Epiphone Valve Junior Buzz/Hum

[b***@feedbackloop.net]

I just got one of the Valve Junior amps, which I really like. however,
there's a buzz I can't seem to get rid of.

First off, this is basically a tweed champ, built cheaply, and with a
few changes. The topology is as follows. Gain stage coupling cap,
voltage divider with two 1M resistors, and a 1M volume pot, which feeds
the grid of the next gain stage.

So there's a 120cycle buzz that is controllable by the volume pot when
no guitar is plugged in. It happens with every preamp tube I try.

Here's where it gets weird. I replaced the coupling cap and while it
was out I hooked up the amp. There was still noise. The only things
before the volume knob were the two 1M resistors for a voltage divider.
I tried shorting out the voltage 1M to ground, and now it was silent
when the volume knob was full up or full down, but in the middle there
is still noteable buzz. However, if I remove the quick connect that
connects the volume pot to the circuit board the buzz goes away.

So, if you ground out the wiper send to the grid of the 2nd gain stage,
the buzz goes away. If you put resistance between the wiper and
ground, the comes back. But if you remove the connection from the
board that connects the wiper to the grid, the buzz goes away.

I should know how to solve this, but nothing seems to be working. I've
tried a little bit of shielding on the leads going to the pot, and
moving the leads going to the pot, but there's no change.

Ideas?

Thanks!

Ben Adrian

[b***@feedbackloop.net]

Edit... It still makes the buzz when you disconnect the quick connect
that connects the pot.

Ben

[b***@feedbackloop.net]

Another edit. When I ground the grid of the second gain stage, the amp
is totally quiet (except for a bit of single ended amp hum that doesn't
bother me). When I leave it open and connected to nothing it's picking
up the internal buzz, which makes me think the board layout is bad.

I'm thinking maybe I can just run a shielded wire feom the volume put
straight to the tube socket... but then I'd have to remove a trace from
the board.

But it definitely seems like the board trace is getting induced hum.

Cheers.

Ben Adrian

[Stephen Cowell]

Post by b***@feedbackloop.net
Another edit. When I ground the grid of the second gain stage, the amp
is totally quiet (except for a bit of single ended amp hum that doesn't
bother me). When I leave it open and connected to nothing it's picking
up the internal buzz, which makes me think the board layout is bad.
I'm thinking maybe I can just run a shielded wire feom the volume put
straight to the tube socket... but then I'd have to remove a trace from
the board.

I cut PC board traces on the merest whim... you're
not meaning that you'd have to take an X-acto knife
and physically scrape all the conductor off, certainly.

Post by b***@feedbackloop.net
But it definitely seems like the board trace is getting induced hum.

Welcome to the world of voodoo... you'll never know
without doing the experiment. Watch where you get
your grounds from... try to get the grid return ground
near to that stage's cathode ground. Good luck!
__
Steve
.

[b***@feedbackloop.net]

Well, I'm concerned that the actual trace placement is what is causing
the induced hum.

Here's my plan.

All the transformer leads were left at full length and bundled on the
inside. I'm going to unbundle, trim where necessary, and route nicely.

I'm going to take the 100 ohm resistors on the heaters that go to
ground and connect that to the cathode of the EL84, a la the basic AX84
amp.

If that doesn't work I'm going to convert to DC filaments.

In adddition, I'm going to add a standby switch and a tone knob, just
because I can.

If none of that works, I'll then scrape the trace and run a shielded
wire.


Comments, advice?

Cheers!

Ben Adrian

[l***@aol.com]

Post by b***@feedbackloop.net
Well, I'm concerned that the actual trace placement is what is causing
the induced hum.
Here's my plan.
All the transformer leads were left at full length and bundled on the
inside. I'm going to unbundle, trim where necessary, and route nicely.
I'm going to take the 100 ohm resistors on the heaters that go to
ground and connect that to the cathode of the EL84, a la the basic AX84
amp.
If that doesn't work I'm going to convert to DC filaments.
In adddition, I'm going to add a standby switch and a tone knob, just
because I can.
If none of that works, I'll then scrape the trace and run a shielded
wire.
Comments, advice?

