5F6A+5E5A Hybrid - V2-B Excessive Voltage

[ululufut]

I'm building a 5F6A circuit within a tweed Pro Amp chassis, with a modified tone stack to account for the lack of a mid control. The layout I'm using is attached.

The V2-B cathode voltage on pin 8 is around 200V, and the pin 6 plate is running at about 325V, while the V1 and V2-A plates are right around 200V, and the V1-A and B & V2-A cathodes are where I'd expect.

When the amp is taken off standby, there is a weird momentary phasing sound, similar to the sound of a vintage CRT television when it is turned on. After turning the amp on and off a few times to troubleshoot it, it started producing a significant hum on standby, which faded to a barely audible hum after warming up in standby mode.

The leads all appear to be in the places indicated on the layout, so I'm wondering if there is something wrong with my layout. It's pretty much the same as the Hoffman 5F6-A, except for the tone stack. Also, the following are the power supply filter cap values:

A - 47µF

B - 30µF

C - 30µF

D - 22µF

Each cap is run in parallel with a .1µF film capacitor.

What am I missing?

[Pietro]

i had the old television buzz sound in an amp once. i think in my case i had used the wrong prim. winding on the PT. 115 v instead if 230v. overloading can cause that sound and excessive voltage is a shure thing.

in your case when it happens when flipping the standby switch could indicate a heavy current draw on the HT winding. B.t.w.  what mains fuse are you using?

for now...do further testing without tubes and measure the heater voltages.

[ululufut]

I got essentially zero voltage on the heaters with the tubes removed, which is actually what I would expect, unless I'm misunderstanding something. I was somewhat surprised, however, that with the tubes removed, the voltages on all of the preamp and output tube plates was about 424V.

If the supply voltage were wrong, I would expect to see issues appear everywhere, but it was only on the second half of the V2 tube where I noticed the higher voltages.

[Pietro]

then i think you are misunderstanding something.
Heater voltage should be present at any time as soon as amp is turned.

with tubes removed you should measure slighly elevated heater voltage of 6.8 to 7.0 approx.


post a picture of your sec psu section and filters. something might been seen...

[ululufut]

then i think you are misunderstanding something.

That very well may be. I measured from 4&5 on the preamp tubes and from 2 and 7 respectively on the octal tubes. I noticed something strange - at least to me - when I did the last meter reads, and this is that the plates all registered about 424V across the board, even in standby mode. This did not happen when the tubes were installed.

The tubes are all glowing as expected, and nothing unusual in terms of appearance, so the heaters are definitely getting current.

There isn't room in the 5E5 chassis for the filter caps with a 5F6-A turret board in there, so I had to retrofit them on the backside of the unit.

I'm happy to upload more images if these aren't sufficient.

[jjasilli]

Schematic is essential.  Layout not good enough


Solder joints on caps look suspicious. 


I would remove the bypass filter caps for now.


The holes in the chassis need grommets!  That's unsafe.

[sluckey]

Re: 5F6A+5E5A Hybrid - V2-B Excessive Voltage

V2B is a cathode follower. It's supposed to have BIG voltages on all three pins.

[Willabe]

You HAVE to cover those big filter caps on the back of the chassis!

That is NOT safe!

And, agree;

The holes in the chassis need grommets!  That's unsafe.

[davidwpack]

I 3rd the grommets. Do Not put it off. Very unsafe!

[ululufut]

Schematic is essential.  Layout not good enough


Solder joints on caps look suspicious. 


I would remove the bypass filter caps for now.


The holes in the chassis need grommets!  That's unsafe.

Yes - I agree about the grommets. I didn't have any left but am intending on installing them and rewiring it, before the unit is installed into a cabinet.

The filter cap solder joints are clean and flush underneath, with plenty of clearance from the chassis - I agree on pulling the bypass caps while it's not working. I'm less suspicious of the filter section than unknowns, in that I'm getting the predicted voltages everywhere, save for the cathode follower - as Sluckey pointed out, that apparently is normal. I knew it would be higher than the other values, but I wasn't able to find any documentation on the relative difference in relation to the other preamp plates, so I was suspicious.

I've uploaded the schematic - not 100% sure I documented the tone stack exactly right, but hopefully it's close.

[ululufut]

You HAVE to cover those big filter caps on the back of the chassis!

That is NOT safe!

And, agree;

The holes in the chassis need grommets!  That's unsafe.

Re: grommets - yes, I got impatient with testing the circuit & no grommets in the drawer; have every intention of adding them before putting the amp to use.

Re: cap cover - I have a cap for it, although I'm surprised that would be considered so unsafe, unless you're referring to them being exposed when installed into a cabinet. It is uncovered only while testing the circuit, unless that inadvisable as well, in which case I will kick the habit.

[ululufut]

You HAVE to cover those big filter caps on the back of the chassis!

That is NOT safe!

And, agree;

The holes in the chassis need grommets!  That's unsafe.

It was actually like 385V earlier, which leads me to relate an earlier mistake I fixed right after powering it up - I'm afraid it might have messed some things up, so I'll fess up, and make a rhyme in the process. I originally, and quite stupidly, connected the 8Ω and 4Ω OT leads to the two output jacks respectively, and - get this - also had them paralleled. Worse yet, even after checking and marking off all leads, I went back again to find that I had V2 pins 3 & 7 incorrectly connected, one lug off in the same direction for each. After fixing both right away and checking again, the hum.noise issue actually seemed to worsen - but the plate reading went from 385V to 325V.

Just in the last hour I measured all the resistors to see if anything was burnt out. I did discover that the two V1 mixing resistors are each reading at 150k, which is weird. They say "270k" right on the units, so I know it's the right value resistor. Same with the 27k feedback resistor - that read at 4.5k.

