• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

high voltage after cap. change

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Oddly enough, I just did this same amp and saw the same thing...
The problem was the grounding switch, when grounded to black.
Old wiring had no preference of polarization, and was dangerous.

The cord needed replacement anyways, so I opted to ground on
third green wire. I used 2 diodes in series, and another two in
parallel going the opposite way, so there would be 1.4V of lift.
I made sure to use big enough diodes that the fuse could blow.
2A fuse means you are using at least 3A diodes or bigger...
You can use a modern diode bridge, if you wire + and - together.
Put one AC to green ground, and the other AC to the chassis.
I do mean only AC marks on the bridge, not any actual AC....

I disconnected the option to ground on black. Can now ground
only on white, or the default open ended option is now diode lift.

That stopped the crazy high voltages I was measuring.

There is a 50uF/50V on the diode bridge board (actually piece
of strangely treated cardboard, maybe asbestos? I don't know).
That one is the negative bias. Should measure -57V at the cap,
and -45V at the tube (or the center tap of the balance pot)...
Let me clarify this is not the same diode bridge as the one
I mentioned earlier for grounding, entirely different things...

Since they got a 50V cap doing a 57V job, and this one was
clearly leaky (measured only -22V at the cap). I put a pair
of 200V 33uF panasonic cap (105C rated) in its place... This
was 66uF, closest I could make with parts I had on hand...

As for re-capping the topside, there is some sort of cover
for the caps, so any new ones need the same shape and size.
I pushed the old (Mallory) caps out of the paper tubes and
put 105C Nichicon's in place. Inside the original paper tubes.

I used two in parallel, but physically bottom to bottom with
a solid core wire hanging out each end of the paper tube.
I insulated enough to get past the metal can of the caps.
I siliconed the caps into the paper tubes, which is easier
undone than it sounds, cause this paper don't want to stick.
And the tops of both new caps in each tube are exposed
in case they need to burst for any reason...

Oh what else? I think I might have rewired the fuse to
make sure it was on the hot (black) side. And but spliced
two of the original wires to make longer (remember one
less was needed for the grounding switch). I re-used all
the original wire without trimming anything, just incase
someone ever needs to put back like original...

If you test this upside down and use gator clips for the
speaker load. Remember the first speaker jack has to
be propped open, else it shorts out the transformer...
This is safer than transformer open, but not as safe as
having the right load attached. The 6L6 tube nearer the
transformer will get crazy hot if the jack isn't propped.
I don't know why the overheating isn't symetrical...
If you plug only into the "external" jack, it doesn't do
anything to remove the safety short.. I don't know
what is the label on the first jack, cause it had been
scratched out by some idiot...
 
Last edited:
Cowboy I'm grasping at straws. Is it wired transformer to resistor to diode to (negative terminal of capacitor, pot and 10K resistor)? This should be simple???

Unsolder from the pot and see what you get. So it's just resistor-diode- cap in your little circuit. Should be -60v or more. (Standby switch OFF for SURE!)
 
That potless measurement might be RMS DC, like
with no cap at all (if the cap died open somehow)...
It does not mean it is steady at that voltage, unless
you measured on a scope or something fast like that.
I got -22VDC here, till I recapped, then got -57VDC.
This was measured with pot connected in both cases.

The cap has to hold down the negative volts through
90% or more of the power cycle. Cap only charges
briefly at the most negative peak. The circuit might be
dragging your cap down too easily during the rest of
the cycle. If the cap is not holding a charge, or open,
or high internal resistance (dried out inside).

And since you see -68V, and I saw -57V. We have
to ask why this cap is only rated 50V? Should we
be replacing with something rated slightly better
in the voltage department? I used 200V Panasonic.

------

You still need to make sure your grounding switch is
NOT somehow in the position that grounds black to
the chassis through a small value capacitor hidden
underneath the switch. Especially if you have white
or green safety ground hard wired (or even with a
diode lift) to the chassis.

Black ground caused all the crazy volts I measured.
I saw unexplainable things over 800V, and beyond...
Might also have contributed to electrolytic cap death.
The cap under the ground switch could be leaking?
Just cause its not elko, doesn't mean lasts forever.

