[Engineer]

"Bah, HUMbug", with apologies to Charles Dickens' Mr. Scrooge.

Hi, Vacuumlanders,
Well, I've just fixed an irritating audio amplifier problem...

Refurbishing a very ratty Heathtkit A8A amplifier (no schematic and
made by a less than skillful original constructor in the 1950's) I did
all the usual stuff (caps, filters, etc) and was left with good sound,
lots of power with low distortion but a persistent 60 Hz hum that was
just audible. Admittedly, I was using a large speaker with good low
end, but this was not acceptable - you could hear the hum 3 feet away,
not just with your ear up against it. BTW, tube line up: 2 x 6L6
(output) 6SN7 (phase splitter, plate of first half directly to grid of
second half, which has equal plate and cathode resistors.) 6SJ7 (2nd
amplifier stage for "xtal/(aux" input) 6SJ7 (1st amplifier stage for
"phono/mic" input)

Here's a summary of what I did:
Initial forensics:
1. Removed 6SN7 phase splitter. Hum gone, not surprised. Put it
back.
2. Removed 2nd 6SJ7 amplifier tube. Hum gone. 6SN7 stages alone very
quiet. Put tube back.
3. Removed 1st 6SJ7 amplifier tube. Hum still there (of course, it was
on both "phono/mic" and the high level "xtal/(aux) input.) So, it's
in the 2nd 6SJ7 stage. Put the tube back.

Non-solutions...
4. Swapped the 6SJ7 tubes just in case of a H to K short. Not the
problem, hum still there.
5. Added temporary extra massive B+ filtering (but it was not a 120 Hz
hum, anyway, so this was just time wasting alchemy!) Actually, I also
did some tests with a separate DC P/S with virtually no B+ ripple - 60
Hz hum still there. Went back to local rectifier B+.
6. Lifted 6.3 VAC centre tap from ground, put 100 ohm pot across
winding and grounded slider to make a "hum pot". Note very tiny
changes on a VTVM across the o/p meter as pot moved, but not a fix,
hum still there.
7. More alchemy... removed pot and grounded first one then the other
side of the heater winding. Tiny changes only. Hum still there, of
course it was! Put the CT back to ground.
8. Thought about heater to cathode emission... Raised heater centre
tap to +20 VDC with a potential divider across B+ (decoupled with 100
MFD.) Hum still there. Pity, I had hopes for this!

Success...!
9. Looked at circuit again and tried think outside of the box... It
had to be the 6SJ7 screen. Both the 1st and 2nd 6SJ7 screen grids
were fed by 470 Kohm, decoupled by only 0.12 MFD caps (I had found old
wax 0.05 MFD paper tubulars there and replaced them with the 0.12's.)
The LP break point (ignoring the parallel screen grid impedance) is
around 3 Hz, seems low enough... and 60 Hz would be some 24 dB less.
But is that enough?
10. But what if the 60 Hz source was in the tube itself and had a much
lower source impedance? (Thevenin, may you be revered for ever...!),
in that case 0.05 or 0.12 MFD won't so the job. So, I decoupled the
2nd 6SJ7 screen with 20 MFD (big, but it was to hand). Ah, ha! HUM
GONE!
11. After a quick calculation and reality check with a DVM (and to
save capacitors) I connected the screens of both 6SJ7's together with
the 20 MFD to ground. The screen voltage is now about 43 VDC for both
6SJ7's, with plates around 77 VDC (B+ = 218 VDC) for the 2nd 6SJ7 and
98 VDC (B+ = 190 VDC) for the 1st 6SJ7 (it has higher bias so takes
less current.)

Moral of the story:
1. Don't assume the circuits you find in old stuff are original,
optimal or even properly designed in the first place (and never assume
this for kit constructed stuff.)
2. Do the forensics carefully, and
3. Work everything out from first principles.
Cheers,
Roger

[Phil Allison]

"Engineer"

Well, I've just fixed an irritating audio amplifier problem...

** Really ?


Success...!
9. Looked at circuit again and tried think outside of the box... It
had to be the 6SJ7 screen. Both the 1st and 2nd 6SJ7 screen grids
were fed by 470 Kohm, decoupled by only 0.12 MFD caps (I had found old
wax 0.05 MFD paper tubulars there and replaced them with the 0.12's.)
The LP break point (ignoring the parallel screen grid impedance) is
around 3 Hz, seems low enough... and 60 Hz would be some 24 dB less.
But is that enough?

