AC15HW1X Loud Hum - Help!

[ryanCBFX]

Hey all. I'm really hoping some more experienced folks than I can help me troubleshoot this thing to figure out what's wrong. I'm trying things as I think of them or learn more, but it's getting aggravating. Some background:

This is an AC15 I have heavily modded in that I replaced literally every component on the board with higher quality ones, so they are all only a few months old. Circuit tweaks are all pretty standard stuff, bright switch chooses between two coupling caps, removed the 330k resistor that limits gain, handful of different values, and LarMar PPIMV to replace the crossline. Pretty much just making it more true to the original AC circuits. I'm confident in the quality of my work and have checked solder joints and beeped connections to make sure everything lines up. Pic of the board after almost all mods below. A few weeks ago I was playing and it started making a loud hum so I turned it off. Life got in the way and a few weeks went by.

Started using it again and it was fine for days. Then the hum started again, and has since become constant. Here are things I have observed about the amp:

1. Hum comes through speaker quietly as amp warms up, then gets very loud right away. It REALLY sounds to me like 120hz (I have compared it with generated tones). There may be some 60 in there but if there is it's drowned out pretty heavily

2. Hum exists with NO preamp tubes. Obviously goes away with no power tubes. The master volume does have some impact on the level of the hum, but not much.

3. All of the grounds really do seem fine. I didn't change the grounding scheme. And if it was a ground loop because of a layout flaw I imagine the noise wouldn't have come on the way it did but would have happened right away

4. Two spots where I did no mods had factory issues. One of the preamp tube socket heater wires seriously had a lead sticking off of the pin that was long enough to contact the chassis and on the impedance selector switch there was a glob of solder on one of the terminals that contacted the shell of the switch I discovered the latter while poking around in that area with my chopstick after hearing the sound cut out. This can't possibly have been good for the OT (see my current conclusion at the bottom)

5. I have rolled tubes and used known good tubes

6. The output transformer primary wires are tightly twisted and run alongside the chassis. Moving them towards the board makes the noise WAY louder, so this is seems to be where the noise is getting in

7. The OT primaries are fairly off in resistance, 110ohms on one side, 130ohms on the other. Because of this the tubes are idling at 925 and 81% plate dissipation. I imagine a lack of common mode cancellation is at least partially responsible for the noise

8. There are definitely some microphonic areas on the board specifically around the preamp stage coupling caps leading to the PI

9. Some voltages seem a bit off, the heaters are running around 5.7v AC for example. Plate to cathode voltage is right around 283 for both tubes but I'm not sure if that's suspicious or not. I can provide more, but I need to get in there and write everything down

10. I've temped in a new, unused MOD cap in place of the B+ and HT filter caps, this did nothing to help. All the caps test fine on my ESR meter (including the cathode bypass cap which is a Vishay), and again, they're all basically new. I haven't temped the cap in for the preamp filter caps because of the noise being present with no preamp tubes

Current conclusion: I'm starting to suspect the OT may be going south. Honestly I wouldn't be entirely surprised given the quality of them and the issue noted above in #4. That said, I'm not confident enough to pull the trigger on a MM Tone Clone. Any advice would be appreciated greatly! I'm really trying here, I just don't have a lot of experience troubleshooting amps at this point.

[maxkracht]

on the impedance selector switch there was a glob of solder on one of the terminals that contacted the shell of the switch I discovered the latter while poking around in that area with my chopstick after hearing the sound cut out. This can't possibly have been good for the OT

OT secondaries shorting to ground isn't a problem. Most amps set up so the output jack will short to ground when there is no speaker installed to protect the OT from a no load situation. SS amps are the opposite.

Few percentage difference between output tubes is also not a big deal.

More photos and a layout might be helpful. If it was working for a while without the hum and started to fail after playing a while, I would suspect a bad solder joint somewhere or another bad connection. Just guesses. Chopstick is your friend.

[ryanCBFX]

Thanks. I had to clear my bench off to do a guitar setup and probably won't have much time to get back to it until Sunday, but I'll definitely go through it, I plan on pulling the board to be sure, and I'll probably just re-flow pretty much everything to make sure it's all solid. It all seemed to beep out fine, but I know that doesn't always mean everything is ideal. I'll grab more pics when I do that, and assuming after I'm done the issue still exists I'll see if I can get a video up here.

[Stevem]

You need to go Thru a process of elimination so you don’t drive yourself up a wall.

First narrow down if this hum is 60 hz or 120hz.

Listen to the power transformer thru a paper towel tub, is this the hum frequency you hear, or is it twice as high.

120hz being heard means it’s after the rectifier(s) .

99% of the time this is due to a poor B+ center tap ground, or a poor filter ground.

Removing the preamp tubes from the output tubes on back up stream to the first preamp tube will narrow down where this fault is.

I have had many a 120 hz hum condition happen especially in fender combo amps where thru process of elimination the first power supply filter feeding the output transformer was bad, yet it showed no physical signs of being so.

Also in terms of 120 hz hum you may have a output tube that is red plating, either due to bad socket connections or just one of the tubes being bad yet not blowing a fuse.

