New 5f2a build oscillation Issue

[dunner84]

Hello

I had some time while away from home for work, and I decided to build a 5f2a with parts I already had on hand. The transformers were purchased new. Everything else was either left over from other projects or scavenged from an old organ chassis. Caps and resistors are all new. Sockets, chassis material, terminal lugs etc, are all scavenged.
The circuit is a 5f2a with a NFB switch and a preamp voice switch (selects different cathode bypass cap values)

When I first fired it up, everything sounded great. It is quiet, and beautiful sounding... Until I turn the volume into overdrive. When the volume gets up near max, the amp cuts out.. I suspect parasitic oscillation, but I am not sure where its coming from.

Here is a list of things I have done so far:

Reflowed all the solder joints
Added shielded wire to the preamp grids
Repositioned the OT on the chassis and re routed the leads.
Rerouted the power tube grid
Swapped the OT primaries
Increased the 6V6 grid resistor to 4.7k
Bent the preamp heaters down to the chassis
poked and prodded and wiggled all of the remaining wires.

Here is where it is at. When I wiggle the second preamp grid wire, the problem goes away. However, any kind of vibration and it just cuts out again. No squeals, no hum, it just cuts out.
Turning the volume down and then back up again cures the problem until the volume gets back up near max

All my voltages look good too.

I will post a pic of my layout, pictures of the initial build, and pictures of how it looks now.

Any kind of advice would be so helpful. I don't have a scope so I can't see what frequency I am dealing with, or even where it is, so that I can filter it out some how.

The chassis is much bigger than a standard 5f2a chassis so there are some longer runs from the power section, and maybe that is the cause, but I figured the spacing between everything only would have helped. Maybe it did because the amp is VERY quiet in terms of hum.

I am just at a loss as to where to go from here.

[dunner84]

Before shots

[dunner84]

Before shots

[dunner84]

Before shots

[dunner84]

Before shots

[dunner84]

Before shots

[dunner84]

Before shots

[dunner84]

Before shots

[dunner84]

After shots

[dunner84]

After shots

[dunner84]

After shots

[dunner84]

After shots

[dunner84]

After shots

[shooter]

volume gets up near max, the amp cuts out.. I suspect parasitic oscillation, but I am not sure where its coming from.

your layout show the grid wire crossing the B+, do you have *lots* of room between them?
did you ohm the pot to see if there's a dead spot where the amp cuts out?
did your remove the NFb as a test?  with nfb disconnected, try tacking a Ecap 4.7uf to 25uf cathode to ground on v1B.
try gator clipping a 470pF to say .001uf from plate to cathode of V1B

[Paul1453]

volume gets up near max, the amp cuts out.. I suspect parasitic oscillation, but I am not sure where its coming from.

your layout show the grid wire crossing the B+, do you have *lots* of room between them?
did you ohm the pot to see if there's a dead spot where the amp cuts out?
did your remove the NFb as a test?  with nfb disconnected, try tacking a Ecap 4.7uf to 25uf cathode to ground on v1B.
try gator clipping a 470pF to say .001uf from plate to cathode of V1B

I also noticed that just rectified B+ right under the grid wire.
I can't see everything well, but I'm guessing the green & black wires are connected to the output jack.
So that means that grid wire is running very near that too.
Rerouting that unfiltered B+ to the OT and main board might help?

I like the before B+ routing much better. It still goes right under the output jack though.
Maybe shield that grid wire?

Only had the Hoffman 5F1 available for a schematic, unsure what changes are on 5F2A.
Looks like F2 has two pots?

[dunner84]

Hi, thank you for the replies.

With NFB disconnected the problem still occurs

The Grid wire does cross the b+. Both are longer runs due to the nature of the chassis. They did run parallel at first with a lot of space between them. Now they cross with little space. I have chops ticked them with no luck.

The green and black wires do go to the output jack. They have been re-routed as well.

I did think about shielding the grid wire. Do you think it is worthwhile? I am not sure what to ground it too, as all of the ground lugs are a bit far from either end.

I am not sure what you mean about a dead spot in the pot? do you mean an open connection to ground at some point? If so, I do not know, and I can check.

I will also try clipping the caps in to see what happens. I am assuming they create some kind of low pass filter?

I will try these and report back.

[Willabe]

I did think about shielding the grid wire. Do you think it is worthwhile? I am not sure what to ground it too, as all of the ground lugs are a bit far from either end.

