SOLVED Seeking 5F6A Build Help: Tests and sounds great, but bypasses tone stack?

[Willabe]

Interesting.

Well, according to that article, any true push/pull class A power amp will be missing a lot in the tone we like from what is really happening when the amp is turned up enough to strain the power supply enough to sag.

(And, an AC30 does come out of class A and into class A/B turned up enough.)

[shooter]

does come out of class A and into class A/B turned up enough.

^^^^^^^


that's one of the infuriatingly fun things to "get right"
I've had the most luck with SE Kt88 or PP EL34's

[gfarina55]

That article is a bit over my head, but it reminded me of something else I noticed: A pronounced 60 cycle hum when I turn the amp up to around 7 and beyond. (I believe it's 60 and not 120). As a player not really an issue b/c at 7 the amp takes paint off the walls, but it's definitely there. Might the two symptoms be related?

Silvergun, I don't have a power resistor but I'll order one to try. A bad choke could cause these ghost notes?

Also, I'm also using new Weber 10A125 30w speakers, light dope, and Weber says it could quite possibly be "cone cry" from all four new speakers. Even so, the amp sounds kind of brash and sharp in overdrive and I'm wondering if there's something in the power section that could be improved.

[SILVERGUN]

That article is a bit over my head, but it reminded me of something else I noticed: A pronounced 60 cycle hum when I turn the amp up to around 7 and beyond. (I believe it's 60 and not 120). As a player not really an issue b/c at 7 the amp takes paint off the walls, but it's definitely there. Might the two symptoms be related?

Silvergun, I don't have a power resistor but I'll order one to try. A bad choke could cause these ghost notes?

Also, I'm also using new Weber 10A125 30w speakers, light dope, and Weber says it could quite possibly be "cone cry" from all four new speakers. Even so, the amp sounds kind of brash and sharp in overdrive and I'm wondering if there's something in the power section that could be improved.

Yes the choke could contribute...and yes the hum could be related.
If you PM me your address I can drop a resistor in the mail for you tomorrow so you won't have to pay more shipping than resistor cost.
I'm not suggesting that I am right but there are different ways to troubleshoot power supply issues and this is one of them. Once you rule out the choke you can cross it off the list. (and I don't want you to have to overpay to try it)

After reading more about power supplies today I'm also assuming that there is a balance between an amp that sounds great due to SOME power supply ripple and an amp that can deal with extreme overdrive.
SOME ripple can manifest into a more harmonically rich tone whereas an over-filtered (less ripple riddled) power supply will handle overdrive better.

So, what I'm suggesting is that your amp might be sounding great until YOU push it too far into overdrive and CAUSE the nonlinear distortion effect of adding new frequencies (harmonic distortion). You could fix it with better filtering, but by fixing it you could sacrifice TONE.

Keep in mind, I am a welding machine guy and I'm learning right beside you, but this is pretty cool stuff. You might be able to "tune" out the issue with a different value filter cap, which will be a compromise between retaining some good ripple but enabling your brand of full tilt overdrive.

A great man once said "Very low level hum is not offensive; it is "the sound of electricity". An utterly hum-free amp sounds artificial, phony. You are free to disagree."
I agree.

[SILVERGUN]

On a separate note:
Let's go back to your original post...
You said that you incorporated Robrob's NFB switch.

Does your OT have a 2 ohm tap?
Did you use it for your NFB tap?
Does the ghost noting change at all if you select a different NFB switch setting?


You will also want to confirm if that is 60hz or 120hz that you are hearing.
Use your phone to generate those test tones and match it up. That's about as simple as it gets.

[gfarina55]

Thanks Silvergun! Yes, the OT has a 2-ohm tap only. I have used Rob's NFB mod and the three settings (Bassman/none/Marshall) have no impact on the ghost notes. (I also did all the standard trouble shooting: swapped every tube, tried all inputs, settings, etc.) I've also listened to 60vs.120 and I'm convinced this is 60. Again, the hum isn't an issue from a player's perspective as it's not very loud, but I wondered if it might be a clue.

At this point I think the ghost notes and the "brashness" of the overdrive might have different causes. I think the ghost notes may be cone cry, as it seems to be  common to 10A125 speakers. (I have to lug my C-30 over to the studio to test the 5f6a speaker cab at some point.) The "brashness" might be related to what you are describing.

Thanks for the offer about the power resistor; I just ordered an Ohmite "Dividohm" 8-ohm 225W Power Resistor to use as a dummy load. Would this work? Also, what exactly do you mean by "better power filtering"? I read this phrase in a lot of posts/articles and I'm not sure about the specifics. Is there a "better" scenario than the 4x22/500V filter cap setup?

[SILVERGUN]

Thanks Silvergun! Yes, the OT has a 2-ohm tap only. I have used Rob's NFB mod and the three settings (Bassman/none/Marshall) have no impact on the ghost notes. (I also did all the standard trouble shooting: swapped every tube, tried all inputs, settings, etc.) I've also listened to 60vs.120 and I'm convinced this is 60. Again, the hum isn't an issue from a player's perspective as it's not very loud, but I wondered if it might be a clue.

Since ghost notes can commonly be caused by a faulty power supply any clues are helpful. It can be difficult to differentiate between 60 and 120hz because they are directly related harmonics. A slight 60HZ hum is pretty common in amps. My new favorite saying is "it's the sound of electricity".

