New Bassman+ build with issues

[tubeswell]

Okay I finished this hybrid monster this evening and fired it up and found it making a random staticky hiss with odd popping and crackling noises, even with all the volume controls and tone controls down. In fact, when I pulled all the pre-amp tubes out there was still a very faint similar hiss that immediately got noisier when I plugged the PI tube back in. I swapped pre-amp tubes around but to no avail. The 6L6GCs were good in another amp (no issues). I noted the voltages in the attached schematic.

I adjusted the output tubes to the following parameters:

Plates - about 430

Screens - between 429 and 431

Grids - about -33V

Tube current - 46.6mA in one tube and 46mA in another.

Figuring the idle dissipation at about 20W, so they can't be over-biased.

I note that when I hook my meter up to the treble pot wiper side of the .22uF coupling cap to the PI, the hissing noise gets a lot quieter, but it is still there (albeit more faintly).

Any ideas anyone?

[MWJB]

The tubes are UNDER biased (too hot, too much current) in a bassman style amp idle disipations of 12-15W are typical, look at 17.5W as the higher end. In 99% of amps 30-35mA per tube will suffice.

If the noise is still there with the preamp & PI tubes pulled, then look at the power amp. How long are the power tube grid wires, are they shielded?

2x1n4007 looks a bit wimpy for that amp, use 4 or 6 1n4007.

[tubeswell]

Thanks MWJB. I thought that the current-dissipation aspect was related to plate voltage, so I figured that with a plate voltage of ~430 that 46mA would give me about 20W, which is 66% of the 30W rating of these tubes (JJ 6L6GC) but I'll go back and re-bias to 35mA.

However that won't affect the problem because the problem was still there even when I first powered up when I had the bias set really high anyway as a precaution.

The problem is that the amp is making a hissing sound much like the sound an amp makes when it is on full blast with nothing plugged in - but it makes this sound even when the vol controls are completely cut, and the vol makes no difference to the hiss. (Only the hissing isn't your normal quiet hiss - like when the amp has nothing plugged in and you dime the vol controls - rather, the hissing is sort of white noise like that, but at full vol, and is also accompanied by random electrical popping and crackling noises).

I wonder whether it is coming from somewhere down the signal path because I notice that when I brush my fingers along the chassis, the sound ends up amplified in the speaker, as if I have some signal grid wire hooked up to the chassis. I am going back checking all the wiring, and although I was following my layout and schematic developed in conjunction with feedback here, I must've miswired something, but I'm a bit stumped as to what it might be, because every time I recheck it, it seems to stack up as intended.

I don't know if it is related to the output tube grid wires - they are unshielded at the moment and are about 6 inches long, but that isn't any different to any other amps I've made and they didn't have this problem.

I forgot to mention that the heaters are elevated at about 40V. Seeing as how the cathode voltages on the CF stages are quite high (212 on the trem CF stage, and 147 on the DC pair driving the tonestack, and 140 on the reverb bypass CF stage) - do you think 40V elevation is enough? (or should I go to 60V or 80V?) (What is the max cathode-filament voltage in 12A_7 tubes? - 180V?)

[MWJB]

The JJ's will take the higher current, but 5F6A styleamps are typically not biased that hot, if you like it stick with it, but don't try running any other current production 6L6 that hot.

Picking up sound from the chassis sounds like grounding/layout issues.

Out of interest why elevate the heaters? Shouldn't be necessary.

Don't worry about cathode to filament voltages, plenty of similar designs run at over 500v with no issuies in this department.

[tubeswell]

Okay Thanks MWJB.

Tonight I 'upgraded' the SS rectifier to 6 x 1N4007 with 6 x snubber caps and I shielded the output tube signal grid cable from the PI coupling caps and I rebiased the output tubes to 35mA @ 445 on the plates. The funny noises have gone, and it is just the hiss present...

But the hiss is there whether I dime the vol controls or not. Tonight I tried plugging a guitar in and confirmed it for sure - the vol controls and the reverb don't work. The vol is permanently on. The reverb recovery works (but not the driver) and the tone stack works and so does the presence control, and the tremolo (altho' the latter is not as strong/deep as I'd like it, I'll have to tweak the LFO circuit, maybe take the .022uF from the plate to the cathode out? - it is a stock 6G16 trem circuit.)

The thing is its plugged into the input, which goes through V1 and through the vol pots and through the reverb/bypass circuit to get to the DC pair (so try figuring that one out ). Maybe (incredibly) both vol pots are haddit? The coupling caps off V1 go straight to the pot inputs and the wipers come off that and go straight to the 270k mixing resistors, which go straight to a junction for either a (500pF) coupling cap - to the reverb dwell control and driver; or to a (.047 uF) coupling cap - which goes straight to the reverb bypass CF stage grid. The signal either supposedly goes though the pan and recovery stage with tone control (if wet), or is bypassed through the CF stage. There it is controlled by a 250kN mixer which goes to the grid of the DC pair. I have to think about this deeply. If I bump the pan, I hear a crash in the amp, so the recovery stage is evidently working, and if I swing the mixer to the dry side, I don't hear the pan when I bump it, so the mixer is working.