Hi Ben,

Your key (and excellent) description stating "120hz" hum indicates it
has nothing to do with tube heaters and likely has nothing to do with
the PCB layout (though remotely possible). 120hz hum comes from the B+
supply (rect & filter) - it's the ripple freq produced by the fullwave
rect from its 60hz power source. Therefore, whatever is causing the
supply ripple to be objectionably audible, your above ideas won't
affect it. Those ideas would only affect 60hz hum. You may even have
a marginal new filter cap, it is not unknown. A less-than-happy B+ gnd
is another. Some unwanted means of magnetic coupling of the preamp ckt
to the primary supply component(s) is another. Even an unhappy filter
cap and/or its rail resistor futher from the main filter can cause this
by causing inadequate stage decoupling. Similarly, any other causes
will somehow relate to the B+ supply ripple. A scope would help you.

[Stephen Cowell]

Post by l***@aol.com

Post by b***@feedbackloop.net
Well, I'm concerned that the actual trace placement is what is causing
the induced hum.
Here's my plan.
All the transformer leads were left at full length and bundled on the
inside. I'm going to unbundle, trim where necessary, and route nicely.
I'm going to take the 100 ohm resistors on the heaters that go to
ground and connect that to the cathode of the EL84, a la the basic AX84
amp.
If that doesn't work I'm going to convert to DC filaments.
In adddition, I'm going to add a standby switch and a tone knob, just
because I can.
If none of that works, I'll then scrape the trace and run a shielded
wire.
Comments, advice?

Hi Ben,
Your key (and excellent) description stating "120hz" hum indicates it
has nothing to do with tube heaters and likely has nothing to do with
the PCB layout (though remotely possible). 120hz hum comes from the B+
supply (rect & filter) - it's the ripple freq produced by the fullwave
rect from its 60hz power source. Therefore, whatever is causing the
supply ripple to be objectionably audible, your above ideas won't
affect it. Those ideas would only affect 60hz hum. You may even have
a marginal new filter cap, it is not unknown. A less-than-happy B+ gnd
is another. Some unwanted means of magnetic coupling of the preamp ckt
to the primary supply component(s) is another. Even an unhappy filter
cap and/or its rail resistor futher from the main filter can cause this
by causing inadequate stage decoupling. Similarly, any other causes
will somehow relate to the B+ supply ripple. A scope would help you.

Yeah... try a much bigger filter cap and see if that
helps... if so, then that's your problem. Push-pull
amps get away with less filtering, since the final
section cancels some of this out.
__
Steve
.

[b***@feedbackloop.net]

I tried adding more filtering, and it reduced the low level single
ended hum, but I still have the buzz that's affecting the volume knob.


I also twisted the filament wires to where they connect to the board.

It could be 60 cycle hum and I'm just hearing a higher harmonic of it
in the 8 oinch speaker.

A scope would be cool, we have one at work for setting azimuth, maybe
I'll see if I can swipe it.

Ben

[Stephen Cowell]

Post by b***@feedbackloop.net
Well, I'm concerned that the actual trace placement is what is causing
the induced hum.
Here's my plan.
All the transformer leads were left at full length and bundled on the
inside. I'm going to unbundle, trim where necessary, and route nicely.

You can twist opposite phases and get cancellation...
highly recommended, better than shielding.

Post by b***@feedbackloop.net
I'm going to take the 100 ohm resistors on the heaters that go to
ground and connect that to the cathode of the EL84, a la the basic AX84
amp.

Hey, OK with me... single-ended amps don't
cancel filament hum naturally.

Post by b***@feedbackloop.net
If that doesn't work I'm going to convert to DC filaments.

Man, you're willing to go all out!

Post by b***@feedbackloop.net
In adddition, I'm going to add a standby switch and a tone knob, just
because I can.

Sure... watch the layout!

Post by b***@feedbackloop.net
If none of that works, I'll then scrape the trace and run a shielded
wire.

What does the trace do? Is it a preamp grid connection?
You can cut both ends of the trace and sub a wire... no
need to scrape the trace off the board. Then, if you decide
to go back with the trace, you can jumper the cuts.

First, I'd cure the hum, then go modding. One thing at
a time... believe me, I've been there. You don't want to
have to undo it all, then start over doing what you should
have been doing in the first place.
__
Steve
.

__
Steve
.

[b***@feedbackloop.net]

steve said
What does the trace do? Is it a preamp grid connection?
You can cut both ends of the trace and sub a wire... no
need to scrape the trace off the board. Then, if you decide
to go back with the trace, you can jumper the cuts.