At this point I think the ghost notes and the "brashness" of the overdrive might have different causes. I think the ghost notes may be cone cry, as it seems to be  common to 10A125 speakers. (I have to lug my C-30 over to the studio to test the 5f6a speaker cab at some point.) The "brashness" might be related to what you are describing.

Yeah, that could be the case. The "brashness" is subjective and could just be the inherent quality of the amp fully overdriving. Are you driving the input with a boost or distortion pedal? Ghost notes (or harmonic distortion) can often come from a defective power section.

Thanks for the offer about the power resistor; I just ordered an Ohmite "Dividohm" 8-ohm 225W Power Resistor to use as a dummy load. Would this work? Also, what exactly do you mean by "better power filtering"? I read this phrase in a lot of posts/articles and I'm not sure about the specifics. Is there a "better" scenario than the 4x22/500V filter cap setup?

The offer still stands. The resistor you ordered will work as a dummy load for testing, however you would have wanted to order something closer to 2 ohms since that is your OTs output impedance.
My suggestion was to temporarily use a power resistor in place of the choke to see how or if it affects your ghost note issue. It is common for a power supply to either have a resistor OR a choke in this position, so if you are troubleshooting a possibly faulty choke you can do so by tacking in a resistor (of a similar resistance as your choke) to see if it clears up or changes the ghost note issue.


Before my reading yesterday I would have told you to go ahead and replace the filter caps with higher values and if your choke was Ok then this would probably fix the issue. I didn't think you could over-filter an amp to a fault (within reason).
But after reading that article I am convinced that "over-filtering" can sterilize your amp. And we don't want that.
And yes (to a previous question you brought up), you could still have a damaged filter cap from the initial screen resistor incident. I was just giving you a simple way to eliminate the choke as a possible suspect. Disconnect one wire from choke, tack a resistor in it's place and test.
PM me your address and I can still get one out today. I haven't seen the over-heated postal person yet today.

[gfarina55]

Thx Silvergun! The power resistor I bought is variable, so I can "set it" to 2 ohms. I believe I saw one of these on Slucky's page so I found one on Ebay. I'll use it and also patch in some other filter caps to troubleshoot sometime soon.

[SILVERGUN]

Thanks for inspiring me to continue reading about the subject.
Here's more brain food so you understand where I'm coming from:

[gfarina55]

Glad you're learning faster than I am Silvergun!

So, I've been reading about grounding all night (I've read Aiken and the Hoffman info multiple times) and I'm wondering if the 60 cycle hum might be reduced w/a better grounding scheme. See my setup in the photos below. I used a Rob-style preamp ground bus connected to one of the input jacks, and I have the AC earth bolted to the side of the chassis. I have another bolt to the right of the PT that is my power amp ground. I've connected the PT high voltage CT, the filament CT, the long ground wire from the filter cap doghouse, and a wire from pin 8 of the power tubes (through resistors for quick bias measurements) to this ground. The OT is grounded to the speaker jack.

I found this thread on TDPRI that cites Rob and Hoffman and suggests some improvements:

https://www.tdpri.com/threads/try-robs-pt-ct-trick-one-wire-to-quiet-them-all.1015082/

It suggests that a better way to ground the PT CT is to connect it to the negative side of the "first" filter cap. (Which cap is "first"?) However, the example used in the thread is a Champ in which the filter cap is in the chassis close to the PT. In a 5f6A, the caps are in the doghouse way over on the other side. Is it advisable to run the PT CT the entire way to the doghouse, and then the ground wire back to the power amp ground? Is there a better way to accomplish this?

In the same thread, a TDPRI member wrote: "a trick I learned from Hoffman years ago is to ground the heater CT at the input jack grounding point. In the amps I've done this on, 60Hz hum is gone." Would it be advantageous to actually run the filament CT wire the entire way across the chassis to the termination of the pre-ground bus at the input jack?

Finally, is there a better grounding scheme that might reduce 60 cycle hum? I know that 120 cycle hum is more often a symptom of ground loops, but is my grounding somehow allowing some ripple into the signal? (Silvergun, I'm still considering the caps themselves and the choke.)

[Willabe]

It suggests that a better way to ground the PT CT is to connect it to the negative side of the "first" filter cap. (Which cap is "first"?) However, the example used in the thread is a Champ in which the filter cap is in the chassis close to the PT. In a 5f6A, the caps are in the doghouse way over on the other side. Is it advisable to run the PT CT the entire way to the doghouse, and then the ground wire back to the power amp ground?

Yes.

The 1st B+ filter cap is the cap that supplies the power tube plate. The nosiest ground is that cap because it has the most current going through it and has the most ripple on that cap. So you ground the CT to that caps ground lead, then run a wire to the chassis power amp ground or to the far end of the ground buss.   

Read this;

http://www.valvewizard.co.uk/Grounding.pdf

That wire nut is not safe to use.

[SILVERGUN]

More good info here:
Grounds (el34world.com)


Current Flow (el34world.com)

[gfarina55]

OK, thank you, good to know! I may rethink my grounding scheme.

A bit of troubleshooting this morning: the hum goes away when I pull V.1, so I think the heaters are responsible. Has anyone elevated the heaters in a 5F6A (Fender layout)? Where exactly do you tap the voltage from?

