. If there is
an operator controlling the sound system, there's a lockout function which
takes the performer out of the loop in terms of unintended switch punches.
You must open the battery panel, press and hold *two* switches to re-enable
the user controls. Sure beats gaffing over the buttons...which I always do
when handing talent a mic not so equipped.

On rare occasions we had mikes "forget" their programmed frequency. In
the middle of a worship service there is not much you can do about
that when it happens. At least not fast enough.

Not seen any of that, either. In any case, the problem can be minimized by
having a wired backup. Again, an operator is desirable so that the backup
doesn't need to be left 'open.' If none is available, a switched
mic--turned off, but left potted up--is a fair alternative. The talent
needs to be aware of the plan, of course.

jak

 

[Joerg]

Hello Jak,

You'll like the Shures mentioned before. No such problems in my experience.
No pops, no noise when the receiver is operating without a transmitter,
extremely wide choice of frequencies...auto frequency search. ...

That would be an alternative. This time though we'd insist on a serious
test drive before a purchase. The additional Shure system we have is
more than 10 years old but the noise tolerance (lamp switches etc.) is
not stellar, to put it mildly. The Sennheisers are a lot better.

Not seen any of that, either. In any case, the problem can be minimized by
having a wired backup. Again, an operator is desirable so that the backup
doesn't need to be left 'open.' If none is available, a switched
mic--turned off, but left potted up--is a fair alternative. The talent
needs to be aware of the plan, of course.

Frequency memory losses were rare, maybe a couple times a year. Yes, we
do have wired backups and the choir people are well trained. But some
performances require lots of movement and the cables would become a hazard.

Regards, Joerg

 

[Joerg]

Hello John,

He relies on Christ. If He brings you to it, He will see you through it.

Yes, and that is part of many of the sermons. We have a group of
missionaries going to Russia tomorrow morning. They don't have their
tickets yet, none speaks Russian.... but they trust it's going to work
out ok.

Regards, Joerg

 

[Joerg]

Hello Winfield,

Nobody should be bound to the podium. Solve the problem.

When a friend was speaking I saw the field strength meter on the
wireless receiver for his mike go from near full scale to zilch. Oh no,
not now... The battery had given out. He has such a thunderous voice
that I could not notice a difference and I was way in the back, behind
the audience.

So it may just be a matter of training ;-)

Regards, Joerg

 

[BOB URZ]

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

 

[Jim Thompson]

Hello Winfield,


When a friend was speaking I saw the field strength meter on the
wireless receiver for his mike go from near full scale to zilch. Oh no,
not now... The battery had given out. He has such a thunderous voice
that I could not notice a difference and I was way in the back, behind
the audience.

So it may just be a matter of training ;-)

Regards, Joerg

Joerg, Can you recommend a Sennheiser model, with lapel microphone?
Thanks!

...Jim Thompson

 

[Jim Thompson]

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

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

...Jim Thompson

 

[Joerg]

Hello Jim,

Joerg, Can you recommend a Sennheiser model, with lapel microphone?

My experience is limited to their 9V wireless gear which I would not
recommend because of limited battery life and premature battery
failures. There is a new "SK 100 G2" unit which looks very interesting
and that is the one I am going to check out once we upgrade:

http://www.sennheiser.com/sennheiser/icm_eng.nsf/root/21411

The corresponding receivers are typically 19" rack mount and I believe
they do sell kits that contain transmitter, mike, receiver, diversity
antennas, power supply and the cables. The Sennheiser web site is not
very good so you may have to call them to see what the best deal would
be for the whole set. Believe it or not, their "search function" didn't
find their own SK100 mike transmitters, Google did...

Since this is professional gear the output is usually differential, for
connection to a large mixer panel.

To be honest, this professional gear takes more time to set up in the
field when compared to cheaper systems like the Radio Shack I mentioned
before.

My only gripe besides the 9V with our existing Sennheisers is the pop
noise upon turn-off and mute. But if your wife's scout presentations are
typically uninterrupted that won't matter. Also, these systems can be
upgraded to umpteen mikes if that was ever needed.

Regards, Joerg

 

[Pooh Bear]

Could you explain what you actually mean here by "misled"? DC gain has
never had any relevance to amplifier stability. No one is going to
design an amp feedback loop based on DC conditions. Its only the phase
at the unity gain frequency that matters.

I misled myself I'm afraid. It's that long ago that I forget the exact
details.

Graham