Ghost Note Causes

[drewl]

Hello all,

So what else can cause ghost noting in a tube amp besides filter caps and output trans?

Got a 75 Marshall Superlead that is heavily modded, extra tube with two extra gain stages added, Some kind of gooped "super secret" mod that seems to be a resonance boost.

Filter caps have all been replaced, output trans was tried, and output tubes look new.

It ghosts without the added gain stage engaged, but with all the wires floating around in this thing, I have to think that might be having an affect.
The shielded wiring is all grounded properly, but with this much gain who knows.

Thanks.

[Helmholtz]

If the ghosting is not caused by hum modulation, the amp may tend to oscillate (scope?).

Does the presence setting influence the ghosting?

[pdf64]

Is this ghosting at high power outputs, or something else?
Bear in mind that a 100W 1959 SL might be expected to have more HT ripple (and hence intermodulation at high outputs) than a 50W1987, as they have half the power amp HT filter capacitance but roughly twice the HT current draw.

[drewl]

I'll check again, but HT ripple didn't seem bad

Definitely more noticeable at higher output.

[Chuck H]

... and output tubes look new.

How do they test for current matching at idle or conducting signal? Ripple is ripple and if the tubes are conducting unevenly there will be more ripple in the output creating sum/difference notes. "look new" doesn't tell us that the output tubes are all PERFORMING similar enough. This could also be affected by components. Coupling caps leaking, drifted resistors, dead screen resistor, etc.

And though the description so far seems to indicate the ghosting is in the LF I'll mention speaker cone cry anyway. You can try the amp through another cabinet just to rule it out.

EDIT: Since you asked about causes I'll go ahead and mention a weird one I had. A cold solder joint that would only act funny at certain resonances. It seemed like ghost notes in that it was frequency specific though it was in the low midrange.

[drewl]

Well I finally had a chance to really go through this amp.

The ghosting happens with the extra tube stage removed, or switched off. I tried adding another filter node for it, but it didn't help.

The signal is clean going to the output tubes, checking the signal output, you can see some nasty "stuff" noise or whatever riding on the signal.

I tried jumpering in other filter caps, but the caps were replaced, tried different output tubes, different cab, different guitar.
Checked all grounds, the only thing I haven't checked is the OT, but owner said he tried another FWIW.

You can really hear the ghost notes up areound the 12th to 15th frets on the D and G strings, a weird out of tune lower note in between a 4th or 5th down

Any other ideas?

jmp_superlead_100w_1959.pdf

[Helmholtz]

The signal is clean going to the output tubes, checking the signal output, you can see some nasty "stuff" noise or whatever riding on the signal.

Scope pics?

Are power tubes balanced?

Did you try with presence up?

It might help to know how/where the added circuit is inserted.

[drewl]

The extra gain stage is inserted before the cathode follower stage.
switching it off, removing the tube makes no difference.

The tubes are balanced within a couple mv, and I tried different ones.

Increasing the presence increases the ghosting/oscillation.

Signal hitting output tube grids is nice and clean, after OT it looks like this:

[Helmholtz]

I don't see hum modulation or oscillation.
But there's strong duty cycle shift and distortion.

Do grid signals on both sides look the same?

[drewl]

Hmm,

It's preamp is like a 2203, even without the extra gain stages.

Here's the grids with preamp set cleaner

[Helmholtz]

Here's the grids with preamp set cleaner

Grid signals look bad, What are the peak- to-peak voltages?

Can you feed a sine signal of maybe 2V to the PI input?

[drewl]

Bad in what way?
I was slamming the input, had sig gen set high from testing a power amp.

Getting around 100v p-p max at the output tube grids.

Here's the output with a 2v p-p sig fed to PI.

[drewl]

Just to add, when pushing the power amp, HV drops like 60v and ripple increases from about 15v pp to about 65v p-p

[Helmholtz]

Just to add, when pushing the power amp, HV drops like 60v and ripple increases from about 15v pp to about 65v p-p

It's normal that B+ drops by maybe 10% at large output.
Also normal that ripple increases.

[Helmholtz]

Here's the output with a 2v p-p sig fed to PI.

That's no sine but a triangle. An unusual kind of distortion.
Are the grid signals clean sine?
If not, lower input signal until the output looks like a sine (presence at zero).

What is the max. clean sine output voltage into an 8R or 16R dummy load?

What is the frequency of your signal?