[Raybob]

Thanks for inspiring me to continue reading about the subject.
Here's more brain food so you understand where I'm coming from:

Totally OT here but that information was written before 1953 when the ISO made us go to 440 tuning.  Reading here about the summation and differential tones, It mentions 256 as C and 320 as E. The differential tone is C two octaves below. With A=432, C=256, D=144, E=320, etc.  What a simple concept music was until they had us go to 440 as standard, where all other tones are irrational numbers.

[SILVERGUN]

OK, thank you, good to know! I may rethink my grounding scheme.

A bit of troubleshooting this morning: the hum goes away when I pull V.1, so I think the heaters are responsible. Has anyone elevated the heaters in a 5F6A (Fender layout)? Where exactly do you tap the voltage from?

Before you elevate. Try a different V1.

[SILVERGUN]

Thanks for inspiring me to continue reading about the subject.
Here's more brain food so you understand where I'm coming from:

Totally OT here but that information was written before 1953 when the ISO made us go to 440 tuning.  Reading here about the summation and differential tones, It mentions 256 as C and 320 as E. The differential tone is C two octaves below. With A=432, C=256, D=144, E=320, etc.  What a simple concept music was until they had us go to 440 as standard, where all other tones are irrational numbers.

It's still simple for most guys who don't think about frequency numbers. 

[gfarina55]

Yeah, I tried different tubes in every position. No reduction in 60cycle hum.

[Willabe]

A bit of troubleshooting this morning: the hum goes away when I pull V.1, so I think the heaters are responsible.

That doesn't mean it's the heaters. You can check the heaters by using a large 6v lantern battery. My $$'s on the grounding.

Has anyone elevated the heaters in a 5F6A (Fender layout)? Where exactly do you tap the voltage from?

Yes. I like it.

I'd make sure your grounding correct before you chase elevating the heaters. 

[gfarina55]

Willabe, could you look at my grounding in the photos I posted yesterday (a few threads above) and see if anything looks suspect? I used the preamp ground bus in Rob's layout, my OT is grounded at the speaker jack, the AC cord is "earthed" to a dedicated bolt on the side of the chassis, and my power amp ground has the following connections:

>Grounds from filter doghouse
>PT CT
>Heater CT
>Power tubes pin 8 (through a resistor for easy bias measurements)

Just to be clear, you suggest that I run a long wire from the PT CT to the doghouse and connect it to negative lead of the first filter cap, even though that negative lead is running through another wire back to the power amp ground point (where the PT CT is "currently" connected...pun intended.)

[PRR]

> Is it advisable to run the PT CT the entire way to the doghouse, and then the ground wire back to the power amp ground?

Yes.

> Is there a better way to accomplish this?

No. "Better" how? Less wire cost?

In a HIGH current (transistor) amp we'd put PT Rect and Caps all real close and strap them with copper busbars. But in a tube amp the current ultimately flows through a vacuum, not very well, and even long skinny wires conduct better than that.

The "first cap" is the one fed directly by the rectifier. (In a 2-Diode CT rectifier, one leg is the PT HV CT.)

[gfarina55]

By "better" I meant more elegant (less wire), but I'll try this and report back. Thanks!

[Willabe]

....could you look at my grounding in the photos I posted yesterday (a few threads above) and see if anything looks suspect?

>Grounds from filter doghouse
>PT CT
>Heater CT
>Power tubes pin 8 (through a resistor for easy bias measurements)

I can't tell be looking at a picture. Read the link on grounding I posted for you.

In it you will see you can't ground all the filter caps at the power ground and use a ground buss for all the preamp and PI circuitry.

Just to be clear, you suggest that I run a long wire from the PT CT to the doghouse and connect it to negative lead of the first filter cap, even though that negative lead is running through another wire back to the power amp ground point (where the PT CT is "currently" connected...pun intended.)

Read the grounding link. Then we can talk if you still need a little more clarity.

[gfarina55]

Thx Willabe.

Regarding the PT CT, I ran a wire from it to the leg of the filter caps (where the white wire is connected in the photo below) and it made absolutely no difference in the hum. (Again, there is a wire running right back to the power amp ground, so I don't fully understand this concept.)

I've read the valvewizard info to which you linked (also before I built the amp), and also the Aiken Amps grounding page, and some others, and I've followed all I can understand. I get the concepts, but I'm still trying to understand "where the wires go".  When I built the amp, I tried to stick to this: https://el34world.com/charts/grounds.htm

But mine has some differences:

1) I followed Rob's layout which only has 4 filter caps (why?). My filter cap board (photo below) is exactly like Rob's. In my photo, the white wire goes to the power amp ground point, and the black to the preamp ground bus, just as valvewizard states. Correct, yes?

2) I have no *wire* that connects the power amp ground to the preamp ground bus, and no *wire* from the OT secondary (grounded at the speaker jack) to the power amp ground. Should I, or does the chassis suffice?

3) Hoffman's grounding scheme doesn't show a PT heater CT. My transformer has this, and I was told to connect it to the power amp ground. Is this correct?

The other thing is that I notice the hum more when I turn the bright channel vol up. I assumed this is b/c it "brightened up" the hum, but could something be awry only in one Chanel?

[gfarina55]

Maybe this helps!

[sluckey]

That's basically the same ground scheme that Hoffman and I use.

[Willabe]

I looked at robrob's layout ground wiring. And found your link to robrobs updated 1st B+ filter cap ground wiring that's on a web forum. What robrob did in the updated ground wiring is what I'm talking about.

