Noisy Express

[peesinstew]

Well, yeah, it has been some trouble, but mostly because I was dumb.
My main incentive for switching over was that I don't know where my noise is coming from and I wanted to try to eliminate the heaters as a potential source. Now that I patched everything up and put her back together, she's working, hum-free I might add. It's just too bad that the hiss and squeal are louder than the hum ever was.

I tried a couple more things tonight to troubleshoot the hiss: replaced the cathode caps and the V1 plate resistors without any luck, replaced the volume pot and the treble cap, tried a diode in the B+ line to the preamp (I saw this in another thread about hiss here). I'm waiting on a MOV that I ordered, but really I feel that I'm running out of things to change.

I still get some microphonic oscillations occasionally, but I can stop them now by pressing on certain parts of the chassis, pushing or pulling on the V1 socket, moving a wire or two around, or turning down the volume, treble, or presence. I have all of the wires where they seem to be quietest, but things still can get finicky.

There is hiss when I turn up the mid pot or the bass pot with the treble down, but it's quieter and lower frequency, so it doesn't bother me much. It does make me think that at least some of it is coming from V1a or maybe through the B+ rail, although last week I tried adding a 0.1uF cap from directly at the top of the V1 plate resistor junction and the ground of the V2b cathode, and it didn't help. Also, when all the tone stack knobs are at zero and I sweep the volume from zero to full, the hiss will be quiet and then swell and then go back down when it reaches the end with the grounded tone stack. I guess this means that at least some hiss is coming in from the volume pot, the bright switch, or the shielded V1b grid wire.

well, I'm sleepy. I'm going to bed before I melt something again.

[Firestorm]

Also, when all the tone stack knobs are at zero and I sweep the volume from zero to full, the hiss will be quiet and then swell and then go back down when it reaches the end with the grounded tone stack. I guess this means that at least some hiss is coming in from the volume pot, the bright switch, or the shielded V1b grid wire.

When you set all the tone knobs at zero, you completely load down the preceding stage (so at this point none of the hiss can be coming from V1a). With that setting, both ends of the volume pot are at ground so when you sweep pot, you are lifting the V1b grid above ground and getting noise entering the grid. I think that isolates (most of) the problem to V1b.

Where is the volume pot grounded (I can't tell from the pics)? It it's grounded in a noisy place, you may pick up some noise through the shield on the wire from the volume pot to V1b. On Francesca, the tab on the volume pot is soldered to the pot case just above the ground buss -- very little resistance between those points.

I also noticed that you have the cap on the presence control oriented 90 degrees differently from Francesca (and just about every tweed Fender and Marshall). I'd move it just the way it is on the Francesca pics. Caps can sort of "beam" noise out the ends -- a lot of unstable old Fenders settle down when certain caps are rotated 90 degrees. Just a thought.

[novosibir]

Has anybody already tried an elevated heater supply to reduce the hum?

Larry

[paulster]

Larry

Want to try one of these?

6.3V AC in, 6.3Vdc out. And very quiet!

Paul

[dartanion]

I've done this with cathode biased amps with about 15V of elevation and it works pretty well even at that level. I can imagine that raising them higher would help more.

[peesinstew]

I've been doing alot with this the past couple of days, but I still can't get to the root of this.

The Volume pot was originally grounded via a bus along the back of the pots, then I tried a bus that was only attached to the chassis at the input jack, then I tried grounding the pot grounds to the pot bodies and the board grounds to a bus. The only real noticeable difference I've heard is when I have the chassis open and wave my hand over it or touch things with my DMM leads. When the chassis is shielded, I haven't been ale to hear any difference between the grounding schemes.

I changed the orientation of the presence cap but didn't hear a difference. It does make it much easier to get to the pot lugs though.

besides that, I anted to further isolate things, so I detached V1a from the rest of the circuit and completely removed the tonestack and bright switch components. With DC preamp heaters at 6.4V and 9400uF (2x4700) of filtering, I originally thought that all the hum was gone, but with the amp like this, I realized that I was getting hum at 120Hz instead. It's quiet, but it's there, and it increases with the volume. My DMM says there's about 200mVAC on the positive heater wire.