Since I got rid of black as a grounding option and
started using a polarized three prong plug, it has
ceased to be a problem. Even if the grounding cap
leaks a little, it will no longer cause any harm.
 
Last edited:
Original was 50uF/50V. I replaced with 2x33uF/200V/105C,
cause those were the best and closest match I had on hand.
Obviously two caps in parallel don't look very authentic...

With 66uF in total, I measured -47V at the point on the
schematic that indicates -45V. I didn't sweat the 2V diff,
since was slightly over on the safe side, and sounded great.
The voltage on the cap itself was -57V.
 
Last edited:
to answer your questions, yes i put a new cap in, the cap i took out was a 50uf- 50v, schematics value is 50uf-50v, pot side of diode to ground 21ohms, 65 bassman has 2 prong plug. looks like i bought a bad cap. i guess i will try another 50uf-50v. i am using blue atoms
 
Many links to cutaway Sprague Atoms, see whats inside...
Sprague Atoms @ Mouser [Archive] - The Gear Page
Not bad, they cut up some scary fakes for comparison too.
But even the real Atoms seem to contain a much smaller
modern capacitor inside..

I note prominently on the outside: 65C. - Six Five See!
Atoms might be the right shape and size, but I would not
be caught dead using a low temp cap in that application.
Too hot for 65C cap just sitting in my car in the summer.

We can hope the temp printed on the cover is for show,
as those are a legacy style for restorations. The real cap
inside might be a much higher temperature grade? But
we have no way of proving that assumption....

I siliconed modern Nichicon 105C caps into the original
Mallory paper tubes, for reason of the mechanical size.
At least I can guess reasonably sure what I hid inside.
Fakes inside of fakes is a possibility too, I suppose...
 
Last edited:
what would happen if i replaced the 50uf/ 50v atom with a f&t 47uf/50v cap

Electrolytic caps typically have a tolerance of +/- 20% or more. Back when that amp was made +100/-50% tolerances were typical. IOW, there is no meaningful difference between 47 and 50uF, except that 50 is not a modern, standard value. The only reason to use those Sprague caps is if nothing else has long enough leads. It is their only favorable characteristic. The F&T is a much better cap.

I would definitely use a larger voltage rating; minimum 63V, better 80 and almost everybody in the biz just puts in a 100uF @ 100V. The only real constraint on the value is that the bias supply must charge up as fast or faster than the heaters and B+. 100uF is not a problem, more than that is, so don't go bigger unless you're religious about that standby switch.

During installation, remember the bias supply is negative and the positive side of that cap goes to ground. You can measure bias volts at pin 5 of the power tubes, use the balance pot to get them equal (or near 'nuff). Do this on standby without tubes. Then hit the switch and see that they are still equal and in that amp pretty close to where they were on s'by.

Much of your testing can be, and since you're having problems, should be done with a 100W lamp in series.

As previously mentioned, there's no reason to use a temp rating under 105C in a tube amp, You can often get away with 85C, but why ....
 
well someone said using a higher voltage might raise the bias volts. or wouldn't this bother anything? this is new to me i am more musician than amp tech.

Higher uF might become a problem if taken too far.
Anything up to 100uF would not be "too far" in my
opinion. Bigger values will provide a smoother bias.
But this negative bias has reach full charge faster
than the High Voltage supplied to the plates. Bigger
caps charge slower, so yeah, you can go too far.

The next common value above 50 is 68uF. I tried
66uF (as a pair of parallel 33's) and worked fine.
I had no good reason to use 33's, except that it
saved me from waiting on parts I did not have.
The worst offence, merely that is does not look
original to use two smaller radial caps where one
medium sided axial type is expected. Does not
cause any electrical problem...

Higher Volts and higher temp can only be good.
They do not affect the sound or operating point,
only living much longer under harsh conditions...

You can't take reliability too far, except price.
Any simple device that's expensive and popular
becomes target for fakers. And fake parts are
the opposite of reliable.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.