He moved the PT B+ CT from the chassis power amp ground lug to the 1st filter cap negative lead. Then he ran a wire from that junction, ie; filter cap negative lead/B+ CT, over to the chassis power amp ground lug/bolt. That's the same as PRR, Merlin, Kevin O'Connor/London Power and others recommend. 

Here's why;

Short answer = create a short ground loop for the 1st filter cap/CT isolated from the other grounds. 

Long answer;

Fender and many others used a random ground scheme. They used the chassis as the ground buss where it was convenient to make a ground chassis connection. Fender did separate the power amp ground, ie; power tubes and PI grounds from the preamp grounds. Look at the Fender Tweed, brown face and black face amps layout drawings, it's there. Although on the Tweeds, the power amp filter cap chassis ground is inside the dog house on the back of the chassis, so it's not shown on the layout. Our host Doug uses a grounding scheme very close to Fenders. He separates the power amp chassis ground from the preamp ground but he uses a ground buss for the preamp. It works.

KOC, Merlin and others use a wired ground scheme. They put the B+ filter cap close to the circuitry it feeds and wire the grounds from those circuits directly to the filter caps ground lead that feed that circuit. This keeps the loop short and isolated.

Doing that they do not use the chassis as a ground buss. Random ground currents in the chassis can cross each other in the chassis and impose a signal that will modulate a different circuit, that causes the buzz.

They also make individual stars on the buss that are then wired together. They don't just run each ground wire to the closest buss point. A preamp tube would have the grounds for the K, G1 return and filter cap that feeds that tube tied together as a star then run a wire (buss) to the next star. Merlin has drawings showing this.

What this does is creates individual ground loops for each filter cap that are short/small and isolated from each other. Current for the tube comes up from ground, thru the K R, thru the tube, out the plate, thru the plate R, to the + end of the filter cap, thru the B+ filter cap, out of the B+ filter cap thru it's ground lead and back to the K. And round and round it goes. But that current is isolated from the other loops so it doesn't mess with them.

With the power tubes, current up thru the K, out the plate to 1 side of the OT primary, thru that 1/2 of the OT wind to it's CT, out thru the CT, to the + of the B+ filter cap, thru that filter cap, out the negative lead and back into the power tubes K. And round and round it goes.

The 1st B+ filter cap is the cap that supplies the power tube plate thru the OT CT. The nosiest ground is that cap because it has the most current going through it and has the most ripple on that cap. So that's why this ground should be wired up correctly to minimize noise.

That's why I'm telling you as PRR, KOC, Meriln, and robrob, to make that B+ filter cap loop isolated, ie; B+ filter cap negative lead/OT CT, then add a wire from there to your tube power tube ground star, that wire isolates the loop from the power tube star ground. It's the purple arrow in Merlin's drawing.

Then, if your using 2 chassis B+ filter grounds, ie; power amp ground/preamp ground, run a wire from the CT /red arrow wire, power tube ground star/purple arrow wire and the B+ filter cap negative lead to your chassis ground. In Merlin's drawing, it would be the preamp daisy chain ground wire. He uses 1 chassis ground, not 2 chassis grounds for the B+ filter caps. 

Look at this; the red arrow points to the OT CT, the purple arrow to the power tube star ground.

All said, you want to keep the filter cap/circuitry ground loops short, ie; star and place the filter cap close to the circuit id supplies, and use less/least amount of chassis ground connections as is reasonable/possible, ie; star buss. This works well. Ground stars forming a galaxy = galactic grounding scheme.   

And I attached robrobs updated power amp OT CT ground connection.

[Willabe]

Look at this, B+ CT/1st B+ filter cap move. 

From PRR;

https://el34world.com/Forum/index.php?topic=27590.msg305790#msg305790

Then his result after moving wire from chassis to where it should be;

https://el34world.com/Forum/index.php?topic=27590.msg306339#msg306339


 

[gfarina55]

Hi Willabe,

Thanks for this detailed explanation! I didn't understand from your previous posts that you were suggesting that I move one of the filter caps, but creating isolated "individual ground loops" for each filter cap/cap set is genius. B/c I'm building a "traditional" 5F6A I think I'll keep the caps in the doghouse on this build, but I'm definitely going to try the wired grounding strategy on my next build. Thank you again for such a clear and detailed explanation. With a "wired ground scheme", is it advantageous to isolate the inputs?

Robrob's layout does indicate the first 3 filter caps grounded at the power amp ground point, and the fourth at the preamp ground bus, so that's what I did. I know the old school way is to ground them right to the doghouse, but I can't imagine this is safe or quiet. I spent a couple hours in the studio with the amp yesterday, and I have to say the 60 cycle hum that I'm getting is negligible when I'm playing the amp at proper volume. I'm just trying to get it as quiet as possible w/in certain limitations, and this exploration of grounding has been instructive.

Regarding the "ghost notes" that I was encountering earlier in this thread, I believe the issue is that I switched from a Tele to a Les Paul recently. After some troubleshooting, I've found that the Les Paul also creates ghost notes on other amps, and I think the 5F6A just doesn't have the right power filtering to handle overdriven humbuckers. When I played my Tele through the 5F6A the amp began to make a lot more sense.

[Willabe]

Thank you again for such a clear and detailed explanation.

Seems not clear enough.     

