Patrick, many thanks for your comments. See below in your text...
Cheers,
Roger

"Patrick Turner" wrote in message
...


Engineer wrote:

Hi, vacuumlanders.
I wonder if anyone can shed some light on the following...
I rescued an Electrohome "Concord IV" amplifier from a junked
console
(at the end of a neighbor's driveway - there were no tubes in it
and
the console had been in a basement for decades holding up one end
of a
workbench!)
It's a nominal 8 to10 watt PP amp using two 6V6GT tubes with an
integrated preamp for a cheap record changer (called "hi fi" but
not really!) It was clean enough top and bottom when dusted off.

After checking for no B+ shorts, I plugged in the rectifier tube
(only)

i.e. No 6V6's (or other tubes) plugged in. Only the 5Y3.

and started to bring up the B+ on a variac; got to 135 VDC
(less than 1/2 way up) and left the caps to form for a bit... so
far
so good But while poking around with a DVM to check the B+ to all
the
tube bases, I measured 135 VDC on one 6V6 plate pin and only 124 on
the other. I immediately smelled a rat, powered it down,
discharged
the caps (slowly) and checked the OPT primary. Worst fears
substantiated - one half of the primary was open.

If one 1/2 of the primary was open you would have seen about 124V at
one
6V6 anode
and no voltage on the other.

No, because there were no 6V6's plugged in (see above). The 124 VDC
measured was an 11 volt drop across the aprox. 4 Mohms (not 5.5 megs
as I first said) "open" winding (see below) at the tiny current
(around 3 uA?) drawn by my Fluke 111 DVM on that voltage range (spec
says "10 Mohms", but no current drawn specs.)

No need to lose faith in the OPT unless you prove what you observe
to be
true.

I did, unfortunately. As Monty Python would have said: "It's an
ex-OPT! It has given up it's mortal coils..." (well, 1/2 a primary.)

The halves of the OPT primary have slightly different DC resistance,
different
Ia may be flowing, so slightly different anode voltages.

Faith very lost! As we agree, no Ia flowing as no tube in there!

Never mind, we're
still ahead of the game (good chassis, power transformer and other
stuff,) It's still slated to be the stereo partner to my "Scratch
10"
small monobloc - see ABSE.

I pulled out the OPT and checked the primary again: 296 ohms from
one
plate terminal to CT, 5.5 Megohms (ugh!) on the other! OK, so we
replace it, no big deal. But with what? What was the original
plate-to-plate impedance used? Regardless, what do we really need
for two 6V6's?

That does not explain how you got supply voltage to appear at both
anodes.

Yes, it does. The 135 VDC was measured after the 296 ohm OPT 1/2
primary resistance, i.e. no voltage drop at the 3 uA, or so, DVM
current. The low 124 VDC was seen through the 4 Meg "open" winding"
(leakage) at 3 or so uA DVM current draw.

But ok, so lets say indeed the OPT is stuffed.

It is, indeed, seriously stuffed...!

On the variac, I put ran the good half of the primary up to 100 VAC
and measured 9.96 VAC on the secondary. That's a 10:1 turns ratio,
or
20:1 turns ratio Plate-to-Plate to speaker, so impedance ratio
400:1.
But what was the speaker impedance?

The load for 6V6 should be about 8k a-a, so
the secondary load would be 8,000 / 400 = 20 ohms, which seems high,
but
the a-a RL might indeed be slightly lower when matched to the very
common
use of
16 ohm speakers when the amp was made.

That's about what I thought, although the speaker DC res. measurements
(below) seem a bit low for a 16 ohm system. Anyway, now I'll be
looking for an 8K a-a OPT with a 4, 6 & 16 ohms speaker secondary.
The 3 Electrohome speakers and X/O will be scrapped (still in good
shape but not good enough quality.)

Well, as an avid pack-rat, I had pulled the 3 speakers and X/O as
well
! (12 inch woofer, DC res: 10 ohms, 8 inch mid range, DC res. 7.4
ohms and a horn tweeter, DC res. 9.3 ohms; all look like RSC (Radio
Speakers of Canada) brand.) If they are nominally 8 ohms, that
makes
the P-to-P impedance 3200 ohms, IMHO very low for PP 6V6's. But if
the speakers are 15 ohms (distinctly possible for that era), the
P-to-P impedance would be 6000 ohms, more reasonable. The separate
6V6 cathode bias resistors are 470 ohms each. Compare this to 510
ohms
in each cathode in my "Scratch 10" 6V6 PP amp. where I use a 10000
ohms P-to-P impedance OPT and get 10.4 watts max with 325 VDC on B+
and 9.6 watts with 315 VDC on B+. The power transformer is
320-0-320
(open circuit) with a capacitor input filter but I've not measured
the
B+ on a simulated load.

So, what primary impedance should the replacement OPT have? 6000,
10000 ohms, other? I don't mind re-engineering the O/P stage to
get
a genuine 10 watts in class AB1, or close enough. I've not yet
drawn
load lines on 6V6 plate curves, it may come to that, but two 6V6's
are
such an established design that it should be a no-brainer to pick
the
best OPT primary impedance.

About 8k a-a is about right, so look for an OPT to match to 4,8,16
and take your pick.
Hammond may have something, but I have no idea what their quality is
like
on PP OPTs,
because its important if you have negative feedback;
the circuit will need carefully applied stabilisation techniques to
make
sure
it does not oscillate with no load connected, or with a capacitive
load.

My experience with Hammond OPT's is good, even with fairly high values
of NFB (sorry, no dB figures to hand.) Of course, you need to
experiment with the small cap. across the NFB resistor.

Don't throw out the old OPT, give it to someone so they can wind a
nice
choke
since the core material is probably better than average.

I thought of the 1/2 primary as a choke but I've no idea of the value.
Also, the max current might be a bit low as is. Anyone who wants the
dead OPT for the core can have it for the postage. Mail me at:
"analogdino 'at' rogers 'dot' com".

Alternatively, a skilled person could re-wind such a tranny,
perhaps to a better design.

Patrick Turner.

What do people think? BTW, I'm also open to sales offers of a
suitable PP OPT.
Cheers,
Roger

No OPT offers yet! None found on eBay-Canada, either, so getting a
quote from a Hammond distributor on a 125E (10K a-a), but may now go
for an 8K a-a, as you suggest.
BTW, Patrick, please take a look at my "Scratch 10" on ABSE and let me
know if I can make it better.
Thanks again for reply.
Cheers,
Roger