The trace is the grid connection from the wiper of the volume pot. The
volume pot is connected to the board by a multi pin quick connector. I
detach it and disconnect the entire audio circuit before the trace
leading to the grid and I get the noise. However, if I ground the
trace or the pin, the buzz stops. So the second gain stage of the tube
is still live and idling, but the buzz is gone when I ground the grid.

So this means that the 2nd gain stage is amplifying the buzz, the buzz
isn't coming directly from the first gain stage, because it's out of
the circuit, and the level of the buxx can be adjusted by the volume
knob.

If the buzz was coming from noise in the B+ on the second gain stage,
shouldn't it be the same level regardless of volume knob position?

So it's not noise from the B+ of the first gain stage because it's no
longer in the circuit, and it not ripple on the plate of the second
gain stage because that would not then be controlled by the volume
knob.

I move the leads all around and get no change in the noise level.

I'm pretty stumped.

Ben

[Stephen Cowell]

Post by b***@feedbackloop.net
steve said
What does the trace do? Is it a preamp grid connection?
You can cut both ends of the trace and sub a wire... no
need to scrape the trace off the board. Then, if you decide
to go back with the trace, you can jumper the cuts.
The trace is the grid connection from the wiper of the volume pot. The
volume pot is connected to the board by a multi pin quick connector. I
detach it and disconnect the entire audio circuit before the trace
leading to the grid and I get the noise. However, if I ground the
trace or the pin, the buzz stops. So the second gain stage of the tube
is still live and idling, but the buzz is gone when I ground the grid.

All this going between completely open and shorted... not helpful.
Use a 1M resistor to shunt the input, then see what your noise
level is like. Use the ground nearest the cathode of the tube for
the shunt ground.

Post by b***@feedbackloop.net
So this means that the 2nd gain stage is amplifying the buzz, the buzz
isn't coming directly from the first gain stage, because it's out of
the circuit, and the level of the buxx can be adjusted by the volume
knob.

Yes, big antenna.

Post by b***@feedbackloop.net
If the buzz was coming from noise in the B+ on the second gain stage,
shouldn't it be the same level regardless of volume knob position?

Right. It's actual noise you're picking up... what kind of light
is over your bench?

Post by b***@feedbackloop.net
So it's not noise from the B+ of the first gain stage because it's no
longer in the circuit, and it not ripple on the plate of the second
gain stage because that would not then be controlled by the volume
knob.
I move the leads all around and get no change in the noise level.
I'm pretty stumped.

You're picking up noise from the environment... a buzz is
normally a fluorescent light. These things tend to go away
when the case is closed back up...
__
Steve
.

[b***@feedbackloop.net]

Update.

Last night I found that there was a jumper on the board that carried
the signal to the grid of the second gain stage. It went as follows:
wiper of volume pot, wire to quick connect, trace to jumper, jumper,
trace to grid.

I cut the jumper and the noise went away. So I wired up a direct wire
from the wiper to the nub of the cut jumper sticking out of the circuit
board, and it worked, the 120 cycle buzz was gone. I put the chassis
back in the cabinet and there was now a noteable 60 cycle buzz, and
there was noise when I turned the volume knob.

I put the chassis back ou top of the cabinet and it got quiet again,
but now I can juggle the speaker cable and get a 60 cycle buzz to
appear and disappear. But it does seem to get louder when I put it in
the cabinet!

Plus, when I turn up the volume knob when it's quiet, the volume will
reach a certain level (about 2-3 o'clock) and parasitic oscillations
start to appear. It got too late and I had to stop to go to bed. I
think it's a problem with the unshielded wire I put in and maybe a
ground/chassis issue. Moving the new lead with a chopstick while the
amp is live does give me a change in sound.

It should be noted that this amp has a star ground except for the input
jack and the speaker jack, which are both grounded throgh the chassis.
I run alligator clip leads from the speake ground to the star ground,
and from the input jack to the star ground, and nothing quiets down.

Steve, I don't have a bench. I live in an apartment, and I have a
little corner workspace I use when I sit on the floor. When I do any
major soldering, I take it to my kitchen counter. But there is not
flourescent light over my bench or even in my house at all.

Tonight my plan is to run a shielded lead in place of the unshielded
lead I ran. Also, to connect the speaker jack and/or the input jack to
the star ground if people seem to think that will help. I also read
that grounding the filter cap for the early gain stages closer to the
input jack can reduce noise. I'm wondering if there should be one star
ground for everything except for the input jack and preamp filter cap,
which ground to the input?