I didn't understand from your previous posts that you were suggesting that I move one of the filter caps, but creating isolated "individual ground loops" for each filter cap/cap set is genius. B/c I'm building a "traditional" 5F6A I think I'll keep the caps in the doghouse on this build, but I'm definitely going to try the wired grounding strategy on my next build.

I'm not telling you to move any filter caps. Leave them where they are in the dog house. 

I'm saying to move the PT CT wire to the 1st filter cap negative lead.

Just to give a name and description of some ground schemes to help sort them out, let's say there's;

1. Standard traditional ground scheme.

2. Better ground scheme(s). (Separate power ground/preamp ground using a ground buss with 2 chassis grounds or a full ground buss with 1 chassis ground. Either of these are normally very quite.)   

3. Better+ ground scheme. (Galactic grounding. You can have this with out moving some/all of the filter caps, just that the wired loops will be longer. At some point it doesn't make sense with long wire loops running around the chassis.)

4. Best ground scheme. (Galactic grounding, all caps moved, distributed in the chassis, stars made, small loops, and all jacks isolated from the chassis and wired to a ground star point, input, speaker, verb, etc. KOC favors this. Often can be over kill.)
 
What you have now for the grounding scheme is somewhere between standard and better, robrob used 2 chassis grounds and a ground buss for the preamp circuit grounds, instead of random chassis grounding. But he missed the PT B+ CT ground, which he now has changed over to like many others. (And he missed the PI ground, see below.)

....I'm just trying to get it as quiet as possible w/in certain limitations, .....

I'm suggesting a slightly better ground scheme. The difference in wiring is subtle, but can make a large difference. 

Look at the hi-lighted robrob drawing below.

1. Move PT B+ CT wire, red/yellow, now hi-lighted red, to the 1st B+ filter cap negative lead. 

2. Move your white power ground wire, gray in robrob drawing, now hi-lighted orange, to the 2nd filter cap negative lead, it's the screen cap. The hi-lighted yellow wire is now isolating that 1st B+ filter caps current loop from the rest of the filter cap loops.

3. I just saw this; robrob should have grounded the filter PI cap with the power ground because this amp has a negative feed back loop. That -FB loop is feed from the OT secondary. So it gets grounded with the power amp.

Disconnect the PI caps ground, gray wire in robrob's drawing, from the preamp buss, connect a jumper, hi-lighted purple from that PI cap over to the screen cap.

So now all 4 dog house filter caps are grounded together but the 1st filter caps loop is isolated. 

Robrob's layout does indicate the first 3 filter caps grounded at the power amp ground point, and the fourth at the preamp ground bus, so that's what I did. I know the old school way is to ground them right to the doghouse, but I can't imagine this is safe or quiet.

There's a subtle difference in wiring up the ground that can make a big difference in noise levels. He now does this, it would be nice if he went back and up-dated his layouts to reflect his new change in wiring that 1st filter cap ground. 

It's just as safe to ground it in the dog house as grounding it in the chassis. Both are soldered to the chassis.

Fender did move that ground after the tweed amps to the B+ CT ground, instead of having the dog house ground go through the chassis over to the PT B+ CT chassis connection. We don't know why they did it, might have been they found it quieter, might have been easier to build, might have been both?   

[Willabe]

I've found that the Les Paul also creates ghost notes on other amps, and I think the 5F6A just doesn't have the right power filtering to handle overdriven humbuckers.

Could be the PUP's, the guitars wiring, grounding matters there too.

Grounding schemes can cause all sorts on weird, strange sounds, noises. That's why you want to eliminate as much as you can to rule that out as the possible problem. 

With a "wired ground scheme", is it advantageous to isolate the inputs?

Some do, some don't, it's the extra mile. The higher the gain of an amp and the louder you play, the more these things seem to get noticed.

But layout and lead dress can make problems too.

Getting the layout, lead dress and grounding as good as you can, goes a long way to eliminating noise, hum, buzz and other assorted bugs.

[gfarina55]

He moved the PT B+ CT from the chassis power amp ground lug to the 1st filter cap negative lead. Then he ran a wire from that junction, ie; filter cap negative lead/B+ CT, over to the chassis power amp ground lug/bolt. That's the same as PRR, Merlin, Kevin O'Connor/London Power and others recommend. 

Yes, I understand this. As I mentioned above, I tried exactly this and there was no difference in the 60 cycle hum or other noise. However, my 4th filter cap is going to pre-amp ground as in Rob's layout. Your annotated version now shows all 4 filter caps together and to run the wire back to the power amp ground? Isn't this contrary to the Vavlewizard info you cited?

In it you will see you can't ground all the filter caps at the power ground and use a ground buss for all the preamp and PI circuitry.

In any case, I will try it....

[Willabe]

Yes, I understand this. As I mentioned above, I tried exactly this and there was no difference in the 60 cycle hum or other noise.

So you moved the PT B+ CT wire to the 1st filter cap negative lead?

But did you move the power ground wire from the 1st filter cap negative lead to the screen cap negative lead?

Your white wire, I show it as orange.(I explained why this is done. It's subtle but it maters.)

I just found that robrob also has the -bias ground grounded to the ground buss, can't do that. That -bias comes from a tap on the B+ wind and it's fed into the power tube grids. It has to be grounded at the power ground.

And he has the Presence control pot grounded to the mid pots ground that goes to the ground buss. But the Presence circuit is fed from the PI's tail and OT. So he ended up tying the preamp buss through the mid pot ground to the OT ground presence pot/PI and the -bias ground. That defeats the separation of the 2 chassis grounds.