10. But what if the 60 Hz source was in the tube itself and had a much
lower source impedance?


** How is that possible ??

The only source of 60 Hz is the heater voltage.


in that case 0.05 or 0.12 MFD won't so the job. So, I decoupled the
2nd 6SJ7 screen with 20 MFD (big, but it was to hand). Ah, ha! HUM
GONE!


** You sure ?


11. After a quick calculation and reality check with a DVM (and to
save capacitors) I connected the screens of both 6SJ7's together with
the 20 MFD to ground. The screen voltage is now about 43 VDC for both
6SJ7's, with plates around 77 VDC (B+ = 218 VDC) for the 2nd 6SJ7 and
98 VDC (B+ = 190 VDC) for the 1st 6SJ7 (it has higher bias so takes
less current.)


** With 470 kohms feeding that 20 uF electro - the screen voltage is bound
to be less than before due to leakage current. The 6SL7 has simply dropped
gain.

Very bad practice to use an electro there.




...... Phil

[Peter Wieck]

On Apr 7, 5:13 pm, "Engineer" wrote:
"Bah, HUMbug", with apologies to Charles Dickens' Mr. Scrooge.

Hi, Vacuumlanders,
Well, I've just fixed an irritating audio amplifier problem...

Refurbishing a very ratty Heathtkit A8A amplifier (no schematic and
made by a less than skillful original constructor in the 1950's) I did
all the usual stuff (caps, filters, etc) and was left with good sound,
lots of power with low distortion but a persistent 60 Hz hum that was
just audible. Admittedly, I was using a large speaker with good low
end, but this was not acceptable - you could hear the hum 3 feet away,
not just with your ear up against it. BTW, tube line up: 2 x 6L6
(output) 6SN7 (phase splitter, plate of first half directly to grid of
second half, which has equal plate and cathode resistors.) 6SJ7 (2nd
amplifier stage for "xtal/(aux" input) 6SJ7 (1st amplifier stage for
"phono/mic" input)

Here's a summary of what I did:
Initial forensics:
1. Removed 6SN7 phase splitter. Hum gone, not surprised. Put it
back.
2. Removed 2nd 6SJ7 amplifier tube. Hum gone. 6SN7 stages alone very
quiet. Put tube back.
3. Removed 1st 6SJ7 amplifier tube. Hum still there (of course, it was
on both "phono/mic" and the high level "xtal/(aux) input.) So, it's
in the 2nd 6SJ7 stage. Put the tube back.

Non-solutions...
4. Swapped the 6SJ7 tubes just in case of a H to K short. Not the
problem, hum still there.
5. Added temporary extra massive B+ filtering (but it was not a 120 Hz
hum, anyway, so this was just time wasting alchemy!) Actually, I also
did some tests with a separate DC P/S with virtually no B+ ripple - 60
Hz hum still there. Went back to local rectifier B+.
6. Lifted 6.3 VAC centre tap from ground, put 100 ohm pot across
winding and grounded slider to make a "hum pot". Note very tiny
changes on a VTVM across the o/p meter as pot moved, but not a fix,
hum still there.
7. More alchemy... removed pot and grounded first one then the other
side of the heater winding. Tiny changes only. Hum still there, of
course it was! Put the CT back to ground.
8. Thought about heater to cathode emission... Raised heater centre
tap to +20 VDC with a potential divider across B+ (decoupled with 100
MFD.) Hum still there. Pity, I had hopes for this!

Success...!
9. Looked at circuit again and tried think outside of the box... It
had to be the 6SJ7 screen. Both the 1st and 2nd 6SJ7 screen grids
were fed by 470 Kohm, decoupled by only 0.12 MFD caps (I had found old
wax 0.05 MFD paper tubulars there and replaced them with the 0.12's.)
The LP break point (ignoring the parallel screen grid impedance) is
around 3 Hz, seems low enough... and 60 Hz would be some 24 dB less.
But is that enough?
10. But what if the 60 Hz source was in the tube itself and had a much
lower source impedance? (Thevenin, may you be revered for ever...!),
in that case 0.05 or 0.12 MFD won't so the job. So, I decoupled the
2nd 6SJ7 screen with 20 MFD (big, but it was to hand). Ah, ha! HUM
GONE!
11. After a quick calculation and reality check with a DVM (and to
save capacitors) I connected the screens of both 6SJ7's together with
the 20 MFD to ground. The screen voltage is now about 43 VDC for both
6SJ7's, with plates around 77 VDC (B+ = 218 VDC) for the 2nd 6SJ7 and
98 VDC (B+ = 190 VDC) for the 1st 6SJ7 (it has higher bias so takes
less current.)