Also in terms of tracking down a 120hz hum issue you can set your voltmeter to AC and take some reading looking at the ripple voltage.

Also if your power supply filter grounds are all just daisy chained that will be a source of a hum level that’s higher then it should be .

There should be at least 3 star grounds wire runs in any amp with output tubes .

The B+ center tap lead off the PT should go right to the first power supply filter and then a heavy ground wire should go from there back to being bolted or soldered onto the chassis near the Pt.

The next ground for the output tube screen filter should have its own wire that then gets stacked on that first chassis ground at the PT.

However many stages of preamp filters you have after that can then most times be daisy chained into one wire run that once again gets stacked on the other two.

This works because the preamp stages you so little current while the output tube section uses a lot more.

Note that although rare, I have had new filters being bad.

[ryanCBFX]

Appreciate the replies. Took me longer than I wanted to get around to it, but I overhauled the thing to be sure. I wanted to track down a specific bad joint or something, but with all the factory wires run on the bottom of the turrets, I just assumed there was too much lead free solder somewhere making a bad connection, so I completely desoldered, cleaned and re-soldered everything on the turret board, the output jack section, and cleaned up a bunch of the wire runs. I put it all back together, and it's quieter now. I'm very happy to accept there could have been one or more bad joints. But there's more to it because it's not 100%

Basically, when I start it up, it doesn't immediately start to hum extremely loudly like it was doing as soon as it warms up. It does pass signal on both channels and the controls all work properly so I'm confident the preamp is fine, though obviously it isn't sounding great. There is some sort of power draw going on and B+ drops almost 100V (~230V). If I remove the power tubes, I can run it with all the preamp tubes and the rectifier and everything seems fine. B+ around 330V, which is healthy for one of these. I've tried multiple known-good sets of power tubes, same result. Just to note, I am running the amp on a bulb tester to make sure I don't melt anything. I wonder if what I was hearing before was caused by even greater power draw because I wasn't using the current limiter. Hesitant to repeat it. I had the old B+ written down, and this is going to sound cliche and awful, but it was on a sticky note on my desk and my kittens think those are food Anyway, iirc it was in the 200s which is definitely not right.

Anyway, I've traced out the power section and I can't see a problem with the way it's wired. The components all seem to be fine, and I had temped in other filter caps to make sure that wasn't it. I'm kind of back to the output transformer since I know how flimsy they are. Basing this largely on that with no power tubes it's fine, and the missing link there is the B+ to the OT center tap, and then obviously all the OT secondaries. It's cathode biased, so there's no DC being fed to the grids, and obviously the coupling caps stop all the PI DC. So the only two places I could see causing the problem are the OT primary or secondary. Certainly curious if anyone has any more advice, I'd appreciate it! I'm ready to spring for a nice mercury or something (a nice upgrade anyway), I'd just really like to be sure.

Cheers,
Ryan

[Stevem]

When you get hum that is not there at first you likely have a output tube that is starting to red plate.

Unless it's only doing this in a minimum way then viewing the output tubes with the lights off in the room should show it.

Also it may just be one output tube doing this.

To test this once the hum starts up pull out one tube at a time to check if it goes away.

Please report back .

[ryanCBFX]

Thanks! So before I started pulling them I figured I'd just measure the bias. The tubes are running SUPER cold actually, these are the numbers:

V5:
Plate voltage (to cathode): 211.2V
Primary resistance: 130.5 ohms
Voltage drop: 2V
Plate current: 15mA
Plate Dissipation: 3.17W

V4:
Plate voltage (to cathode): 210.4V
Primary resistance: 120 ohms
Voltage drop: 2.76V
Plate current: 25mA
Plate Dissipation: 5.26W

And the B+ was floating right around 220V which is crazy low. I tried a third known good set of EL84s (two of the DJs that came in my Matchless) and got similar results, under 40% plate dissipation. Obviously not right. I can still pull tubes one by one if there's something to be learned, but they definitely aren't red plating.

The OT secondary wires to B+ and the tube plates are still bundled up and run across the front of the chassis like they were, and I think the new lead dress I set up in there has helped reduce the noise, but if I chopstick those up towards the board / tube sockets it still picks up a ton of noise. My thoughts are that if the power tubes aren't what's drawing all the current, and the draw only happens when the power tubes are installed, wouldn't that point to a partial short in the OT primaries? Only other related source of current draw I can think of is the screen supply, but the resistors are good and all the tubes I've used work just fine in my other EL84 amps, including one of the AC15 heritage editions.

Cheers,
Ryan

[maxkracht]

100v difference between power tubes and no power tubes doesn't seem crazy to me. Sounding bad is another issue. If nothing is heating up too much and your tubes are biased cool, it's probably fine to turn it fully on, at least long enough to measure voltages. You can also sub in a higher wattage bulb if that makes you more comfortable. The bulb is meant to drop voltage, so low B+ is expected.

[ryanCBFX]

Running it with a 100 watt bulb. Imagine that should be overkill for an AC15. Anyway, I'm going to start from scratch using wall voltage, measure everything so I don't accidentally blow out any caps (they're all at least 475V, should be fine). If I run into an issue I can use the variac to get it to sensible levels. Will report back with the numbers and what I observe.