That's all part of the power ground. Not supposed to do that. (No disrespect to robrob, he has a great web site with lots of very good info on it.  )

However, my 4th filter cap is going to pre-amp ground as in Rob's layout. Your annotated version now shows all 4 filter caps together and to run the wire back to the power amp ground? Isn't this contrary to the Vavlewizard info you cited?

In it you will see you can't ground all the filter caps at the power ground and use a ground buss for all the preamp and PI circuitry.

I wrote that wrong, sorry.

The PI is the splitting point for the ground scheme if the amp has a -FB loop and is using 2 chassis ground points. With 2 chassis grounds and a -FB loop the loop just made the PI part of the power amp, so it goes with the power amp ground. No -FB loop, PI goes with the preamp buss chassis ground. That's the split and it falls to 1 side or the other. With a single chassis ground, like Merlin and KOC use, they have a full ground buss going from the power amp to the input jack where the chassis ground is mounted. Although, KOC used to favor the ground buss chassis connection around the PI section of the buss, don't know if he still does?   

It's hard figuring out what's going on from the layout without a schematic. I don't think robrob posts a schematic, just a layout. 

As I said If you have a -FB loop/Presence control, and this amp does, that's all part of the power amp, so it all gets grounded with the power amp. The -FB is tapped off the OT secondary. If you didn't have a -FB loop then you would ground the PI with the preamp buss going to the input jack ground. (You said you read the Merlin link, he's very clear in there about this.) 

Look at this;

[tubeswell]

This topic got me reading more today and I stumbled across this interesting article:
Class AB Power Supply Ripple (ampbooks.com)

Thanks for the link. Motto of the story, if it ain't broke...

[gfarina55]

I did not rearrange the grounding on the filter caps when I connected the PT CT to the negative leg of the first filter cap, but I will. Your annotations to Rob's layout is incredibly helpful to a layperson like me, and I'm surprised there's not a 5F6A groundling layout on one of these forums. I would have loved to have seen this before I built the amp!

So you are saying the PI ground should also go to the power amp ground? What/where exactly is the PI ground? Is it the 10K "tail" resistor currently connected to the NFB resistor and the Presence & Mid grounds?

[Willabe]

Filter caps layout = from the left in the drawing;

1st 2 caps = PT B+ CT , 3rd cap = screen, 4th cap = PI.

I re posted the hi-lighted drawing below and here's step by step instructions;

1. You disconnect the black wire from the presence pot lug 1 that's going to the mid pot lug 1. Red line cut.

2. Leave the black wire from the mid pot lug 1 that's there in place going to the ground buss.

3. Then run a wire from presence pot lug 1 to the 4th filter cap negative lead, PI cap, hi-lighted purple wire.

4. Disconnect the ground wire that's there now, that goes to the closest end of the ground buss. In rob's drawing it's gray, I have 3 red lines cut through it.

5. Now disconnect the short gray wire, it's the ground, that goes from the -bias caps/R/new red ground wire to the ground buss. I have 1 red line cut through it.

6. Then run a wire from the -bias supply ground, junction of 2 @ 8uF -bias caps and 1 resistor, back to the power chassis ground, I have a red wire there.

7. Then connect that 4th PI filter cap's negative lead to the 3rd filter cap, screen, red wire.

8.Last, run a wire from that 3rd filter cap, screen cap, I show it as orange, back to the chassis power ground.

That should get the ground straightened out.

There's 1 last thing, the -bias wire, it's probably a red/blue on your PT, in rob's layout it's dark blue, should not be wrapped around the heater wires. There's very little current going through that wire, only the -bias caps charging current, but, that -bias SS rectifiers pulses produce a very nasty strange wave that could get on/into those heater wires. Move it away from the heater wires.   

Here's the hi-lighted drawing again;   

[gfarina55]

Good call on the filament wire and the bias wire. Mine are intertwined b/c the two wires are intertwined on my Victoria 35210, so I copied that, but I'll untangle them.

Does this diagram represent everything you're saying? The only ground we didn't talk about is the filament CT. Can this go directly to the power amp ground as I have it in my diagram?

Also, both OT wires go only to the speaker jack, and there's no actual wire that goes from the speaker jack to the power amp ground, correct?

[Willabe]

Does this diagram represent everything you're saying?

Yes!              

The only ground we didn't talk about is the filament CT. Can this go directly to the power amp ground as I have it in my diagram?

Yes, leave it where it is, that's why I didn't mention it.

Also, both OT wires go only to the speaker jack, and there's no actual wire that goes from the speaker jack to the power amp ground, correct?

Yes, if your not using plastic Cliff jacks, Marshall amps. A Switchcraft metal jack, Fender amps, by itself makes a chassis ground connection, just like the input jacks. It goes through the chassis, from the speaker jacks ground lug, through the jacks mounting shell that's bolted to the chassis with a star washer to get a good bite into the chassis. No extra wire needed.

[gfarina55]

Ok, fantastic! Yes, I'm using switch craft jacks and star washers.

And just to clarify, this "power amp ground" is just a point on the chassis, separate from the earth ground and the PT Faraday cage ground, correct? (Mine is a dedicated bolt as you can see in the photos).

I'm going to try this "Better +" ground setup, and then when I have everything exactly as I want it, I'm going to rewire the amp from scratch. (Just for fun, and b/c I want to have cleaner solder joints, use different wire, and some different components.)

