[GRe]

"Alex" wrote in message
...
Hello,

You know those Japanese tuner-amplifiers of the 60's which have dual AM
receivers, one possibly with FM and a stereo amplifier, typically powered
by 6BM8 tubes...

I got one of those, Pioneer SM-B200A. It is not working -- the tubes get
hot and "burning". Most likely dead electrolytic and/or leaky interstage
paper caps which bring positive bias to grids. I need to fix the beast
before getting rid of it.

I have no schematic, but tracing the wiring, realised that 6BM8s have
fixed bias originating from some selenium rectifier. Cathodes of 6BM8s are
grounded. More over, there are flinsy open (to dust) carbon trimpots to
"balance" (!) bias of the output tubes. At the same time there is no way
of knowing values of the currents in the tubes. There are no even small
current sensing (measuring) resistors in the cathodes. That design does
not make much sense.

I would get rid of the fixed bias in favour of automatic bias. Automatic
bias will be more robust and forgiving to tubes mismatch. It will not care
if this selenium rectifier dies altogether. I would throw resistors from
each cathode to ground, plus another between the cathodes to be able to
check unbalance.

Now my question is, from your experience, what should these resistor
values be, to get a benign class AB? I would guess 300R from each cathode
to ground and 100R between the cathodes?

Regards,
Alex

Adding the 100R will only decrease the voltage imbalance between cathodes
but will increase anode current imbalance. And that's not what you want.

Datasheet Rk for ECL82/PCL82/UCL82 in AB-PP:
B+ (V) 100 170 200 230 250
Rk (ohm) 140 125 170 200 220
Rk's above are spec'd in bean counter mode, Patrick would say.
So, in single cathode resistor config the values must be doubled.

More details:
http://frank.pocnet.net/sheets/030/e/ECL82.pdf
http://frank.pocnet.net/sheets/030/p/PCL82.pdf
http://frank.pocnet.net/sheets/030/u/UCL82.pdf

Equivalents:
ECL82 = 6BM8 (Htr. 6.3V)
PCL82 = 16A8 (Htr. 0.3A)
UCL82 = 50BM8 (Htr. 0.1A)
Finally ECL82 = PCL82 = UCL82 except for heater V/I and MINOR
differences which are not relevant in the context of your question.

Regards,
Gio

[Big Bad Bob]

On 12/27/10 13:30, Alex so witilly quipped:
I have no schematic, but tracing the wiring, realised that 6BM8s have fixed
bias originating from some selenium rectifier. Cathodes of 6BM8s are
grounded. More over, there are flinsy open (to dust) carbon trimpots to
"balance" (!) bias of the output tubes. At the same time there is no way of
knowing values of the currents in the tubes. There are no even small current
sensing (measuring) resistors in the cathodes. That design does not make
much sense.

I suggest pulling the tubes (except B+ rectifiers, if any), and checking
the selenium rectifier output. As I recall they're not highly reliable,
but I could be wrong. I suppose you could cheat and sub in a modern
silicon rectifier. A bad bias rectifier (or bias filter cap) would
explain "the symptoms". [Don't forget to discharge power supply caps
after you power down since no tubes = no discharge current].

[GRe]

"Alex" wrote in message
...

"GRe" wrote in message
...
I would get rid of the fixed bias in favour of automatic bias. Automatic
bias will be more robust and forgiving to tubes mismatch. It will not
care
if this selenium rectifier dies altogether. I would throw resistors from
each cathode to ground, plus another between the cathodes to be able to
check unbalance.

Now my question is, from your experience, what should these resistor
values be, to get a benign class AB? I would guess 300R from each
cathode
to ground and 100R between the cathodes?

Regards,
Alex

Adding the 100R will only decrease the voltage imbalance between cathodes
but will increase anode current imbalance. And that's not what you want.

Thanks.

Will it help to use independent resistors from each cathode to ground and
then a large non-polarised electrolytic between the cathodes? I would
think this will give the best DC current balance, highest gain while in
class A (low volume) -- hence lowest output impedance, but might result in
higher distortion in transition from A to B (at higher volume), but who
cares?