With the amp in this condition, I still have some hiss, which also increases with volume and presence. Much less than when V1a is included, but still there and still more than I would expect from a normal 2 gain stage amp, and yes I realize that a wreck isn't 'normal' but still...
I think I'm getting noise in every stage to some extent, but since both sides of V1 are amplified so much more, it seems that they're really causing the problem.

I wound up switching back to AC heaters and elevating the center tap to see if that would be any better. At 18V elevation, I do hear hum, but it's very quiet and I can live with it. When I turned the volume up, though, I seem to get more 120Hz buzz than when I had the DC heaters installed (with the DC heaters, it was more of a 120Hz hum than buzz). I don't think I changed anything else in that time except the heater wires themselves.

I ordered some parts the other day, including a hum balance pot, a MOV, and a real bottom plate so I can stop using aluminum foil. So maybe these will help when I get them, but I don't know. Is it possible that I just live in a noisy area and won't be able to get rid of this? There are dimmers throughout my apartment, plus I live right next to the 'L' train tracks. Will a MOV help if I have noisy power coming in? I'm pretty stumped otherwise.

I've basically taken out V1a and replaced or removed every part, component, and wire in the V1b path except for the tube socket, chassis, and the board itself. I've reflowed or desoldered/resoldered almost every joint and I've changed the grounding scheme more than I can remember. Some things have helped a little bit, but there's still more noise than I would like. I'll put all the parts back in tomorrow and see if she turns out any different than before, but based on what's coming out of V1b, I don't have my hopes set too high. I'm not giving up yet, but I'm struggling to think of other things to try.

[Firestorm]

Are you sure the hum is 120 cycles? When you disconnected V1a, did you ground the input to the volume pot (the wire that originally came from the tone stack)? If not, you have a little antenna that will pick up the 60-cycle hum that is ubiquitous -- same thing as an amp where the input shorting jack isn't shorting.

As to the MOV, I don't think they do anything until there's a voltage spike that exceeds the rated level. I believe these are referred to as "sacrificial" devices, because a big spike will take them out. Smaller spikes wear them out over time.

Proper bottom plate? Good. Hum balance pot? I've never seen one that worked as well as two matched 100 ohm resistors (or a DC bias). Plus, you'll have to go back to a pure 6.3VAC heater supply.

As to the noise -- I assume we're talking about excessive hiss (even for a TW) -- I'd suspect the dropping resistors in the power supply. These generate a lot of contact noise that can get into the signal path. It looks like all of your power resistors are metal oxide (including the big cement ones since they are 2K2 (over 1K, these are usually MO rather than wire wound). This is probably the noisiest resistor in your amp because of the amount of current flowing. Get a 1K/25 WIREWOUND resistor for the "choke" position (either cement or those silicone cylindrical ones). Wirewounds are dramatically quieter in high-current positions.

Your 9K1s are also MO. You can get 3.25W 9K1 wirewounds, but these are about $3.50 each, instead get the 3W 9K1 metal film resistors (it looks like this is what is in Francesca, btw). Mouser has them from Vishay/BC for 54 cents each. When you put it all back together, you should also use a metal film 1 Meg resistor on the input jack -- all the noise generated there gets amplified by all the stages. If you want to learn more about resistor noise, check out Randall Aiken's site: http://www.aikenamps.com/ResistorNoise.htm. I learned an awful lot from him.

[novosibir]

The link above doesn't work!

Use this one: http://www.aikenamps.com./ResistorNoise.htm

Larry

[peesinstew]

Are you sure the hum is 120 cycles? When you disconnected V1a, did you ground the input to the volume pot (the wire that originally came from the tone stack)? If not, you have a little antenna that will pick up the 60-cycle hum that is ubiquitous -- same thing as an amp where the input shorting jack isn't shorting.

well, I don't have a frequency meter, but it corresponded to between a Bflat and B on my guitar and it sounded higher than 60Hz.

