Trainwreck Rocket 120hz Hum

[kendalosborne]

Hello all -

I've got a Trainwreck Rocket DIY build that has some slight 120hz hum. It's really not that bad, but I'm an audio engineer and session guitarist and ideally would like this amp to be as silent as possible. I've gotten a lot of great tips from this forum to help quiet down the noise a LOT already, but I still have some lingering 120hz hum. I also fully accept that this is as quiet as it can be given layout, tubes, transformers, etc.

Here's the best info I can give:

1. It 1:1 copy of the Matt Taylor "Rocket Version A0 Schematic" aside from some tiny voicing tweaks (bright cap values, etc).

2. Yes, it is definitively 120hz. Even still, I have a humdinger on the filaments and 60 is Very quiet.

3. Amp is well built, clean, quality components, wirewound and metal film resistors in important spots, etc. Cap foils to ground, shielded wiring where applicable, and all the other normal tricks are employed.

4. Ground scheme is a 2-star system with power amp grounds separate from preamp grounds. Power amp grounds back by the PT. Preamp grounds on opposite side of amp near tone stack area (and yes I have tested it grounding closer to the input jack - no change in the 120hz hum). I have NO bus bar on the back of the pots. Individual wires go back to the ground lug, and they're all quite short.

5. Filter caps are new F&Ts, all proper values.

6. Voltages are spot on, amp itself sounds fantastic. No complaints there.

7. I have a very well matched set of power tubes and have rolled tubes for V1/V2/V3.

8. This particular amp is quieter with phase inverter grounded to preamp ground. Have tested this both ways.

9. Amp is virtually SILENT with gain pot down. Noise reading in my DAW is below -90dB. Crazy quiet. I really only hear the hiss and hum when the gain pot is turned up.

10. With this in mind, the only thing going on in V1 is: isolated input jack (cliff), wirewound grid leak, wirewound gridstopper straight to V1 socket, cathode resistor+bypass cap on the board, metal film plate resistor on the board. Standard coupling cap to the gain pot. Input jack and Cathode of V1 are connected together and those connect to the preamp ground. Gain pot grounds to V2 cathode, which goes back to preamp ground. Nothing wacky really?


I've done about as much as I know to do. Spent many, many hours on it. Again, I realize this may be as good as it gets and if that's the case, I'm still happy. I'm just trying to see if I'm missing something, or if any of you could help me think outside the box, or give me suggestions for things to consider at this point to help reduce the 120hz hum.

Any ideas or things to try are welcome.

Thanks,
Kendal

[Latole]

Did the builder use some shielded wires at the input and volumes control pots ?

Show us good picture of the wiring

[jamaio]

3. Amp is well built, clean, quality components, wirewound and metal film resistors, etc. Cap foils to ground, shielded wiring where applicable, and all the other normal tricks are employed.

Cap Outer Foil

The banded end (outer foil) of the cap should always go to the lower impedance side of the circuit. If it is a coupling cap coming off the plate of a tube then the banded end goes towards the plate (band is at + potential). If it is a cathode bypass cap the band goes to ground (band is at - potential). If is a PI input cap the band goes away from the grid of the PI tube (band is at - potential).

https://www.aikenamps.com/index.php/where-to-connect-the-outside-foil-on-capacitors

[uki]

Hello and welcome to the forum!

If you post some high resolution pictures that can help a lot.

Did the builder use some shielded wires at the input and volumes control pots ?

Show us good picture of the wiring

This ! Shielded cables from pots to board, any signal wire that is close to high voltage wires, is a good idea to be a shielded cable.

[HotBluePlates]

... I've got a Trainwreck Rocket DIY build that has some slight 120hz hum. ... It 1:1 copy of the Matt Taylor "Rocket Version A0 Schematic" ... Amp is virtually SILENT with gain pot down. Noise reading in my DAW is below -90dB. Crazy quiet. I really only hear the hiss and hum when the gain pot is turned up. ...

Post the schematic you used, or a link to it.

I found a Matt Taylor Rocket schematic, but there is a Volume pot not a Gain pot.  So I don't know if I'm looking at what you used.

[kendalosborne]

This is the schematic I used. I just called it a gain pot because it makes more sense to me.

Like I said, this is basically identical to the schematic aside from just a few voicing changes:
1. There's a 3-way bright switch instead of a fixed bright cap.
2. There's a mid boost toggle that changes the value of the fixed mid resistor.
3. There's a resistor on the bass pot to alter the taper / value of that pot.
4. The treble cap in the tone stack is a 68pF but it actually reads more like a 62pF, and I liked the sound of it, so I left it in.

Yes there is good shielded coax on the input and Gain(Vol) pot to V2. The shielded coax going to V2 goes under the board, and I have considered moving that. The wiring is all very short in this amp.

You can see the power amp ground back by the PT and the preamp ground by the edge of the board. This chassis has a flanged edge so unfortunately the photos don't show the pots in great detail.