The RT has DC resistance on both sides (about 1K on the primary and about an Ohm on the secondary.)

All the heaters light up. Hmmm (I wonder if it is the grounds for the vol pots and the dwell pot too maybe? - I'll check that tomorrow as well as looking at other reverb driver tubes)

Anyway its late for now and I've got to stop for tonight.

[MWJB]

Vol pots are grounded with a good ground reference?

[tubeswell]

Getting a little further into this, I found out that the Normal Vol channel control works as intended to cut the volume. The bright vol control doesn't work. So I tested the pot with a DMM and it didn't seem to be malfunctioning but I replaced the pot (and cap) with brand new stuff (that I tested as well) anyway but that didn't fix the problem. I checked and reflowed the vol pot grounds with my big (80w) iron (and the pre-amp common ground connection point) and all test okay on the DMM.

Here is a diagramatic of the the layout I followed if anyone has any bright ideas. The green cables are shielded, the red cables are the DC power supply

The cathodes of the output tubes have 1R resistors going to probe sockets (not shown)

[MWJB]

The diagram looks OK in respect to the bright vol., trouble is the diagram might be fine and your problem is with the actual amp - the diagram is what you intended to do, the amp is actually what you have done, so a pic of the amp would be better.

I assume you have tried another 12AY7?

[txstrat]

The diagram looks OK in respect to the bright vol., trouble is the diagram might be fine and your problem is with the actual amp - the diagram is what you intended to do, the amp is actually what you have done, so a pic of the amp would be better.

I agree, +1 on the ground issue. The DMM on the treble wiper might be a temporarily ground reference thus decreasing the noise. I'd have a look at all grounds around this part of the circuit and check if they're OK.

Additionally you could tie the bright channels grid to ground (that'll leave only the normal channel be a possible source of noise) and see if the noise decreases.

[tubeswell]

Pics

I may have to upload more than 1 reply of pics.

When I disconnected the 500pF coupling cap from the mixing resistors to the dwell pot input, the Bright Vol pot worked. Don't ask me why. The normal channel pot still worked. (However the input signal path to the reverb driver circuit was at this stage disconnected). Do I really need a 500pF cap here? (because to my mind there is no other source of DC going into the dwell control pot, so surely the 500pF here must be superfluous)

However, the overall hiss was still there through the speakers, even with both pots cut. So I've got more than one issue.

Okay so when I put the 500pF cap back in, the bright vol pot didn't work again. (Go figure that one out. What is happening here BTW? Is this having some kind of effect on the output impedance of the V1 stage? or the input impedance of the reverb driver stage? Surely not - because why would the other vol pot still be working anyway? I can't understand this. Is there not good enough impedance bridging between the vol pot mixers and the reverb driver??? The dwell pot is 1M, giving a high input impedance for the reverb driver, or is there more to it than that? I have tried 3 different 12AT7s in the reverb driver slot. Or is it also something to do with the bright vol pot having a bypass cap?). Anyhow, when I test the ground-return wires from all these pots back to the pre-amp grounding point, they all seem to have connectivity with my DMM.

So back to the overall hiss still being there - it was even when I disconnected the coupling caps from the PI plates to the output tube grids. So I figured that maybe this is happening in the power amp somehow. (Could this be to do where I have positioned the 27k GFB resistor on the board between the LTP tail and the LTP-P.A. coupling cap sections? Does this resistor typically emit a lot of EMI? Is that why Fender put it on the other side of the PI-power amp decoupling caps?

[tubeswell]

more pics

Further to the previous post...

[txstrat]

Since the hiss is present with only the power tubes in it might be the bias circuit too close to the power tube grid caps (the .1 uF).
You could try to pull the power tube whose cap is right next to the bias circuit and see if the noise changes. Or try to shield the bias from the cap with a piece of tin soldered to ground.

[tubeswell]

Hi txstrat

That did occur to me and at one stage yesterday I did have both 'tail ends' of the PI-to-output tube coupling caps disconnected - and the hiss was still there. So I think if it was the proximity of the bias circuit, that would've stopped the hiss but it didn't. I will try shifting the locality of the GFB resistor later today.

[txstrat]

TW
I've had another look at your pics and noticed the two 220k carbon comp resistors of the bias supply to the grids of the power tubes. Since CC resistors are prone to hiss you might try carbon film or metal film instead and see if it changes something.
Since this is one of the last possibilities in the line up to the tubes themselves it might be worth a try.

[loudthud]

Have you built amps before with the same type white board? Old Fenders suffer from the dreaded "tweed disease" caused by moisture absorbed by the eyelet board. They make a "frying egg" noise.

The other thing I wonder about is the polarity of the feedback. I'd disconnect the feedback until the amp at least passes a signal.

No grid stoppers on the 6L6s? Add them!