Moral of the story:
1. Don't assume the circuits you find in old stuff are original,
optimal or even properly designed in the first place (and never assume
this for kit constructed stuff.)
2. Do the forensics carefully, and
3. Work everything out from first principles.
Cheers,
Roger

Good Story!

For giggles:
a) try replacing that 0.10-now-20uf cap with a 1uF film cap.
b) look for a weak chassis/ground-rail connection that may be creating
a hum-loop...
c) conversely, check for a burnt or cracked bit of insulation causing
an inadvertent ground.
d) wire/lead dressing has much to do with hum on (especially) vintage
Heath equipment.

You are working from an OEM schematic, I hope? Heath made their pieces
in their uncounted thousands with their customers largely being the
beta-testers. Check if possible for the latest version of the
schematic, or latter day service literature. But I entirely agree with
you on the value of taking nothing for granted on kit-built
equipment.

Peter Wieck
Wyncote, PA

[Engineer]

On Apr 8, 9:46 am, "Peter Wieck" wrote:
On Apr 7, 5:13 pm, "Engineer" wrote:


"Bah, HUMbug", with apologies to Charles Dickens' Mr. Scrooge.

(snip)

Good Story!

Thanks, Peter, it vented my frustration a bit!

For giggles:
a) try replacing that 0.10-now-20uf cap with a 1uF film cap.

That was my preference but I don't have one to hand. Will try. My
replacement 0.12 MFD is still across the 20 MFD (actually two of them,
as I connected the two screens - well decoupled by the big cap.) If I
do that I'll separate the screens, of course.

b) look for a weak chassis/ground-rail connection that may be creating
a hum-loop...

Already checked - none found.

c) conversely, check for a burnt or cracked bit of insulation causing
an inadvertent ground.

Already checked - none found.

d) wire/lead dressing has much to do with hum on (especially) vintage
Heath equipment.

Ah, ha... yes, this one was a dogs breakfast but I've cleaned up some
of it.

You are working from an OEM schematic, I hope?

No, I could not find the schematic even after a search and putting out
an "APB" for the A8A. I reverse engineered what I saw to sketch the
circuit but, as I said elsewhere, the original constructor may have
use the wrong values (I found a couple wrong.) This is a very basic
circuit so I just fixed everything to first principles of a simple
6L6-6SN7-6SJ7 P-P amp design (proper filter, plate and cathode bias
resistor values) - now it all works as it should, actually very
well.

Heath made their pieces
in their uncounted thousands with their customers largely being the
beta-testers. Check if possible for the latest version of the
schematic, or latter day service literature.

Don't have any...!

But I entirely agree with
you on the value of taking nothing for granted on kit-built
equipment.

Thanks, this was a major point in my post.

Peter Wieck
Wyncote, PA- Hide quoted text -

Cheers,
Roger

[Ian Iveson]

Engineer wrote

"Bah, HUMbug", with apologies to Charles Dickens' Mr. Scrooge.

Hi, Vacuumlanders,
Well, I've just fixed an irritating audio amplifier problem...

Refurbishing a very ratty Heathtkit A8A amplifier (no schematic and
made by a less than skillful original constructor in the 1950's) I
did
all the usual stuff (caps, filters, etc) and was left with good
sound,
lots of power with low distortion but a persistent 60 Hz hum that
was
just audible. Admittedly, I was using a large speaker with good low
end, but this was not acceptable - you could hear the hum 3 feet
away,
not just with your ear up against it. BTW, tube line up: 2 x 6L6
(output) 6SN7 (phase splitter, plate of first half directly to grid
of
second half, which has equal plate and cathode resistors.) 6SJ7 (2nd
amplifier stage for "xtal/(aux" input) 6SJ7 (1st amplifier stage for
"phono/mic" input)

Here's a summary of what I did:
Initial forensics:
1. Removed 6SN7 phase splitter. Hum gone, not surprised. Put it
back.
2. Removed 2nd 6SJ7 amplifier tube. Hum gone. 6SN7 stages alone very
quiet. Put tube back.
3. Removed 1st 6SJ7 amplifier tube. Hum still there (of course, it
was
on both "phono/mic" and the high level "xtal/(aux) input.) So, it's
in the 2nd 6SJ7 stage. Put the tube back.