Yes, use shielded wire for the 6V6 grid. I'd try and ground it at the 'send' end from the 2nd gain stage.

You could/should have put the 1st filter cap, feeding the B+ to the OT and the 2nd filter cap for the screen supply B+ over at the far end of the chassis close to the PT and rectifier tube.

AND, the 1st B+ filter caps ground connection should be connected directly to the PT's B+ CT. That ground is the noisiest ground in the amps as it has the most current. 

Read this on grounding from Merlin;  http://www.valvewizard.co.uk/Grounding.html

If you do this with the 1st 2 filter caps, then they will have short wire runs to what they feed and then when you run the B+ to the preamp 12AX7 it will be much cleaner and less current and less DCV. This is good because now the B+ wire run will radiate much less noise/hash that can be injected into other wires.

The way you have it now the unfiltered and full current of the amp rectified B+ goes past the 6V6 and then has to come back, not the best for noise control.

[shooter]

shielding the grid wire

run a shielded wire, connect the shield to the ground point on the pot, the *other* end only connect the center conductor.  Anytime signal wires cross power wires you want them at 90deg and as far away as possible.

an open connection to ground

yup, remove the 2 NON-grounds from the Vol pot, connect your meter from the ground to wiper, sloooowly rotate the pot and see if the display jumps, goes infinite, gets all jittry, etc.  should be a fairly smooth numbers up, numbers down, as you twist the pot from one end to the other end.

[dunner84]

I clipped in 4.7uf and then 25uf from V1b plate to ground. Both made the problem worse. I used a .001uf from V1b plate to cathode, and it had no effect.

I will try running a shielded wire for the 6v6 grid. I really hope this works. This was supposed to be a straightforward build, which is why I went from scratch. The positive from this experience is the amount that I have learned.. It is just disheartening to put effort into making something tidy only to have it not work.. I will rig up a shielded wire and see how it goes.

[shooter]

Do what Willabe suggests on your grounding and B+ taps, I missed that

[dunner84]

Thank you.
The filtering is currently grounded to a PT bolt but it is a long run. Moving that section over makes a lot of sense. I needed to have a good look through the chassis before I committed to moving the filter caps. It shouldn't be too hard to move the first 3 filter caps closer to the pt. With the new orientation of my OT I don't have a lot of room on the top side of the chassis to bolt the terminal lugs. How close to the PT can I be? As is, I have room for those filter caps about an inch away from the PT. The only long B+ lead will now be the one that connects to the 22k resistor on the last filter cap. Is this going to be ok?

Thank you again for all of the help so far!

[Paul1453]

Don't be too disheartened, this stuff is not as easy as it looks.
Even for a trained repair tech, the design/layout engineering is complex and not obvious except to very experienced designers.
Of which I am not one of those.

I am having significant oscillation issues with many circuits when I try to prototype them on my breadboard.
That is why I am involved in this thread.  I am trying to better learn the tricks and tips of this myself.

On this you have a large chassis with plenty of room.  Did you consider mounting your board closer to your rectifier?
How about splitting the PS and signal parts between two boards?  Mounting the PS parts near the PT and rectifier and the signal board on the other end?
Running clean, filtered voltages across your chassis usually causes less of these issues.

PRR gave me the analogy of having the water pipes come in the east side of your house and the sewer lines exit on the west.
If we keep our inputs away from our outputs we have less chance of cross contamination.   

[dunner84]

I think that is what will essentially be happening. I will be separating the first 3 filter caps and placing them closer to the PT and rectifier. The last 8uf filter cap will be staying on the preamp side for now because it is grounded with the preamp grounds. I will still be shielding the 6v6 grid too..

I am actually glad I learned all of this. My next build is also a scratch build. It will be a 1 channel vibrolux reverb without the trem. So a vibroless reverb... I was supposed to have started building it already, but this 5f2a has eaten up a lot of time.

I am hoping to have time later tonight or tomorrow morning to shield the 6v6 grid and move the filter caps. Fingers are crossed.

[Paul1453]

I am confident you will be able to resolve this issue.   

Like you said you will have gained valuable experience from dealing with this problem.
When we learn from first hand experience, it tends to stay with us for a very long time.
Hopefully we learn enough not to make the same mistakes again.

I am learning that with high voltage tube circuits the design/layout is the most critical part of the build.
Poor decisions made during this phase are often difficult or impossible to correct after putting everything together. 

