Power Tube squealing in Peavey 6505 head

[nevetslab]

In tracking down the cause of loud hum in a Peavey 6505 Guitar Head, I first found the input tube V1 appeared to be the culprit, having sequentially gone thru all five tubes, from Driver tube on back to the front end. But, then thinking I had found the cause, when I put the chassis back into the cabinet and powered it up again to check (was sounding tame still on the test stand), it was now humming loudly when I dialed the Rhythm Ch Pre-Gain control up, with the Post-Gain control turned up full CW. I removed it and returned the chassis upside down onto my test stand, and, tapping on the power tubes, I discovered the age-old problem of solder joint fractures on their PCB assy. Discharged the supplies, then de-soldered and re-soldered the plate, screen, cathode & grid terminals of the four tubes. The heater connections looked ok. Powered back up, then, with speaker attached, I went thru both channels' Post Gain controls fully CW, and turning up the Lead Ch Pre-Gain control, when I got my hand near the upper half power tubes, sensitivity to the presence of my hand next to the outside tube got it to feeding back. Rubbing my finger on the top of either tube on the upper half of the output stage sent the amp into squealling, while doing the same on the lower half power tubes, it decreased the sensitivity to touch/hand presence nearby.

I know there's so bloody much overeall gain in this amp from all of the cascaded gain stages, but this is the first time I had found this apparent capacitance-coupling to my hand next to the power tubes, let alone rubbing on them having such an effect. The tubes weren't yet so hot that I couldn't touch them.

I pulled this set of Ruby matched 6L6GC tubes, and installed a new set of J/J 6L6GC matched tubes. Same problem. So, it wasn't the power tubes per se. I began removing the J/J tubes, and upon removing the last Ruby tube from the box that contained the J/J tubes, that last tube came out with the upper half of the glass bottle broken off! So much for restoring the original tubes.

I then went back thru the preamp tubes, starting with swapping out the driver tube, then the preceeding stages, with no change in this behavior, and left them as I had before, just changing out V1 at the start of this task.

I didn't have Peavey's exact 6505 Schematic...thought I had a copy, but might be in my computer down the street, currently inaccessible due to the building rented out to a client. I was working off of an EVH/Peavey 5150, which mechanically did appear like the 6505 I had on the bench.

Has anyone come across this odd characteristic, being extremely sensitive to proximity to the final tubes in the gain stage? The upper tubes proximity would screech, while the lower half tubes would attenuate, not having any high end sensitivity like the upper half does.

Peavey_5150_(6505)_(EVH_120,_2-92)_Schematics.pdf

[Enzo]

5150 and 6505 are identical. Only difference is the label. Likewise the 5150-2 and 6505+ are identical.

Eddie Van Halen owns the 5150 trademark, and when his agreement with Peavey ran its course, he left and took the name with him, so Peavey changed it to 6505. It was 2005 at the time, and 1965 was the beginning for Peavey. hence 6505.

In answer, no I never saw that particular symptom. I did however always find those Ruby 6L6GCMSTR (or whatever the exact alphabet soup is) worked well and reliably in the 5150.

[DFuller]

This is gonna be a long shot, but sometimes when I get oscillations like that it's from bad filter caps. Scope the plate connection, I'd bet there's a sawtooth waveform present.

[Pedro Vecino]

I haven't seen such a bad effect but I have seen hum issues on the 5150s.
On the 5150 I and II (6505/6505 +), the negative pole of the electrolytic capacitor closest to the guitar input is connected to the chassis through a screw. If it don't have a good mechanical connection it generates hum. Also, the metal case that houses the preamp tubes doesn't appear to be connected directly to the chassis. I take advantage of one of the fixing screws on the board to put a metal connector and take it to the negative pole of the capacitor.
Maybe they could be related in some way.

[Chuck H]

Eddie Van Halen owns the 5150 trademark,..

Well, owned, I guess. I suppose it's owned by his "estate" now.

[Enzo]

True dat, but he wasn't quite dead in 2005.

[nevetslab]

I haven't seen such a bad effect but I have seen hum issues on the 5150s.
On the 5150 I and II (6505/6505 +), the negative pole of the electrolytic capacitor closest to the guitar input is connected to the chassis through a screw. If it don't have a good mechanical connection it generates hum. Also, the metal case that houses the preamp tubes doesn't appear to be connected directly to the chassis. I take advantage of one of the fixing screws on the board to put a metal connector and take it to the negative pole of the capacitor.
Maybe they could be related in some way.

Good suggestion. I'll pull the chassis back out to have a look at that, and revise the grounding accordingly. I'll also have a look at the filters feeding the O/T C/T as well as the screens. Very odd characteristic to be able to 'play' the amplifier by how close you place your hands to the power tubes. No doubt can only do this when you have that much overall gain in the system. I've never experienced this on the traditional Fender amps and the like. I never noticed if it happened on the other EVH 5150's I've serviced. Probably because I didn't have solder joint fractures on the power tube PCB's on those. Odd one. Once I had this back together, it was just noisy from all the gain when cranked up fully in the Lead Ch. Wonder if this happens on the Mesa Dual and Triple Rectifier model heads?