I was only commenting on the 100R cathode to cathode resistor. Forget about
the C to C resistor and the common cathode resistor configuration. For best
anode current balance the single cathode resistors configuration will help.
The value of them depend on your B+, see the datasheets and the table below.

Regards,
Gio



Datasheet Rk for ECL82/PCL82/UCL82 in AB-PP:
B+ (V) 100 170 200 230 250
Rk (ohm) 140 125 170 200 220
Rk's above are spec'd in bean counter mode, Patrick would say.
So, in single cathode resistor config the values must be doubled.

More details:
http://frank.pocnet.net/sheets/030/e/ECL82.pdf
http://frank.pocnet.net/sheets/030/p/PCL82.pdf
http://frank.pocnet.net/sheets/030/u/UCL82.pdf

Equivalents:
ECL82 = 6BM8 (Htr. 6.3V)
PCL82 = 16A8 (Htr. 0.3A)
UCL82 = 50BM8 (Htr. 0.1A)
Finally ECL82 = PCL82 = UCL82 except for heater V/I and MINOR
differences which are not relevant in the context of your question.

Regards,
Gio

[Alex]

Hello,

You know those Japanese tuner-amplifiers of the 60's which have dual AM
receivers, one possibly with FM and a stereo amplifier, typically powered by
6BM8 tubes...

I got one of those, Pioneer SM-B200A. It is not working -- the tubes get hot
and "burning". Most likely dead electrolytic and/or leaky interstage paper
caps which bring positive bias to grids. I need to fix the beast before
getting rid of it.

I have no schematic, but tracing the wiring, realised that 6BM8s have fixed
bias originating from some selenium rectifier. Cathodes of 6BM8s are
grounded. More over, there are flinsy open (to dust) carbon trimpots to
"balance" (!) bias of the output tubes. At the same time there is no way of
knowing values of the currents in the tubes. There are no even small current
sensing (measuring) resistors in the cathodes. That design does not make
much sense.

I would get rid of the fixed bias in favour of automatic bias. Automatic
bias will be more robust and forgiving to tubes mismatch. It will not care
if this selenium rectifier dies altogether. I would throw resistors from
each cathode to ground, plus another between the cathodes to be able to
check unbalance.

Now my question is, from your experience, what should these resistor values
be, to get a benign class AB? I would guess 300R from each cathode to ground
and 100R between the cathodes?

Regards,
Alex

[Alex]

"flipper" wrote in message
...
On Mon, 27 Dec 2010 13:30:43 -0800, "Alex"
wrote:

Hello,

You know those Japanese tuner-amplifiers of the 60's which have dual AM
receivers, one possibly with FM and a stereo amplifier, typically powered
by
6BM8 tubes...

I got one of those, Pioneer SM-B200A. It is not working -- the tubes get
hot
and "burning". Most likely dead electrolytic and/or leaky interstage paper
caps which bring positive bias to grids. I need to fix the beast before
getting rid of it.

I have no schematic,

I found one. If you would like I can post it to
alt.binaries.schematics.electronic

Thank you very-very much for the offer. Schematic will help. Could you
please e-mail it to
alex-at-microcoin-dot-com ?

Regards,
Alex

[Alex]

"GRe" wrote in message
...
I would get rid of the fixed bias in favour of automatic bias. Automatic
bias will be more robust and forgiving to tubes mismatch. It will not
care
if this selenium rectifier dies altogether. I would throw resistors from
each cathode to ground, plus another between the cathodes to be able to
check unbalance.

Now my question is, from your experience, what should these resistor
values be, to get a benign class AB? I would guess 300R from each cathode
to ground and 100R between the cathodes?

Regards,
Alex

Adding the 100R will only decrease the voltage imbalance between cathodes
but will increase anode current imbalance. And that's not what you want.

Thanks.

Will it help to use independent resistors from each cathode to ground and
then a large non-polarised electrolytic between the cathodes? I would think
this will give the best DC current balance, highest gain while in class A
(low volume) -- hence lowest output impedance, but might result in higher
distortion in transition from A to B (at higher volume), but who cares?