[Mike_J]

My first build was a 5f6a bassman.  When I first started testing it there was motorboating.  Solution for that problem was 1.5K grid resistors to the grid (pin 5) of the power tubes.  This is what Fender did on the blackfaced amps but hadn't done yet on the tweed amps.  It would be worth a try because I didn't see that you had a grid resistor on your power tube from looking at the pictures of your build.


Thanks
Mike

[Willabe]

My first build was a 5f6a bassman.  When I first started testing it there was motorboating.  Solution for that problem was 1.5K grid resistors to the grid (pin 5) of the power tubes.  This is what Fender did on the blackfaced amps but hadn't done yet on the tweed amps.  It would be worth a try because I didn't see that you had a grid resistor on your power tube from looking at the pictures of your build.

Yes, good catch. 

I missed that. 

[dunner84]

Thank you for the advice. I initially had a 1.5k grid resistor on the 6v6, and even tried bumping it up to 4.7k. That is what is currently on there. I just ran it from pin 5 to pin 6, and from pin 6 to the board. Is this an ok way to do it? The 6v6 pinout showed pin 6 as unused. In my before shots, I actually have this backwards, but I did eventually fix it.

[Mike_J]

Thank you for the advice. I initially had a 1.5k grid resistor on the 6v6, and even tried bumping it up to 4.7k. That is what is currently on there. I just ran it from pin 5 to pin 6, and from pin 6 to the board. Is this an ok way to do it? The 6v6 pinout showed pin 6 as unused. In my before shots, I actually have this backwards, but I did eventually fix it.

The way you have it wired would work.  Looking at the Princeton 5f2a layout it is apparent that the amp and chassis were pretty small.  As such, the wires were probably pretty short.  Your chassis looks like it is quite a bit larger than what the original probably was.  The longer wires could be causing the oscillation problem you are having.  I noticed the grid wire on the layout going from V-1-7 to the wiper of the volume control was pretty long.  I don't remember you  saying anything about using a cable there.  If your wire is very long going to V - 1 - 7 it might be a cause of the oscillation.


Thanks
Mike

[Mike_J]

Sorry, missed the picture on reply 9.  Looks like you already tried the shielded wire to the volume pot wiper.


Thanks
Mike

[dunner84]

Today I moved the power filtering over so that it is closer to the PT and rectifier tube sockets. I also added a shielded cable to the 6v6 grid, and replaced the shielded cable to preamp grid I made it a bit longer so that i can move it around and play with its placement. Things are worse. The amp hums. The volume fluctuates all over the place, the overdrive, when present, sounds awful and it all changes as I move that shielded preamp wore and the new shielded 6v6 grid wire.

I have also redressed a few wires off of the preamp tube so that they all cross at right angles and have as much gap between them as possible.

Touching the 6v6 grid wire with my finger cuts out the hum, but the placement of the preamp one dictates a lot too...

[dunner84]

It doesn't make any sense, but I think it is fixed! I just clipped the braiding for the shielded cable on the 6v6 and all the problems went away.... I am so confused, but I am pretty happy.. I just gave it a 10 minute workout while probing around at the troublesome wires without any issues!

I just want to say that I really appreciate all the help I got on this forum. I don't spend a lot of time here, but I think that will be changing especially since my next build is based on Doug's 1 channel ab763.

Here is a link to my initial thread for this build. It is on another forum, so sorry if that is a no-no or breaking the rules. There are some more pictures of the project there as well as a few answered questions to some things I encountered along the way if anyone is interested.

http://www.tdpri.com/forum/shock-brothers-diy-amps/597938-new-5f2a-build-mods.html

[Paul1453]

It does make some sense.  There was obviously some kind of Ground loop issue.

And you tried something that fixed it.  Good Job!   

I knew you could get this one to work right.   

[shooter]

I just clipped the braiding for the shielded cable

was the braid attached at both ends?

[dunner84]

No. That is why I think it is pretty strange. I am happy it is working though! Thanks again for the help.

[Willabe]

No. That is why I think it is pretty strange.

So now the shield on the 6V6 grid is not hooked up on either end?

[dunner84]

That is correct. Like I said though, I don't really get it. There are 3 ground points on the amp. 1 at the preamp, 1 for the power cord, and one for the power filtering, so I am not sure how a ground loop could have been created. Now when I chopstick that grid wire, it is dead quiet.

[Paul1453]

It doesn't make a lot of sense, but since you had that shielding wire connected in the preamp section that is where I would suspect the grounding issue is if you have anymore problems.