Non-solutions...
4. Swapped the 6SJ7 tubes just in case of a H to K short. Not the
problem, hum still there.
5. Added temporary extra massive B+ filtering (but it was not a 120
Hz
hum, anyway, so this was just time wasting alchemy!) Actually, I
also
did some tests with a separate DC P/S with virtually no B+ ripple -
60
Hz hum still there. Went back to local rectifier B+.
6. Lifted 6.3 VAC centre tap from ground, put 100 ohm pot across
winding and grounded slider to make a "hum pot". Note very tiny
changes on a VTVM across the o/p meter as pot moved, but not a fix,
hum still there.
7. More alchemy... removed pot and grounded first one then the other
side of the heater winding. Tiny changes only. Hum still there, of
course it was! Put the CT back to ground.
8. Thought about heater to cathode emission... Raised heater centre
tap to +20 VDC with a potential divider across B+ (decoupled with
100
MFD.) Hum still there. Pity, I had hopes for this!

Success...!
9. Looked at circuit again and tried think outside of the box... It
had to be the 6SJ7 screen. Both the 1st and 2nd 6SJ7 screen grids
were fed by 470 Kohm, decoupled by only 0.12 MFD caps (I had found
old
wax 0.05 MFD paper tubulars there and replaced them with the
0.12's.)
The LP break point (ignoring the parallel screen grid impedance) is
around 3 Hz, seems low enough... and 60 Hz would be some 24 dB less.
But is that enough?
10. But what if the 60 Hz source was in the tube itself and had a
much
lower source impedance? (Thevenin, may you be revered for ever...!),
in that case 0.05 or 0.12 MFD won't so the job. So, I decoupled the
2nd 6SJ7 screen with 20 MFD (big, but it was to hand). Ah, ha! HUM
GONE!
11. After a quick calculation and reality check with a DVM (and to
save capacitors) I connected the screens of both 6SJ7's together
with
the 20 MFD to ground. The screen voltage is now about 43 VDC for
both
6SJ7's, with plates around 77 VDC (B+ = 218 VDC) for the 2nd 6SJ7
and
98 VDC (B+ = 190 VDC) for the 1st 6SJ7 (it has higher bias so takes
less current.)

Moral of the story:
1. Don't assume the circuits you find in old stuff are original,
optimal or even properly designed in the first place (and never
assume
this for kit constructed stuff.)
2. Do the forensics carefully, and
3. Work everything out from first principles.
Cheers,
Roger


Er...so where did the 60Hz originate before you thought you had
decoupled it?

Out of which box did you think?

Perhaps you have coupled, rather than decoupled.

cheers, Ian

[Engineer]

On Apr 8, 11:12 am, "Ian Iveson"
wrote:
Engineer wrote


"Bah, HUMbug", with apologies to Charles Dickens' Mr. Scrooge.

(snip)

Er...so where did the 60Hz originate before you thought you had
decoupled it?

It did not come from the B+ and was not a hum or ground loop, nor was
there any obvious 60 Hz inductive or capacitive pick-up (although I
would like to rewire all the heaters with more tightly twisted
pairs.) IMO, it has to be the 6SJ7 heater, i.e. in the tube itself,
somehow coupled to the screen grid through a low enough impedance that
the 0.1 MFD did not sink it adequately. I will check the tube base
for any leakage from pin 2 to pin 7 but, since the hum was no less
with either side of the heater grounded, this is unlikely.

Out of which box did you think?

Ah, ha... the box we all build around ourselves when we think we know
it all and don't check the basic physics and engineering. "What you
know that ain't so is the problem". I've forgotten who first said
that.

Perhaps you have coupled, rather than decoupled.

I don't think so... g

cheers, Ian

Cheers,
Roger

[Engineer]

On Apr 8, 5:54 pm, "Engineer" wrote:
On Apr 8, 11:12 am, "Ian Iveson"
wrote:

Engineer wrote

(snip)

I will check the tube base
for any leakage from pin 2 to pin 7 but, since the hum was no less
with either side of the heater grounded, this is unlikely.

Correction: pin 6 (screen) to pin 2 (unlikely) or pin 7 (closer)
Sorry!

(snip)

Cheers,
Roger

[Ian Iveson]

If the grid and cathode are both infected with the same hum, then
coupling it to the screen would help, but may not be the best
solution.

Perhaps you need twice the capacitance in such a circumstance, because
the frequency is half of what it expected.

sorry for top post....time for bed


"Engineer" wrote in message
oups.com...
On Apr 8, 11:12 am, "Ian Iveson"
wrote:
Engineer wrote


"Bah, HUMbug", with apologies to Charles Dickens' Mr. Scrooge.

(snip)

Er...so where did the 60Hz originate before you thought you had
decoupled it?

