Alessandro Deluxe Reverb Reissue Parasitic Oscillation?

[Charlie Wilson]

Hello I have an amp that I am working on that has some kind of reverb oscillation issue. The complaint was that when the reverb level control is turned up it hums more than usual for this type of amplifier. The amp overall sounds a bit too bright and hi fi. In working on the amp I noticed that the power transformer gets hot pretty quickly and the OT gets a little warm also. I am suspecting a parasitic oscillation issue probably in the reverb section of the circuit. I have swapped cables and pans so I ruled that out. I do have an oscilloscope but I rarely if ever have had to track down this sort of problem. Any advice would be appreciated. One thing I noticed is that the lead dress around the reverb tubes is not traditional Fender. The grid wires(green) are up in the air and very sensitive.
Thanks,
CW

[Stevem]

You seem to have 2 different issues here.

The hum if 60hz, then its from the PT and could be just due to having the wrong end of the pan near the PT.

The output side of the pan and its cables need to be as far away from the PT and the wires in the amp that carry AC voltage for minimum 60hz hum.

If it's 120 hz hum then that's from the AC ripple riding on the B+ supply of the amp.

This brings up another point that can make for osccilations, that being the condition of the power node filters as they uncouple one gain stage from effecting the others, and it would not be the first time that poor filtering made for a osscilation issue kicking in.
If you have any new filters around jump out a few of them in the preamp section and see if that effects anything?

If it's osscilating does pulling the reverb driver tube out kill it off?
Your OT running hot may be a issue if it's due to the amp outputting a osscilation signal at a good level that is above the frequency that the speaker can reproduce.
A quick simple test is to just hook up your voltmeter set for AC across the speaker and check to see if the meter reads anything over millivolts , if it does then yes the amp is osccilating.

[pdf64]

When a power supply seems to get hot, a good place to start the investigation is the circuit that draws the most power; what cathode / plate current do the power tubes idle at?
To re-phrase that, have you checked that the amp's bias setting is appropriate for the particular power tubes in use?

[Structo]

I know it's not a Fender but most old Fenders got pretty hot.
Especially combos.

If the switch bat is too hot to touch, you may want to investigate a way to cool the amp more. (fan, ventilation holes, etc.)

[Charlie Wilson]

Thanks for the replies guys. So I checked the bias on the power tubes and it was 37mA . This is weird because the owner told me the tubes have never been replaced and no one has to rebiased the amp since he purchased it. I set them at 20mA and the amp sounds better and the transformers do seem a bit happier. I compared the noise level of the reverb with a stock 65 DRI and it is twice as noisy when the reverb control is turned up. The thing is I don't if it has been this way from day one of the Alessandro rebuild or if it is a problem that developed since then. I have checked all components and they all seem fine. I did tack in a Hammond reverb driver and it made the reverb sound better but did not make it any quieter. An issue the amp also seems to be having when I play through it for a half an hour is that it looses the high end and gets mid rangy and muddy(both channels). I also noticed that the main filter has about 2 volts of AC on it which seems a bit high for a new F&T cap.
CW

[xtian]

Reverb is noisy in what way? Hum? Because I think the reverb transformer gets its B+ from the same node as the screens, and if you've got too much ripple (2V as you said) at the reservoir, that might be part of the problem.

A good way to test this theory is to clip in (really--with alligator clips) an additional e-cap at B+. Clip it in with the amp off and caps discharged--otherwise you'll get a surprising spark. Confirm the orientation.

If hum is reduced, and ripple drops under 1V, then I'd suggest replacing the e-caps, or at least the reservoir caps.

I see that Stevem, in post #2, is suggesting the same thing.

[Charlie Wilson]

Thanks Xtian. Yeah hum. I tacked(too chicken to use alligator clips) a 20uf cap across the 47uf and the AC dropped to less than a volt. This did seem to quiet down the reverb when the control is turned up. Man I hate the sound of these new Accutronic pans. I think the amp overall sounds much better.
CW

[hitchcaster]

that reissue fender iron is probably not the best something to consider.... if he's using silver plated Teflon wire, it's generally pretty brite

[pdf64]

Thanks for the replies guys. So I checked the bias on the power tubes and it was 37mA . This is weird because the owner told me the tubes have never been replaced and no one has to rebiased the amp since he purchased it. I set them at 20mA and the amp sounds better and the transformers do seem a bit happier.