Thank you for your time and knowledge Willabe!

[Willabe]

And just to clarify, this "power amp ground" is just a point on the chassis, separate from the earth ground and the PT Faraday cage ground, correct? (Mine is a dedicated bolt as you can see in the photos).

Yes.

Almost forgot, the -bias wire and it's ground wire, you can run that as a nice twisted pair at least for part of the way.

[gfarina55]

OK, so I'm rewiring the entire amp in part to implement this new grounding scheme. Willabe, as per your advice I now have 6 wires running to/from the doghouse (see photo). The three red wires are B+, and the three white wires are as follows:

1) Neg side of the 2 x "Plates" filter caps that will connect to the HV CT
2) Neg side of "Screens" filter cap running back to the power amp ground
3) Neg side of the "PI" filter cap that will run to the Presence ground

My question is this: Can I twist the three white wires together? They are quite long, and I'd love to twist them together and tuck them into the corner before installing the board.

[Willabe]

Yes.

But it's better when you can run twisted pairs, +/-.

Edit; Twisting those 3 ground wires together will defeat the purpose of running separate grounds. Any current induced in/on 1 of those ground wires will be induced into the other 2 ground wires.

Your amp was not laid out for what you want to do. But since you've gutted the amp, you would have to move the 1st B+ node, the 2 stacked B+ OT filter caps inside the chassis next to the PT.

Running tightly twisted pairs of B+/ground greatly helps to nullify any/all magnetic field noise from being radiated into anything else near those wires. They cancel each other because their out of phase from each other. 

I don't know why you gutted that amp.     

[Willabe]

Here's the problem I mentioned with layout.

3. Better+ ground scheme. (Galactic grounding. You can have this with out moving some/all of the filter caps, just that the wired loops will be longer. At some point it doesn't make sense with long wire loops running around the chassis.)

[Willabe]

Can you post a full length pic of the back side of the chassis?

[gfarina55]

Yes.


Edit; Twisting those 3 ground wires together will defeat the purpose of running separate grounds. Any current induced in/on 1 of those ground wires will be induced into the other 2 ground wires.

Your amp was not laid out for what you want to do. But since you've gutted the amp, you would have to move the 1st B+ node, the 2 stacked B+ OT filter caps inside the chassis next to the PT.

Running tightly twisted pairs of B+/ground greatly helps to nullify any/all magnetic field noise from being radiated into anything else near those wires. They cancel each other because their out of phase from each other. 

Ugh....I went ahead after you said it was OK to twist the wires together, and the amp is (was) just about finished. At this point I have no way to separate the white ground wires and run them under the board, and there are ground connections missing under the board.

A simple question: Is there any way to improve the grounding of Rob’s layout without moving the filter caps? (Moving the caps is no longer an option.) Or, is Rob's layout the best grounding scheme for a "traditional" Bassman layout?

I don't know why you gutted that amp. 

I gutted the amp b/c I wanted neater wiring and better solder joints, a more elegant bias supply, CC resistors and pushback wire, and to implement this ground scheme.

I've annotated the photos. The 3 white wires are the ground wires running to the doghouse. The 3 red wires are the three B+ wires from the doghouse, and I've tried to keep them separated. (They do run parallel for a few inches, but they are about a half-inch apart.) The yellow wire running from the top left of the board to the power amp ground is the + side of the bias filter caps.

[gfarina55]

So, how to proceed? Should I simply revert back to Rob’s grounding scheme? Or is there some middle ground? For example:

-I can easy connect the Presence pot ground (now connected to the neg side of the PI filter cap via a long wire) directly to the power amp ground, and this will eliminate one wire running to the doghouse. Would this be an improvement, and if I do it, should I “re-separate” the neg side of the PI and Screens filter caps?

-Should I keep the power amp ground and pre amp ground separate, or should I re-connect the pre-amp ground to the neg side of the PI filter cap as Rob does?

-Should I keep the bias ground connected directly to the power amp ground (yellow wire in photo)?

-If I take the above steps, there will only be two ground wires running to the doghouse: the HV CT to the Plates cap and the return from the Screens cap. Would it be an improvement to keep these, or should I just reconnect the HV CT directly to the power amp ground?

[sluckey]

You should simply quit obsessing about this. If there was only one good/successful grounding scheme, every amp would be using it.

Follow one basic rule... Keep power amp grounds close to the PT and keep preamp grounds away from power amp grounds.

There are many different successful grounding schemes that share this basic rule.

[Willabe]

You should simply quit obsessing about this. If there was only one good/successful grounding scheme, every amp would be using it.

Follow one basic rule... Keep power amp grounds close to the PT and keep preamp grounds away from power amp grounds.

There are many different successful grounding schemes that share this basic rule.

Sluckey's right about this.

So, how to proceed? Should I simply revert back to Rob’s grounding scheme?

No I wouldn't. I explained to you why I would change what I showed you to change.

Or is there some middle ground?

Yes. I already showed you the middle ground.

Now all you need to do is add 3 twisted pairs of wire to what I already showed you;

There's rules of thumb for lead dress. One of them is to use twisted pairs of wire when/where you can. It's just good practice when doing a build. We know that 2 wires of opposite polarity or opposite phase if   -tightly and evenly-   twisted together will greatly help cancel any noise and/or magnetic fields from being radiated/induced into other wires and other adjacent circuitry. That's why you see PT and OT wires twisted together.