I listened both ways, with the Volume input grounded or floating. with it grounded and the Volume about halfway up, there was some hiss and a 120Hz hum. With the input floating and the Volume up higher, I could hear the same hiss and hum as well as some other buzzing noises.

As for the resistors, that article is a good read. I've been to Randall Aiken's site before, but each time I go back, I'm able to take in and understand a little more. Can the power resistors make that much of a difference though, despite all the filtering around them?

I do have some other parts that I want to get anyways, so I guess I'll have an order to put together.

By the way, welcome to the garage. I feel honored that your first couple of posts are for my benefit, so thanks for helping me out.

[peesinstew]

And Larry, your builds look great! What parts do you use to make sure your preamp heaters are DC filtered properly? Would I be better off using the 5VAC and 6.3VAC taps together to try to get 12VDC or just using the 6.3VAC taps to convert to 6.3VDC? If I can completely eliminate the heaters as a possible source of noise, I'll gladly do it.

[novosibir]

I'm using custom made transformers and do have on the PT's an additional 13.5V secondary. My DC heaters are regulated 12.6V with a LM317K as the stabilizer transistor.

Larry

[peesinstew]

ok, thanks. I see that the minimum differential input to output voltage is 3V with the LM317. I guess that rules out running at 12.6VDC with my current transformer set (unless I to use a charge pump which I'd prefer not to), but I could still bridge the 5V and 6.3V taps and regulate down to 6.3VDC. If I do this, where's the best place to ground pin 9 of each tube? Should I use the same heater wires as if I were running AC and carry the ground back to the rectifier/filter/regulator ground point?

paulster,
Do you make those boards you posted the pic of? Are they regulated? If so, how do you get 6.3VDC out with 6.3VAC in? If no, how do you make sure it's 6.3VAC?

If I'm asking too much, let me know. I don't want to step on any toes here.

I hate that I still have noise in my amp, but wow, I'm learning alot in the process.

[paulster]

ok, thanks. I see that the minimum differential input to output voltage is 3V with the LM317. I guess that rules out running at 12.6VDC with my current transformer set (unless I to use a charge pump which I'd prefer not to), but I could still bridge the 5V and 6.3V taps and regulate down to 6.3VDC. If I do this, where's the best place to ground pin 9 of each tube? Should I use the same heater wires as if I were running AC and carry the ground back to the rectifier/filter/regulator ground point?

paulster,
Do you make those boards you posted the pic of? Are they regulated? If so, how do you get 6.3VDC out with 6.3VAC in? If no, how do you make sure it's 6.3VAC?

If you're going to use DC preamps then ground pin 9 of one of the preamp tubes (I go for V1) and ensure that you remove the centre-tap connection from ground or you'll short out your DC supply.

I had these boards made for myself using very low dropout regulators so I can happily generate 6.3Vdc regulated from a 6.3Vac supply. 6.3Vac rectified generates 8.9V less the diode forward voltage drops and regulator differential. The combination I use means you could get around 7.5Vdc out.

I hadn't originally planned on offering them for sale, but I'll price up the components tomorrow from UK versus US suppliers to see whether it's economical to just supply a board and a bill of materials if people are interested in one.

[peesinstew]

ensure that you remove the centre-tap connection from ground or you'll short out your DC supply.

oh yeah, I've already made that mistake once. I'm not planning on repeating it.

My SS electronics knowledge is pretty limited (as is my knowledge on tube circuits ) and I still don't know much about what parts do what or what's available. I didn't realize that low dropout regulators existed. I've built a couple pedals and a power supply for them, and I etched my own boards when I made those so I can do that again if I need to, but figuring out what components would be best is where I struggle. So far in the wreck I've just been hanging things off terminal strips to experiment around, but if I do find something that works, I will probably want to do it right.

[Firestorm]

By the way, welcome to the garage. I feel honored that your first couple of posts are for my benefit, so thanks for helping me out.

Thanks for the welcome. One of the few sites where I actually felt like jumping in. Solving these maddening little puzzles is really kind of fun. And it feels so good once you beat one of them into submission.
(And thanks Larry for fixing Aiken's URL).