It did not come from the B+ and was not a hum or ground loop, nor
was
there any obvious 60 Hz inductive or capacitive pick-up (although I
would like to rewire all the heaters with more tightly twisted
pairs.) IMO, it has to be the 6SJ7 heater, i.e. in the tube itself,
somehow coupled to the screen grid through a low enough impedance
that
the 0.1 MFD did not sink it adequately. I will check the tube base
for any leakage from pin 2 to pin 7 but, since the hum was no less
with either side of the heater grounded, this is unlikely.

Out of which box did you think?

Ah, ha... the box we all build around ourselves when we think we
know
it all and don't check the basic physics and engineering. "What you
know that ain't so is the problem". I've forgotten who first said
that.

Perhaps you have coupled, rather than decoupled.

I don't think so... g

cheers, Ian

Cheers,
Roger

[Peter Wieck]

On Apr 8, 10:47 am, "Engineer" wrote:
On Apr 8, 9:46 am, "Peter Wieck" wrote:

On Apr 7, 5:13 pm, "Engineer" wrote:

"Bah, HUMbug", with apologies to Charles Dickens' Mr. Scrooge.

(snip)

Good Story!

Thanks, Peter, it vented my frustration a bit!

For giggles:
a) try replacing that 0.10-now-20uf cap with a 1uF film cap.

That was my preference but I don't have one to hand. Will try. My
replacement 0.12 MFD is still across the 20 MFD (actually two of them,
as I connected the two screens - well decoupled by the big cap.) If I
do that I'll separate the screens, of course.

b) look for a weak chassis/ground-rail connection that may be creating
a hum-loop...

Already checked - none found.

c) conversely, check for a burnt or cracked bit of insulation causing
an inadvertent ground.

Already checked - none found.

d) wire/lead dressing has much to do with hum on (especially) vintage
Heath equipment.

Ah, ha... yes, this one was a dogs breakfast but I've cleaned up some
of it.

You are working from an OEM schematic, I hope?

No, I could not find the schematic even after a search and putting out
an "APB" for the A8A. I reverse engineered what I saw to sketch the
circuit but, as I said elsewhere, the original constructor may have
use the wrong values (I found a couple wrong.) This is a very basic
circuit so I just fixed everything to first principles of a simple
6L6-6SN7-6SJ7 P-P amp design (proper filter, plate and cathode bias
resistor values) - now it all works as it should, actually very
well.

Heath made their pieces
in their uncounted thousands with their customers largely being the
beta-testers. Check if possible for the latest version of the
schematic, or latter day service literature.

Don't have any...!

But I entirely agree with
you on the value of taking nothing for granted on kit-built
equipment.

Thanks, this was a major point in my post.

Peter Wieck
Wyncote, PA- Hide quoted text -

Cheers,
Roger

http://www.historicphotoarchive.com/...athkit_a8a.htm

Might be a source.

Peter Wieck
Wyncote, PA

[Engineer]

On Apr 9, 6:41 am, "Peter Wieck" wrote:
On Apr 8, 10:47 am, "Engineer" wrote:

On Apr 8, 9:46 am, "Peter Wieck" wrote:

(snip)


http://www.historicphotoarchive.com/...athkit_a8a.htm

Might be a source.

Peter Wieck
Wyncote, PA- Hide quoted text -


Thanks, Peter, I already found this one. But it's not really worth it
at over CDN$14.00 (incl. mailing) now that I have virtually redesigned
the circuit. However, I'd be happy to receive a copy of just the
schematic and pay the postage (email me first to avoid a flood!)
Cheers,
Roger

[Engineer]

On Apr 8, 8:10 pm, "Ian Iveson"
wrote:
If the grid and cathode are both infected with the same hum, then
coupling it to the screen would help, but may not be the best
solution.

Might work but I don't really see the need. The cathode is presently
decoupled by the original 25 MFD cap. Just had a thought... what if it
now has a large ESR?! Anyway, since the bias resistor is 570 ohms,
the corner frequency is 11 Hz (if the 25 MFD cap is good)... low but
not "zero". I think I'll put a new 100 MFD there to be sure (they
didn't have cheap 100's back in 1950!) By rights, I should then
recheck the hum with 0.1 MFD alone back on the screen (actually, I
left it in the circuit just in case the 20 MFD was a bit lax at HF.)
Food for thought...


Perhaps you need twice the capacitance in such a circumstance, because
the frequency is half of what it expected.

You mean the original 0.1 MFD was designed to decouple any residual
120 Hz on the screen B+ supply but, instead (or as well as) it is
subjected to 60 Hz from the heater - my speculative lower impedance
internal tube source. I think the low 60 Hz source impedance effect
dominates the frequency effect so it needs the big cap for that. But
see above - got to try the 100 MFD!


sorry for top post....time for bed


Cheers,
Roger