New Rocket build sounds amazing but needs some hum reduction

[jkey04]

Finished my first rocket build about a week ago and it sounds great! My only concern about it is that once the volume is above 10 o'clock it begins to amplify two distinct tones one is a lowend hum and the other is a highend hiss.


In an attempt to fix the problem, I ran shielded wire in several places in the preamp (on the input and grids) and I have checked and redone my grounding scheme several times. The shielded wire helped most of the hiss but didnt touch the hum. I rotated my OT a few degrees in both directions but couldn't hear a difference.

Can I get some help on how to reduce the hum? I would like to just find out what the source of it is. I will post a pic of the guts when I get home this evening.

[M Fowler]

One problem could be bad mechanical hum from the PT due to poor manufacturing. I had this problem and rebuilt an amp three times only to discover it was an improper built Heyboer PT inwhich Heyboer replaced.

Two it could be AC induced hum from filaments.

Third it could be imbalance from output tubes.

Mark

[jkey04]

Here are a few shots of the inside of my amp. The hum definitly sounds like AC interference, but I can't find the source of it.

[Colossal]

Take a look at the turret the cut cap is soldered to. Is that connection well soldered?

[Bob S]

The resistor into V1A looks like it is touching the filament wire. Hard to tell from the pic. I would put some separation between them. Even though they cross at 90 degrees, that's where the enemy lurks.

[M Fowler]

The pots are grounded to the chassis and then have those solder tabs attached with wires going to the buss wire.

The buss bar is grounded near the cap stack or with the cap stack.

The input jack is grounded but doesn't have a ground wire to the buss wire.

I would think you have several ground loops.

I would also use more solder on those turrets.

Mark

[jkey04]

Thanks for all the replies.

I will check the solder connection on the cut control.

I was worried about the filament wire imposing on V1B input, so i will double check that as well.

Mark--can you further explain what i can do to fix these suspected ground loops. Do i need to run a ground wire for the input jack? RJ suggested via email that I use the ground lugs on the PEC pots, because i couldn't solder to their stainless cases. Do i need to remove them?

The Rocket layout that i downloaded from this forum suggested the grounding location that i used for the buss bar (Yes, it is close to cap stack). The cap stack is not grounded at this location. I know it is hard to tell from the pics i posted.

Thanks for the insight. I will try to get to some of this tomorrow!

[surfsup]

I agree with Mark. the last pot has a tab on it that is soldered to the bussbar. The volume pot looks to have a ground to the turret and to teh chassis using two green wires with the chassis connection also connected to the turret/busbar with a black wire. Your other two pots in between have ground runs to the chassis as well off the bar. That all should be cleaned up.

[M Fowler]

One test you could do is disconnect the buss wire grounding by the caps and solder a wire from the input jack grounding tab to the V1 bias resistor/cap ground turret then try the amp.

Mark

[rooster]

Hm, I would suggest turning the power supply caps 90%, just like the typical build. I personally think this is very important, FWIW.

Also, as Mark says, it could also be the filament supply winding in the PT. This has happened to me with a Heyboer, too. To check this, turn the amp on, leave it in standby. If the PT wasn't potted correctly you will hear a physical buzz, like a badass mosquito, right at the PT. This buzz will haunt your build.

The other suggestions are good, too, and I think you were right to add some shielded wire along the signal path going out to the tubes. This is a very quiet build when done right so you are correct in thinking there is something wrong.

[jkey04]

Thanks again for the comments.

I just had some time to work on this hum problem and discovered that as soon as i flip on the power (with amp still in standby) the PT itself does hum. Then when I turn standby off, that same tone is what is being amplified. So i assume this is a bad PT??

Also, I added some more solder to the turrets, removed the buss connection by the filter caps, and added a ground lead from the input jack to the cathode bias circuit of the first stage. This did seem to help some, but the PT really is humming pretty loud.

[M Fowler]

Contact the transformer maker and ask for return.

[rooster]

+1 on the return PT idea.

[Paultergeist]

....I would suggest turning the power supply caps 90%, just like the typical build. I personally think this is very important, FWIW.

May I ask: to what purpose?

[rooster]

OK, realizing that Soldano and even the PC board Express - amongst others - use lineal caps at right angles on the board, I would still point out that these
power supply caps are located on the edge of the board closest to the preamp tubes and away from the control panel (and hopefully not exposing the 'hot' end to any signal carrying wires). In the example here in the photos, although the caps are not actually located that close to the preamp section, the 'hot' end is facing towards the front of the amp.

Of course Fender did this way back in '59 in his Bassman (put the cap on the board in the middle of the preamp section) but the cap was positioned with the shield facing the front of the amp. I doubt this was random but turn it around and see what you get. FWIW, now days many companies like Crate/SLM, some current Vox and Marshall, and Fender, etc., use caps on a board all the time. But, keep in mind that these are vertically mounted (PC) caps, not horizontally mounted lineals.

Anyway, I think Ken considered the very basic construction of an electrolytic power supply cap (as did the General Electric and Bell Lab designers before him) as he developed his layout for the various TW circuits. In his layouts, you can see that the shielded/grounded part of the cap was the only thing the circuit board saw, not the positive end with the power supply leads exposed. I doubt this was a random thing, and this is why I suggested that the caps be turned 90%.