OK, now that's reasonable I suggest you undertake a voltage survey of the tubes, B+ nodes, for future reference.

I compared the noise level of the reverb with a stock 65 DRI and it is twice as noisy when the reverb control is turned up.

Speaker efficiency / tonal balance varies significantly from one model to another; unless both amps are fitted with the same speaker and positioned close together in the room, equidistant from the listener, a comparison between combo amps may not be valid.

Cables with failing screening / ground may be the most frequent cause of a noisy reverb; this includes the cables between amp and tank, and the footswitch cable.
Issues with the footswitch cable can be eliminated by unplugging it (did both amps have their footswitch connected at the time of the comparison?).
Maybe raid your hifi for some decent RCA cables to try between tank and amp.

An issue the amp also seems to be having when I play through it for a half an hour is that it looses the high end and gets mid rangy and muddy(both channels). I also noticed that the main filter has about 2 volts of AC on it which seems a bit high for a new F&T cap.
CW

Check that the speaker magnet isn't getting way hot; consider trying a different speaker.
2Vac (rms, average, peak, or peak to peak?) on the reservoir at idle doesn't seem particularly high to me. Was the measurement made with the power tubes idleing at 37mA or 20mA?

[Structo]

Something I have noticed is when tubes have a lot of hours on them, the high frequencies are the first to suffer.

Also could be a wonky preamp tube that is marginal until it gets real hot, then starts to fade.

[Charlie Wilson]

Thanks again guys. So maybe I made an error in measuring the AC ripple at the main filter but I measured again and adding the piggyback cap actually only dropped it down to 1.5V. The thing is, I measured the ripple on a stock DRRI and it is about 3V so I guess my 1.5V isn't so bad. I started to suspect that the V4 grid wires that are up high and are very sensitive were picking up the noise. So I did a little Dumble style shielded cable on both the V4 grids and I think it did quiet things down. The funny thing is that the amp sounds stronger now. Maybe those sensitive wires were creating an inaudible oscillation. The amp sounded bad before and now sounds pretty good and strong with a reasonable amount of noise when the reverb is turned up so I am calling it done.
CW

[rfgordon]

In a conventional Fender-style layout, the coupling cap from the plate of the reverb recovery stage is on the board, and that results in a fair amount of unshielded grid lead between it and the reverb pot. If you relocate that cap from the board to the tab on the pot, then that previous section of lead (from the board to the cap-on-the-pot) becomes plate lead (low impedance). This can often end parasitic oscillations as well as drop the noise of the reverb circuit.

[pdf64]

In a conventional Fender-style layout, the coupling cap from the plate of the reverb recovery stage is on the board, and that results in a fair amount of unshielded grid lead between it and the reverb pot. If you relocate that cap from the board to the tab on the pot, then that previous section of lead (from the board to the cap-on-the-pot) becomes plate lead (low impedance). This can often end parasitic oscillations as well as drop the noise of the reverb circuit.

I don't agree with the analysis above; a regular RC coupling cap doesn't transform the circuit impedance.
Regarding the BF reverb recovery stage, above the corner frequency (~380Hz) formed by the 3nF wet coupling cap and 100k reverb intensity (+ the ~40k stage source impedance), the impedance either side of the coupling cap is the same.

[xtian]

I agree about the reservoir ripple. I just completed a Tiny Terror build, which came out acceptably hum-free. UNLOADED, the ripple at the 80uF reservoir was only 30mV AC. But in operation, the ripple was 1.6v AC.

[FUCHSAUDIO]

A couple of simple points to note: The reverb driver does not have a cap plate to cathode. Not all Fenders have that, but adding it cannot hurt stability. Sometimes (on peaks) the driver oscillates and sounds like a blown speaker.

Secondly, I cannot see if the .0022 cap is on the reverb return resistor that loads the grid of the return tube. You should have that, as well as adding a similar cap (plate/cathode) as on the reverb driver. Again, a small HF rolloff on the return (inaudible), which can add stability.

I've worked on a few of George's builds and they are nicely executed but sometimes there are odd technical errors or omissions (power tube series grid resistors or screen resistors on UL output transformer taps), and things like this.