5F6A Bassman total custom build - super low output

[tomgilmartin]

OK, so I reversed the leads from the OT. Bad idea, that was definitely positive feedback, completely unplayable. All scratch and nastiness. I put it back and reconnected the negative feedback.

I cut my zip ties and separated the groups of wires. The guitar inputs are very sensitive. Even when I put my fingers near them the hum goes waaay up, theremin-like. I think a key issue may be that I put the grid stoppers at the jacks. I did this because if I didn't, then I would need four pairs instead of two to get the input signals over to the first preamp stage. I am rethinking this now. Sage advice (read, Chuck) says put grid stoppers right at the tube socket. That's how the second stage CF and third stage PI are set up. Thoughts?

[The Dude]

I can't really tell from the pics or layout diagrams. Did you use switching jacks on the inputs so that the jack is shorted when nothing is plugged in?

[Dave H]

Ok, this ones for Tom (just good natured ribbing). Speculating on the inspiration for his design:

A cat should never think outside the box.

[Dave H]

I cut my zip ties and separated the groups of wires. The guitar inputs are very sensitive. Even when I put my fingers near them the hum goes waaay up, theremin-like.

Good move on the zip ties, There are a lot of wires bundled together which we normally try the keep separate. I also see a lot of long wires looping between the input, preamp and power amp panels but I don't see any screened cable. I think twisted pairs work best on low impedance differential signals. I'd use them for the PI outputs and OT inputs but I'd be thinking of using screened cable for the high impedance single ended signal wires. The most sensitive wires in the whole amp are the guitar inputs and they look to be the longest. I'd definitely use screened wire for those and the preamp tube grid wires. How well would a twisted pair work for a guitar lead?

Is that a multi-section can cap on the power amp panel? It's surrounded by the power and rectifier tubes and then inserted in a tube. How hot does it get? The pre-amp sections of the cap are a long way from the pre-amp they are trying to de-couple. Perhaps some extra caps mounted right on the pre-amp would help with the oscillation problem?

Good luck and keep us posted.

[tomgilmartin]

I can't really tell from the pics or layout diagrams. Did you use switching jacks on the inputs so that the jack is shorted when nothing is plugged in?

Yes, they are switching type, Switchcraft MN112A. That part, at least, I wired in the standard way. Well, except that I used isolated jacks to keep with my star ground scheme. Reinvent the wheel and all that. That input circuit is so simple, elegant and functional that I call it genius.

[tomgilmartin]

Good move on the zip ties, There are a lot of wires bundled together which we normally try the keep separate. I also see a lot of long wires looping between the input, preamp and power amp panels but I don't see any screened cable. I think twisted pairs work best on low impedance differential signals. I'd use them for the PI outputs and OT inputs but I'd be thinking of using screened cable for the high impedance single ended signal wires. The most sensitive wires in the whole amp are the guitar inputs and they look to be the longest. I'd definitely use screened wire for those and the preamp tube grid wires. How well would a twisted pair work for a guitar lead?

Is that a multi-section can cap on the power amp panel? It's surrounded by the power and rectifier tubes and then inserted in a tube. How hot does it get? The pre-amp sections of the cap are a long way from the pre-amp they are trying to de-couple. Perhaps some extra caps mounted right on the pre-amp would help with the oscillation problem?

Good luck and keep us posted.

I believe you mean Coax when you refer to screened cable. I have always considered Coax as being for Mhz signals. I haven't considered it as keeping OUT Mhz. I have to think about this some more.

Maybe I will have to try an unshielded twisted pair for a guitar cord. It works for 100 meters on an ethernet cable without picking up too much noise.

Have to go to work, more later...

[Tom Phillips]

...
Maybe I will have to try an unshielded twisted pair for a guitar cord. It works for 100 meters on an ethernet cable without picking up too much noise...

The answer to that is a strong no. You need to shield the low level signal from the guitar pickups.

[tomgilmartin]

The answer to that is a strong no. You need to shield the low level signal from the guitar pickups.

That was sarcasm. You can tell by the ellipsis...
So hard to convey without vocal tone.