Earlier, Steve said I was willing to go all out. I'm just more
interested by the challenge and hopefully reach saticfaction from the
result. I'm not really spending any more money as I have parts at
home. I just enjoy the tinkering.

Cheers!
Ben Adrian

[b***@feedbackloop.net]

One more thing right now.

I do own a Silverface Champ (1969), which is more quiet than the
Epiphone. It does have some power tube filament hum, but it's not as
loud, even though it has the same filter cap values. And there's not
any buzzing or nouises at all.

I chould clarify... the noise in the Valve Junior won't be heard if I'm
playing loud, but if I have the volume up high and roll down my volume
knob to clean up the tone, you'll hear a bit of buzz coming through on
the quiet parts, and it really bugs me.

Cheers!

Ben Adrian

[l***@aol.com]

Post by b***@feedbackloop.net
So this means that the 2nd gain stage is amplifying the buzz, the buzz

No,it doesn't. If, as you say, the pot wiper goes straight to the
grid, all it means is that you've removed (by disconnecting) the tube's
necessary grid resistor, which the pot also functions as, causing the
tube's stage to be unhappy & with near-infinite input impedance that is
bound be rather noisy.

[b***@feedbackloop.net]

When I trimmed the trace to the grid, even without the load, it was
totally quiet, and still had the approaching infinite input impedance.
It had the exact same 120 cycle whether the volume pot was in the
circuit (which the pot then controlled), out of the circuit, or if I
had a 1M to ground alligator clipped in (which was even more noisy as
the jumper leads were long).

Still, I'm happy for al the responses, and I will double check the
second stage with a variety of loads.

Thanks, and any other advice is welcomed.

Ben Adrian

[Stephen Cowell]

Post by b***@feedbackloop.net
When I trimmed the trace to the grid, even without the load, it was
totally quiet, and still had the approaching infinite input impedance.
It had the exact same 120 cycle whether the volume pot was in the
circuit (which the pot then controlled), out of the circuit, or if I
had a 1M to ground alligator clipped in (which was even more noisy as
the jumper leads were long).

When the grid floats, the tube will block... you'll
get no noise. The grid needs a leak resistor in
order to maintain itself near zero volts DC.


Your trace on the PC board may have been
coupling to another trace... that's the way it's
done in the Champ and other Fenders... the
input jack and the volume pot wiper wires are
twisted together in order to get degenerative
feedback. There's voodoo in there... have fun!
__
Steve
.

[l***@aol.com]

Post by Stephen Cowell
When the grid floats, the tube will block... you'll
get no noise. The grid needs a leak resistor in
order to maintain itself near zero volts DC.

Good one. I was betting on the trace and/or surrounding board
condition to present maybe 10+m & maybe a few pf in the bargain - 'nuf
dinky gridleak to make it sound. He seems to be proving both our cases
by chopping the trace to silence.

[b***@feedbackloop.net]

I cut the trace and ran an indisidual wire to the board (where the
trace was cut) from the wiper. After some chopstick manipulation, I
got it to be pretty quiet.

Next I replaced the first 22uf filter cap with an 80 uf. That took
care of most of the single ended power tube hum.

Then I replaced the chassis grounding input jack with an isolated one
and ran the input jack ground to the star ground.

It's working really well now.

Now, a question. Is the 80uf filter cap going to cause quick wear on
the power tube since the amp has a SS rectifier and no standby switch?

Cheers!

Ben Adrian

[Stephen Cowell]

Post by b***@feedbackloop.net
I cut the trace and ran an indisidual wire to the board (where the
trace was cut) from the wiper. After some chopstick manipulation, I
got it to be pretty quiet.
Next I replaced the first 22uf filter cap with an 80 uf. That took
care of most of the single ended power tube hum.
Then I replaced the chassis grounding input jack with an isolated one
and ran the input jack ground to the star ground.
It's working really well now.
Now, a question. Is the 80uf filter cap going to cause quick wear on
the power tube since the amp has a SS rectifier and no standby switch?

No... the HV will actually come up slightly slower with
the big cap than with the small cap, not enough to make
any difference however. It's only one tube... screw it.
__
Steve
.

[l***@aol.com]

This post might be inappropriate. Click to display it.