[Pedro Vecino]

The typical maximum gain setting on the 5150s (lead channel) is usually around 5 or 5.5. It´s an overdrive level similar to that of the second/third channel of a Dual Rectifier in modern mode to 9.
Using gain levels in tests above that reference can be confusing and even disconcerting with noise levels.

[Helmholtz]

The upper tubes proximity would screech, while the lower half tubes would attenuate, not having any high end sensitivity like the upper half does.

As upper and lower half tubes are opposite phase, proximity causes positive feedback only on one side and negative feedback on the other.

Your hand/body picks up a signal from the power tube plate (by capacitance between) and some sensitive circuit point picks up your body signal. So you have a feedback loop. Capacitance depends on proximity and determines squeal frequency.
The effect should stop when you ground yourself.

Do you get the effect with the amp in its box?

[nevetslab]

Your explanation makes total sense, and was what I was thinking too. With the chassis on the service cradle, grounding myself killed the capacitance coupling to my hand. Now, with it back inside the cabinet, turning the Lead Ch Pre and Post gains all the way up, it will feed back. I had the GK 410RBH8 bass cabinet connected, set on a dolly, so I could roll it around, with a 10ft speaker cable. Cabinet position/orientation was involved in the feedback pitch, so as expected...there's enough gain in these amps to get anyone into trouble.

When I had the chassis back on the service cradle this morning, I added a ground wire between J37 Input Jack Normal (connects to Chassis Front Panel Ground) and an added ground lug where C9 capacitor is connected (goes to Chassis ground directly). I also added two tiny solder lugs to the #5 preamp tube PCB mounting screws to the shock-mounted isolated frame, tying both frames together and then to the Chassis ground lug at C9, as seen in the photos below.

I also connected my X10 scope probe to both sides of R210 on the Power Tube PCB (one side is O/T Plate supply, other side is Screen Supply. Plate supply is 470VDC with 4.8V P-P ripple, while the Screen Supply is 466VDC with 0.3V P-P ripple, noticeably filtered by the second stage PS filter cap. The scope screen shot was dancing vertically, hence the partial double image of the ripple.

[Helmholtz]

The scope screen shot was dancing vertically, hence the partial double image of the ripple.

That's strange. Vertical dancing means the presence of a lower frequency signal than displayed (most probably 60Hz). Where did you connect the probe's ground wire?
If you don't ground the probe head to the amp, the potential difference between scope ground and amp ground will add to the signal.

[nevetslab]

Ground lead was clipped onto Neg lead of C24. Scope sensitivity was increased from 2V/Div for the O/T Plate C/T ripple measurement to 0.5V/Div for the Screen measurement. C24 appears to be a filter for the +22VDC source attached to relay coils K1C, K2C, K3C and K4C. Schematic shows a Chassis Gnd symbol at that lead, though I don't see on the PCB foil pattern where that ground makes it to chassis. I see it travel along the edge of the PCB to the right, and makes it to the main filter caps of the amp. Both the scope gnd lead and DMM Blk lead are attached at that C24 Gnd lead.

Scope mainframe (Tek TM515) is plugged into same power strip (3 wire gnd) as the Variac/power Analyzer is, with the Amp plugged into the Variac, so all are sharing the same AC Mains ground in this instance. I didn't stop to troubleshoot the cause of the apparent vertical modulation of the second stage filter cap.

Amp is back together and back in the Guitar Dept.

[Helmholtz]

Ground lead was clipped onto Neg lead of C24. Scope sensitivity was increased from 2V/Div for the O/T Plate C/T ripple measurement to 0.5V/Div for the Screen measurement. C24 appears to be a filter for the +22VDC source attached to relay coils K1C, K2C, K3C and K4C. Schematic shows a Chassis Gnd symbol at that lead, though I don't see on the PCB foil pattern where that ground makes it to chassis. I see it travel along the edge of the PCB to the right, and makes it to the main filter caps of the amp. Both the scope gnd lead and DMM Blk lead are attached at that C24 Gnd lead.

Makes perfect sense, thanks.

Ideal probe grounding point would be the HT filter ground. But as the C24 ground carries the same ground symbol, there should not be any potential difference.
Makes me wonder about the grounding quality in this amp.

The only other reason for the vertical dancing I could think of, would be a faint sign of beginning motorboating/LF oscillation.

[nevetslab]

The HT Filters are, of course, the large Radial Lead caps next to the array of Rectifiers, so no access to their gnd leads to clip onto. I didn't find any evidence of motorboating or LF oscillation connected to dummy load, but, as I had stated previously, with the speaker cabinet connected, at full gain (Pre & Post Gain controls full CW), it will feed back. We have two more of these 6505's in our inventory, with my records on those never having been serviced. This one was also without prior service, so, the game isn't over on these amps.

[Enzo]

The 5150 has multiple grounds, and they are bound together at different points in the amp.