You have 3 B+ filter cap node wires twisted together and their 3 negative return wires twisted together. You should have twisted them together in their respected pairs, A node, B node, and C node. That should help keep the B+ wires from inducing any magnetic field from current and noise into each other, just good lead dress practice.

I would have also twisted the -bias tap and it's ground together too. (You don't have to do that, you should be fine, very low current. And the yellow push back wire wont twist up well with the PT -bias tap wires insulation.)

Also, in the pic you posted of the dog house eyelet board, it looks like you have the caps in backwards? The B+ A node, stacked caps for the OT CT should be closest to the PT. Looks like you have them on the input jack/preamp side? If so that lengthened the filter cap wires that have the most current going through them, B+ A node; OT CT and B node; screen. That should be fixed, turned around, if that's the way you wired it up.

There's reasons why the power ac cord, PT, and heaviest current is on 1 end of the chassis far away from the sensitive low signal input/preamp side of the chassis. Part of good layout and good lead dress.   

You had the amp gutted, you should have moved the stacked B+ filter caps for node A, OT CT. They have the most current going through them and their +/- leads. They could have gone right in between the PT and the eyelet board, inside the chassis. Would have been pretty easy. You could have left the B node, screens and C node, PI in the dog house.   

-I can easy connect the Presence pot ground (now connected to the neg side of the PI filter cap via a long wire) directly to the power amp ground, and this will eliminate one wire running to the doghouse. Would this be an improvement, and if I do it, should I “re-separate” the neg side of the PI and Screens filter caps?

No, leave it like I showed you.

-Should I keep the power amp ground and pre amp ground separate, or should I re-connect the pre-amp ground to the neg side of the PI filter cap as Rob does?

I'd leave it separate.

-Should I keep the bias ground connected directly to the power amp ground (yellow wire in photo)?

Yes.

-If I take the above steps, there will only be two ground wires running to the doghouse: the HV CT to the Plates cap and the return from the Screens cap. Would it be an improvement to keep these, or should I just reconnect the HV CT directly to the power amp ground?

No, I wouldn't. Do it like I showed you. Just add the 3 twisted pairs for the B+; A/B/C nodes. 

[Willabe]

And the wiper wire and ground wire on the Mid pot should be twisted together and moved away from and/or cross the B+ wires at 90 degrees.

You don't have to do this, just another example;

Same with the volume pot wires. Twist them together, use 3 wires on each pot. Add a separate ground wire to the pot that doesn't have 1 going to the ground buss. Twist the 3 wires together, solder the ground wire to the buss right at where the wiper wire/center lug is on the eyelet board. The signal is still pretty weak here, twisting the wires helps protect them from noise being injected.

You will have to put a small mark on the wires with a marker so you know which wire is which. I use several different colors of push back or any wire for this reason. And just to be able to see what wires going where. 

[gfarina55]

Ah, the filter caps are indeed backwards. Working from Rob’s layout and it’s difficult to understand the orientation of the cap board in relation to the rest of the amp.

To clarify your advice Willabe, the three B+ nodes correspond to screens caps, plates cap, and PI caps, correct? You are saying that I should keep all 6 wires running to/from the doghouse, but I should just twist the white (-) and the red (+) wires together into three pairs for each cap (or set of caps). So after I reorient the filter cap board, I will twist together:

* The red PI/B+3 wire and the white wire to the Presence ground
* The red screens/B+2 and the white wire from the neg side of the screens/B+2 cap that goes to the power amp ground
* The red plates/B+1 wire and and the white HV CT wire

So three twisted pairs of red/white wires, each one for a single (or set of) cap(s) running horizontally to/from the doghouse to the power amp ground, in the corner above the circuit board, correct? And this will be better than the "trad" layout even though these 3 pairs are running parallel to each other?

"You had the amp gutted, you should have moved the stacked B+ filter caps for node A, OT CT. They have the most current going through them and their +/- leads. They could have gone right in between the PT and the eyelet board, inside the chassis. Would have been pretty easy. You could have left the B node, screens and C node, PI in the dog house."   

In your first sentence, you mean “PT CT”, correct? I can still move that B+1 cap set. I’m trying to keep a somewhat traditional layout, but I’ll think about this. Regarding twisting the wires on the pots, maybe I’ll do everything above and go through the power up procedure, and if there’s still some hum I’ll attack the mid and vol. pots.

[Willabe]

* The red PI/B+3 wire and the white wire to the Presence ground
* The red screens/B+2 and the white wire from the neg side of the screens/B+2 cap that goes to the power amp ground
* The red plates/B+1 wire and and the white HV CT wire

Yes, yes and yes.

So three twisted pairs of red/white wires, each one for a single (or set of) cap(s) running horizontally to/from the doghouse to the power amp ground, in the corner above the circuit board, correct? And this will be better than the "trad" layout even though these 3 pairs are running parallel to each other?

Yes they can stay where they are, just twist them up nice and evenly/tightly. Start twisting them right from the caps in the dog house and push the twisted pairs through the chassis grommet.

It's not better than the Fender layout, you haven't changed the layout. It is a better way to wire up the grounds than robrobs. I've already explained why I think this. And just better lead dress with twisted pairs, which I explained too. 

In your first sentence, you mean “PT CT”, correct?

No, I mean OT CT. Those 2 stacked caps are node A (B+1). They feed the OT CT that feeds the power tubes.