[eschertron]

I don't think a 100m guitar cable is a good idea, screened or not... (note the elipsis)

but seriously, on my first build I put all the PSU caps together, doghouse style. On power up I got serious motorboating - flub, flub, flub - which was fixed by moving the input stage de-coupling cap right onto the plate resistors for that node (you know what I mean, positive lead where the plate resistor sees B+ and neg lead to where the cathode resistor sees signal ground). Now as a matter of design I place the de-coupling caps for the preamp right on location. Never have looked back

[tomgilmartin]

I don't think a 100m guitar cable is a good idea, screened or not... (note the elipsis)

but seriously, on my first build I put all the PSU caps together, doghouse style. On power up I got serious motorboating - flub, flub, flub - which was fixed by moving the input stage de-coupling cap right onto the plate resistors for that node (you know what I mean, positive lead where the plate resistor sees B+ and neg lead to where the cathode resistor sees signal ground). Now as a matter of design I place the de-coupling caps for the preamp right on location. Never have looked back

This I believe I got right. I have a lovely Sprague Atom for my clean stage filtering connected directly to the pins of V2 CF.

[tomgilmartin]

OK back to weekend warrior mode. I now have to go full-on scientist here. I know it sounds nuts, but actually I considered myself lazy for using the twisted pairs of 600V PVC insulated wire for cables inside my amp. I had intended to use shielded-twisted-pair (STP), but it just takes FOREVER to do it. Below is a pic of lead dress for the STP military cable. This one pair of cables took me over an hour. It is exponentially more difficult to do it this way. Just untwisting the damn braid seems like forever.

With this particular cable, the silicone insulation gives a 600V rating with VERY thin insulation. Thinner insulation has advantages because the conductors (actual metal) are closer together. The self-capacitance and self-inductance of the cable are reduced and less electro-magnetic field can get between the conductors from outside to couple to the cable. No matter how tight I twist the PVC insulated 20AWG 600V wire from McMaster, there is always a gap between the wires of 2 times the insulation thickness. Bigger gap = more crosstalk.

So now I will have a separate wire from each input. The pic shows 2 resistors soldered to the red (+) wires, and connected together. These are the 68K grid stops into the preamp. With this scheme I'm NOT dropping my already-low guitar voltage across these resistors and putting it on a wire in the noisy environment inside the amp until the absolute last possible point.

The blue (-) wires will be tied together with another group of 2 and connected to the star-ground point at the preamp tube socket.

The cable shields will all be tied together and then go to chassis ground, only touching the blue (-) wires through the point where they tie to the chassis. The shield is like an antenna - charge is pushed and pulled into and out of the ground point, but doesn't travel on the blue (-) wire.

Let's see if it works as advertised.

PLEASE DO CALL BULLSHIT ON ME IF YOU THINK I AM GETTING IT WRONG. I LOVE TALKING ABOUT THIS CRAP!
I know it's obsessive overkill, no need to tell me that! :-)

[tomgilmartin]

This I believe I got right I have a lovely Sprague Atom for my clean stage filtering connected directly to the pins of V2 CF.

[Tom Phillips]

...Just untwisting the damn braid seems like forever...

Tom M,
There is a really efficient technique to extract the conductor wires from a braided shield. Rather than write a narrative trying to describe it, I found a video demonstration at https://www.youtube.com/watch?v=JW2esD1NIa8 The extraction process is shown starting at the 2:15 time mark. In the demonstration shown a pigtail wire is attached to the shield but that step is not always necessary if the shield is to be terminated close to the conductor connection in the build.
Check it out.
Cheers,
Tom P

[tomgilmartin]

Tom M,
There is a really efficient technique to extract the conductor wires from a braided shield. Rather than write a narrative trying to describe it, I found a video demonstration at https://www.youtube.com/watch?v=JW2esD1NIa8 The extraction process is shown starting at the 2:15 time mark. In the demonstration shown a pigtail wire is attached to the shield but that step is not always necessary if the shield is to be terminated close to the conductor connection in the build.
Check it out.
Cheers,
Tom P

Awesome, thanks!!

[tomgilmartin]

UPDATE:

Installed the SHIELDED twisted pairs from the inputs to the preamp. 4 wires, one each for each input. Moved the grid stopper 68K resistors to the other end of the cables, right at pins 2 and 7 of V1.

What incredible noise!!!! GOOD noise!

The whistling feed back is completely gone. I still get some whup-whup-whup at volume 10 with nothing plugged in. I can't crank it with the guitar on because it's literally waking up the children at volume 3.

New pics below. Thanks again for all the great advice!!





At the back of preamp V1.