unkown output impedance measurement?

[Henry Kolesnik]

What the easiest ways to measure the output impedance of a push pull
tube amplifier?
Thanks
Hank

 

[Engineer]

What the easiest ways to measure the output impedance of a push pull
tube amplifier?
Thanks
Hank

Henry, do you mean the true Thevenin O/P impedance of the unit as a
"near voltage" source, i.e. very low indeed, less than, say, 0.5 ohm?
Or do you mean what is the nominal speaker impedance that it is set up
for, e.g. 4, 8 or 16 ohms?
Cheers,
Roger

 

[Henry Kolesnik]

normal speaker impedance

What the easiest ways to measure the output impedance of a push pull
tube amplifier?
Thanks
Hank

Henry, do you mean the true Thevenin O/P impedance of the unit as a
"near voltage" source, i.e. very low indeed, less than, say, 0.5 ohm?
Or do you mean what is the nominal speaker impedance that it is set up
for, e.g. 4, 8 or 16 ohms?
Cheers,
Roger

 

[Henry Kolesnik]

This is in an old military radio which I suspect is not 3.2 ohms but
somewhat higher like maybe 600. It's kind of hard to get to the
primary. I guess I should have stated that at the beginning. I kind of
recall some way of measuring the open circuit voltage from a tone and
then with a pot connected across the output, set the pot to where the
voltage is 1/2 and then measure the pot. Do I have that correct?
Hank

 

[N_Cook]

This is in an old military radio which I suspect is not 3.2 ohms but
somewhat higher like maybe 600. It's kind of hard to get to the
primary. I guess I should have stated that at the beginning. I kind of
recall some way of measuring the open circuit voltage from a tone and
then with a pot connected across the output, set the pot to where the
voltage is 1/2 and then measure the pot. Do I have that correct?
Hank

I would go for a set of watty resistors in series, tone source and a DVM on
AC volts, with amp set at low output.
If on some set of resistors of total value of R , the o/p voltage is greater
than with value R+d and R-d , where d is about R/4 then R is the output
impedance. If highest at R-d then go down, in steps, till it peaks, if R+d
is highest then go higher for peak value

 

[Charles]

What the easiest ways to measure the output impedance of a push pull tube
amplifier?

Load the amplifier with an adjustable resistor. When the loaded output
voltage drops to 1/2 of the no-load voltage, the resistor can be removed and
measured with an ohmmeter.

 

[Ian Jackson]

Load the amplifier with an adjustable resistor. When the loaded output
voltage drops to 1/2 of the no-load voltage, the resistor can be removed and
measured with an ohmmeter.

Just in case the amplifier is 'unhappy' with such a low load, it might
be better to load the output with a resistor low enough to produce a
less severe - but measurable - drop in output voltage (say by 10 or
20%). Then do a simple 'pot-down' calculation to get the output
impedance.

 

[Henry Kolesnik]

It's a BC-794 but I think it's been modified and if not something else
is kaput.

 

[Engineer]

Load the amplifier with an adjustable resistor.  When the loaded output
voltage drops to 1/2 of the no-load voltage, the resistor can be removed and
measured with an ohmmeter.

Sorry, not so. This will only give you what I originally mentioned,
i.e. equivalent generator source impedance (see Thevenin), and it will
be very low - no use in selecting speakers. You need the O/P
transformer (OPT) ratio. Make sure the receiver/amplifier is OFF. Use
a filament transformer run off a variac to energise the OPT secondary
(use the speaker terminals) - keep it low, say 2 to 4 VAC. Measure
the OPT primary and secondary voltages at a few levels. Calculate
each ratio and average them. Figure out the correct plate load for
the O/P tube (not covered here) - it will likely be in the range 4
Kohms to 8 Kohms, call it Rp. Then the correct speaker load will be
Rp/OPT ratio squared. Ex: Rp = 5000 ohms, OPT ratio found to be
36:1. Then, speaker should be 5000/(36)^2 = 5000/1296 = 3.86 ohms.
So use a 4 ohm speaker.
Cheers,
Roger

 

[Ian Jackson]

In message

Sorry, not so. This will only give you what I originally mentioned,
i.e. equivalent generator source impedance (see Thevenin),

Surely this IS the output impedance?

and it will
be very low - no use in selecting speakers.

You need the O/P
transformer (OPT) ratio. Make sure the receiver/amplifier is OFF. Use
a filament transformer run off a variac to energise the OPT secondary
(use the speaker terminals) - keep it low, say 2 to 4 VAC. Measure
the OPT primary and secondary voltages at a few levels. Calculate
each ratio and average them. Figure out the correct plate load for
the O/P tube (not covered here) - it will likely be in the range 4
Kohms to 8 Kohms, call it Rp. Then the correct speaker load will be
Rp/OPT ratio squared. Ex: Rp = 5000 ohms, OPT ratio found to be
36:1. Then, speaker should be 5000/(36)^2 = 5000/1296 = 3.86 ohms.
So use a 4 ohm speaker.
Cheers,
Roger

But what does the invariable negative feedback (from the OPT secondary
to an earlier amplifier stage) do to the output impedance?