Alex


Datasheet Rk for ECL82/PCL82/UCL82 in AB-PP:
B+ (V) 100 170 200 230 250
Rk (ohm) 140 125 170 200 220
Rk's above are spec'd in bean counter mode, Patrick would say.
So, in single cathode resistor config the values must be doubled.

More details:
http://frank.pocnet.net/sheets/030/e/ECL82.pdf
http://frank.pocnet.net/sheets/030/p/PCL82.pdf
http://frank.pocnet.net/sheets/030/u/UCL82.pdf

Equivalents:
ECL82 = 6BM8 (Htr. 6.3V)
PCL82 = 16A8 (Htr. 0.3A)
UCL82 = 50BM8 (Htr. 0.1A)
Finally ECL82 = PCL82 = UCL82 except for heater V/I and MINOR
differences which are not relevant in the context of your question.

Regards,
Gio

[Engineer[_2_]]

On Dec 27 2010, 10:11*am, flipper wrote:
On Mon, 27 Dec 2010 23:24:44 -0800, "Alex"
wrote:





"GRe" wrote in message
.. .
I would get rid of the fixed bias in favour of automatic bias. Automatic
bias will be more robust and forgiving to tubes mismatch. It will not
care
if this selenium rectifier dies altogether. I would throw resistors from
each cathode to ground, plus another between the cathodes to be able to
check unbalance.

Now my question is, from your experience, what should these resistor
values be, to get a benign class AB? I would guess 300R from each cathode
to ground and 100R between the cathodes?

Regards,
Alex

Adding the 100R will only decrease the voltage imbalance between cathodes
but will increase anode current imbalance. *And that's not what you want.

Thanks.

Will it help to use independent resistors from each cathode to ground and
then a large non-polarised electrolytic between the cathodes? I would think
this will give the best DC current balance, highest gain while in class A
(low volume) -- hence lowest output impedance, but might result in higher
distortion in transition from A to B (at higher volume), but who cares?

Personally I'd restore it to stock and if you wanted to improve fault
tolerance put some fuses in the output circuit but if you're dead set
on self bias then separate Rks each with their own bypass cap from
cathode to ground will give you the best idle balance.

But fixed bias is technically 'better' since bias will not shift when
it enters Class B and, as I mentioned earlier, I'm sure the existing
balance pot was for dynamic balance (which is what matters), not idle
current.

Alex

Datasheet Rk for ECL82/PCL82/UCL82 in AB-PP:
B+ (V) * 100 *170 *200 *230 *250
Rk (ohm) 140 *125 *170 *200 *220
Rk's above are spec'd in bean counter mode, Patrick would say.
So, in single cathode resistor config the values must be doubled.

More details:
http://frank.pocnet.net/sheets/030/e/ECL82.pdf
http://frank.pocnet.net/sheets/030/p/PCL82.pdf
http://frank.pocnet.net/sheets/030/u/UCL82.pdf

Equivalents:
ECL82 = 6BM8 *(Htr. 6.3V)
PCL82 = 16A8 *(Htr. 0.3A)
UCL82 = 50BM8 (Htr. 0.1A)
Finally ECL82 = PCL82 = UCL82 except for heater V/I and MINOR
differences which are not relevant in the context of your question.

Regards,
Gio

There's another reason to keep the fixed bias... you get to keep an
extra 16 volts or so of B+. The SM-B200 has quite a low B+... only
some 240 VDC comes to mind. Also, the negative bias supply also feeds
the transistorized phono preamp. If the selenium bias rectifier is
dead replace it with a 1N4007 diode.
I've refurbished two of these units (one a basket case)... not easy to
work on. Make sure you replace all the grey, Suzuki 0.05 "oil filled"
caps - most will be leaky, including the coupling caps to the 6BM8's -
wreaks havok on the bias! The weird dual eye-tube is close to
"unobtanium" these days, and most are burned out by now, but you can
sub a single eye-tube and switch the AVC line between the two
receivers (I plant to use the now disconnected speaker reverse switch
for this.)
Good luck!
Cheers,
Roger