I've attached some photos I took just before I swapped out the filter caps. There's a 22uF tacked in on the board, but that is no longer there. One of those Tech-Caps was bad, so I just replaced all 3 and then removed the tacked-in 22uF on the board. Nothing else has changed.

[kendalosborne]

Some other photos

[kendalosborne]

Oh yes - a few other tiny changes I just remembered:

1. V1 plate resistor is a 100k rather than the stock 220k.
2. There's a voltage divider added after the tone stack to cut a little gain before the PI.

BOTH of these mods reduced gain a bit, so I don't think they'd be contributing to noise in any way.

You may also see that there's a few resistors/caps on the far right (input side) of the board - these are no longer being used. Originally The amp had a paralleled V1, but that has been removed. The parts were left on there in case I changed my mind.

[kendalosborne]

3. Amp is well built, clean, quality components, wirewound and metal film resistors, etc. Cap foils to ground, shielded wiring where applicable, and all the other normal tricks are employed.

Cap Outer Foil

The banded end (outer foil) of the cap should always go to the lower impedance side of the circuit. If it is a coupling cap coming off the plate of a tube then the banded end goes towards the plate (band is at + potential). If it is a cathode bypass cap the band goes to ground (band is at - potential). If is a PI input cap the band goes away from the grid of the PI tube (band is at - potential).

https://www.aikenamps.com/index.php/where-to-connect-the-outside-foil-on-capacitors

Yes that’s what I meant. Thanks for the correction

[kendalosborne]

Bump

Any ideas for me to try? Still dealing with a bit of this 120hz hum.

This week I went through about 100 tubes (literally, not exaggerating) trying to find the quietest ones. It was…frustrating and illuminating. Pretty remarkable how many tubes have a noise trade off… low hum BUT terrible microphonics, or the other way around.

It helped. I found a few nice and quiet ones. But good lord. Surely there’s something else that I can do to quiet down some of the 120. 

[Latole]

Some problems are very difficult to fix without having the amp in front of you.

[Williamblake]

You already said you have raised heaters, didn`t you?

[pdf64]

Philips info notes that triode section 6,7,8 has the best hum performance.
eg see bottom of p1 https://tubedata.altanatubes.com.br/sheets/010/e/ECC83.pdf
But that will be in regard of mains freq hum.
How about if the negative lead of C18 is connected directly to that of C1, with any existing connection between C18 and 0V common disconnected?

[PRR]

Is that first cap soldered to the chassis?

[kendalosborne]

Is that first cap soldered to the chassis?

No, Good eye though! I believe that's just a mounting bolt. The green wire with the 3 black lines is the ground line for the reservoir cap that grounds back by the power ground.

[tubeswell]

4. Ground scheme is a 2-star system with power amp grounds separate from preamp grounds. Power amp grounds back by the PT. Preamp grounds on opposite side of amp near tone stack area (and yes I have tested it grounding closer to the input jack - no change in the 120hz hum). I have NO bus bar on the back of the pots. Individual wires go back to the ground lug, and they're all quite short.

I'd go to a galactic ground all on a single buss bar with only 1 connection point to the chassis - see layout. This is the quietest type of grounding I have used in my amp builds and it's a vast improvement over split grounding when you want a quiet amp for studio recording

[kendalosborne]

4. Ground scheme is a 2-star system with power amp grounds separate from preamp grounds. Power amp grounds back by the PT. Preamp grounds on opposite side of amp near tone stack area (and yes I have tested it grounding closer to the input jack - no change in the 120hz hum). I have NO bus bar on the back of the pots. Individual wires go back to the ground lug, and they're all quite short.

I'd go to a galactic ground all on a single buss bar with only 1 connection point to the chassis - see layout. This is the quietest type of grounding I have used in my amp builds and it's a vast improvement over split grounding when you want a quiet amp for studio recording

Thank you! Will try this out.

[scstill]

What about the idea that heaters should not have any loops around the socket? See the photo in this article. Are all the same heaters pins connected to each other? 9 to 9 to 9. 4/5 to 4/5 to 4/5. Etc

http://www.valvewizard.co.uk/heater.html

[pdf64]

What about the idea that heaters should not have any loops around the socket? See the photo in this article. Are all the same heaters pins connected to each other? 9 to 9 to 9. 4/5 to 4/5 to 4/5. Etc

http://www.valvewizard.co.uk/heater.html

Heaters work at mains frequency, 60Hz in this case, so hum induced from that circuit into the signal path would also be at 60Hz. Hence it seems somewhat off topic?
But as an general idea, yes, I think it's good to avoid the wiring loops Merlin describes.
I think that Merlin's idea is improved if the socket is rotated 180degrees, as that provides a gap between terminals 1 and 9 for the wires to pass through.
eg

The wiring need to be arranged to provide 6.3VAC between socket terminals 9 and 4&5.