 

[Henry Kolesnik]

Roger
50 years ago I wound and rewound many a transformer with great success.
But today I'm older, more patient and lazy and looking for an easy way
out that may not be totally accurate.
But I kind of recall reading an excellent article with a trick that I
thought was pretty ingenious but I never saved it. It may have been a
ham mag, or Howard Sams booklet or maybe Rufus Turner. Maybe someone
will remember.
Hank

Load the amplifier with an adjustable resistor. When the loaded output
voltage drops to 1/2 of the no-load voltage, the resistor can be
removed and
measured with an ohmmeter.

Sorry, not so. This will only give you what I originally mentioned,
i.e. equivalent generator source impedance (see Thevenin), and it will
be very low - no use in selecting speakers. You need the O/P
transformer (OPT) ratio. Make sure the receiver/amplifier is OFF. Use
a filament transformer run off a variac to energise the OPT secondary
(use the speaker terminals) - keep it low, say 2 to 4 VAC. Measure
the OPT primary and secondary voltages at a few levels. Calculate
each ratio and average them. Figure out the correct plate load for
the O/P tube (not covered here) - it will likely be in the range 4
Kohms to 8 Kohms, call it Rp. Then the correct speaker load will be
Rp/OPT ratio squared. Ex: Rp = 5000 ohms, OPT ratio found to be
36:1. Then, speaker should be 5000/(36)^2 = 5000/1296 = 3.86 ohms.
So use a 4 ohm speaker.
Cheers,
Roger

 

[Henry Kolesnik]

I just recalled I have an old GR 583-A Output Power Meter and if it
still works it'll tell me. It's been wasting gravity for years, Google
it and you'll see one on Sphere. Thanks to all for the comments, and
I'd like to know any tricks?
Hank

 

[Engineer]

In message
(snip)


But what does the invariable negative feedback (from the OPT secondary
to an earlier amplifier stage) do to the output impedance?

Ian, it reduces the source output impedance of the amplifier as a
voltage generator (increasing the damping factor) but does not affect
the speaker impedance to be used. The speaker impedance determines
the O/P tube plate (anode) load via the OPT ratio.
Cheers,
Roger

 

[Ian Jackson]

In message

Ian, it reduces the source output impedance of the amplifier as a
voltage generator (increasing the damping factor) but does not affect
the speaker impedance to be used. The speaker impedance determines
the O/P tube plate (anode) load via the OPT ratio.
Cheers,
Roger

But the original question was "What the easiest ways to measure the
output impedance of a push pull tube amplifier?" It didn't mention
loudspeakers.

 

[Engineer]

In message

But the original question was "What the easiest ways to measure the
output impedance of a push pull tube amplifier?" It didn't mention
loudspeakers.

Henry mentioned a miltary radio with a high output impedance (but I
don't know any with P-P output...) I have a BC348 and the OPT (from a
single 6K6 O/P tube) is designed for 'phones only, both "high" and
"low" impedance. The "low" impedance 'phone O/P is 300 ohms so using
a regular LS is not possible. I put a typical radio OPT in this radio
(I left the original 'phones OPT in place so it could be restored to
original sometime in the next 100 years!) The new OPT and the nominal
8 ohms LS used puts a reasonably correct plate load on the 6K6, about
7K. Volume is quite good - I'd guess about 2 watts output (but not
measured.)
There is no NFB in the original radio but there is a little bit now as
I routed the Rk decoupling cap to the top of the OPT secondary as a
token NFB (properly phased.) Of course, invariably there is a lot of
NFB on P-P audio amplifiers and, as noted above, it reduces the true
output impedance of the amplifier as a voltage source, but it does NOT
change the correct LS impedance as the latter defines the O/P tubes'
plate-to-plate load.
I don't know why Henry first asked about a "PP amplifier then "morphed
to a "military radio" but it doesn't matter - it's all the same
physics!
Cheers,
Roger

 

[Henry Kolesnik]

Neil Sutcliffe was kind enough to send me a manual on his GR 783A which
is a more recent version of my GR 583A Output Impedance Meter. The o/p
xfrmr on my Hmmarlund Super Pro 210X Type O is 10 ohms and I looked
inside and saw no evidence that it was changed. I confirmed that the
meter works by the same technique on my Ten Tec SP 325 which is 600 ohms
and it measured 600 ohms. The GR 583A is a neat piece of gear, easy as
pie to use.

 

[Henry Kolesnik]

I'll see you on the third half of the show

Usual reason would be to drive several small speakers or to do so over
a
long run - like 100v line. Where a matching transformer is situated
with
the speaker.

--
*I believe five out of four people have trouble with fractions. *

Dave Plowman [email protected] London SW
To e-mail, change noise into sound.