[Engineer[_2_]]

On Jan 1, 3:03*pm, Engineer wrote:
On Dec 27 2010, 10:11*am, flipper wrote:



On Mon, 27 Dec 2010 23:24:44 -0800, "Alex"
wrote:

"GRe" wrote in message
.. .
I would get rid of the fixed bias in favour of automatic bias. Automatic
bias will be more robust and forgiving to tubes mismatch. It will not
care
if this selenium rectifier dies altogether. I would throw resistors from
each cathode to ground, plus another between the cathodes to be able to
check unbalance.

Now my question is, from your experience, what should these resistor
values be, to get a benign class AB? I would guess 300R from each cathode
to ground and 100R between the cathodes?

Regards,
Alex

Adding the 100R will only decrease the voltage imbalance between cathodes
but will increase anode current imbalance. *And that's not what you want.

Thanks.

Will it help to use independent resistors from each cathode to ground and
then a large non-polarised electrolytic between the cathodes? I would think
this will give the best DC current balance, highest gain while in class A
(low volume) -- hence lowest output impedance, but might result in higher
distortion in transition from A to B (at higher volume), but who cares?

Personally I'd restore it to stock and if you wanted to improve fault
tolerance put some fuses in the output circuit but if you're dead set
on self bias then separate Rks each with their own bypass cap from
cathode to ground will give you the best idle balance.

But fixed bias is technically 'better' since bias will not shift when
it enters Class B and, as I mentioned earlier, I'm sure the existing
balance pot was for dynamic balance (which is what matters), not idle
current.

Alex

Datasheet Rk for ECL82/PCL82/UCL82 in AB-PP:
B+ (V) * 100 *170 *200 *230 *250
Rk (ohm) 140 *125 *170 *200 *220
Rk's above are spec'd in bean counter mode, Patrick would say.
So, in single cathode resistor config the values must be doubled.

More details:
http://frank.pocnet.net/sheets/030/e/ECL82.pdf
http://frank.pocnet.net/sheets/030/p/PCL82.pdf
http://frank.pocnet.net/sheets/030/u/UCL82.pdf

Equivalents:
ECL82 = 6BM8 *(Htr. 6.3V)
PCL82 = 16A8 *(Htr. 0.3A)
UCL82 = 50BM8 (Htr. 0.1A)
Finally ECL82 = PCL82 = UCL82 except for heater V/I and MINOR
differences which are not relevant in the context of your question.

Regards,
Gio

There's another reason to keep the fixed bias... you get to keep an
extra 16 volts or so of B+. * *The SM-B200 has quite a low B+... only
some 240 VDC comes to mind. *Also, the negative bias supply also feeds
the transistorized phono preamp. *If the selenium bias rectifier is
dead replace it with a 1N4007 diode.
I've refurbished two of these units (one a basket case)... not easy to
work on. *Make sure you replace all the grey, Suzuki 0.05 "oil filled"
caps - most will be leaky, including the coupling caps to the 6BM8's -
wreaks havok on the bias! *The weird dual eye-tube is close to
"unobtanium" these days, and most are burned out by now, but you can
sub a single eye-tube and switch the AVC line between the two
receivers (I plant to use the now disconnected speaker reverse switch
for this.)
Good luck!
Cheers,
Roger

Forgot to mention... these chassis have a neat centre channel or mono
option. The two stereo channels ground their respective 4 ohm OPT
taps (not the zero ohm tap as is conventional.) The number of OPT
secondary turns is the same from "4 ohms to zero ohms" as from "16
ohms to 4 ohms", both being a 4 ohm impedance match (as we know,
impedance is proportional to the square of the turns.) But now the L-
channel "16 ohm tap" is 180 degrees out of phase with the R-channel
"zero ohm" tap giving a 8 ohm match between the two of them, both
channels driven (needs an 8 ohm mono speaker, of course). No tricky
amplifier bridging necessary to combine the two output channels. Draw
a picture if this is not clear!
Cheers,
Roger