I'm going to swap the three resistors for new ones, but I'm hesitant to power it up again until I have a better idea of what the issue is.

[sluckey]

but the plate reading went from 385V to 325V.

Look at the schematic. You will see that V2 pin 6 connects directly to B+ node D. Quit worrying about it. It's fine.

Just in the last hour I measured all the resistors to see if anything was burnt out. I did discover that the two V1 mixing resistors are each reading at 150k, which is weird. They say "270k" right on the units, so I know it's the right value resistor. Same with the 27k feedback resistor - that read at 4.5k.

You cannot measure those resistors accurately in circuit. To get an accurate reading you MUST DISCONNECT ONE LEAD!

Disconnect the feedback wire from the speaker jack. Sound any better?

[PRR]

> the two V1 mixing resistors are each reading at 150k, which is weird.

Seems un-weird to me. If both Vol pots are full down we expect both 270k in parallel, 135k. If we assume Vol pots are "slightly cracked" to give 30k resistance wiper to ground we get 150k on a nose. Diming the pots gives less effect and "slightly better readings". It won't be "exact" until you lift some legs. And since the reading you have is approximately what we should expect, I'd ignore it. Until you have happy signal yet odd mixing.

The 27k has a shunt path through OT and tail resistor.

Tube-out voltages are ALWAYS higher. All power supplies sag under load. Tube amps often sag significantly. Tubes are the cause of sag on B+. No tubes no sag.

[ululufut]

That makes sense - just looking for a culprit - at least in the meantime I'm learning a few things

[Willabe]

Re: cap cover - I have a cap for it, although I'm surprised that would be considered so unsafe, unless you're referring to them being exposed when installed into a cabinet. It is uncovered only while testing the circuit, unless that inadvisable as well, in which case I will kick the habit.

No, referring to being uncovered when installed in cab.

I don't see any screw mounting holes for the cap cover in your pictures. 

[jjasilli]

Yes, this is called a doghouse, typically used by Fender.  See pics.


BTW: there's no need for all those filter bypass caps (in a guitar amp).  Bypassing the last stage is sufficient.  A good rule of thumb is for the plastic bypass cap to be 1/100 the value of the electrolytic cap.  So a 20uF cap typically would be bypassed by a 0.22 plastic cap.  I start there, and A/B test trying 1/2 or 2X that to check how it sounds.  YMMV; but you may want to A/B the bypass caps on the other filter stages to see if you hear a difference.

[ululufut]

Yep I've got the box - I realized that I had drilled the holes for the cap wires too close to the OT, so since I didn't have any grommets on me an was anxious to test the circuit, I went ahead and wired it up with the intention of re-drilling the holes and drilling out the holes for the box & wires later (it's a Pro Amp chassis so none of these holes are preexisting). I didn't think to see if I could have picked up some grommets at Home Depot so I was assuming a 3-4 day hold on things while the grommets were on order.

[shooter]

yes, I got impatient with

fwiw, once you overcome this problem, your amps will fire-up 1st time every time 92% of the time 

[ululufut]

I think I'm on the way to figuring this out: (see images)

1. I found one clear mistake that I missed when I did a checklist on the chassis wiring - I attached one side of the 15k bias divider to the 47k resistor instead of the 10µF cap. - FIXED

2. Referring to the attached layout (doesn't reflect some component value changes I made but it's irrelevant here), I realized there may be significant grounding issue. I'm using an Allen TP40, which has a heater center tap. However, when I wired up the turret board, I included the 100Ω heater resistors, and the lugs are wired together as they are in the diagram.

QUESTION:

A. Should a ground wire be attached to the rightmost 10µF cap? If so, it is confusing to me as to why a ground would be attached to the positive end of a polarized cap, but that is how it is illustrated in the diagram.

B. I assume that if I were to remove the heater resistors, I could attach the ground wire to the rightmost lug, according to the diagram. However, if I keep them in there, do I need to cut the lug wires that attach it to the other components?



3. The layout diagram includes the label "bias range" pointing to the 1k/3W resistor. Is this merely an identifier, or is it an indicator that the resistor should be attached to something?

[shooter]

not sure what your hybrid schematic looks like, here's the link for 5F6A, it should answer questions on bias questions

https://el34world.com/Hoffman/files/Hoffman_5F6A.pdf

filaments, pick ONE; CT-NO R, no CT R

[ululufut]

not sure what your hybrid schematic looks like, here's the link for 5F6A, it should answer questions on bias questions

https://el34world.com/Hoffman/files/Hoffman_5F6A.pdf

filaments, pick ONE; CT-NO R, no CT R

Yep, that's the same layout I attached, with the modification of the tone stack to account for the lack of the mid control on the Pro Amp chassis I'm squeezing the Bassman circuit into.

I understand the one-or-the other idea re the filament grounding. What I'm not sure about is, given that I'm using the center tap, don't the components framed in red in the below attached illustration still need to be grounded? If so, I assume the easiest way to do it is to yank the 100Ω resistors and attach the ground wire in the same place as indicated in the diagram.

That leaves the academic question as to how it is that the positive end of the bias filter caps would be part of the ground scheme. But I assume there must be a reason - and to paraphrase Tom Hanks in "Saving Private Ryan", if attaching the the + end of the bias filters gets me an amp that works, then that's my mission!

[jjasilli]

It's not Ok to analyze a circuit from a layout diagram.  A schematic is required.  Layout diagrams are useful for inserting components during a build; but they do not clearly indicate actual circuit connections.

Bias voltage is (-), not (+), like B+, for example.  (-) circuits require reversed cap polarity; hence the (+) side of the cap is grounded. 

I found one clear mistake that I missed when I did a checklist on the chassis wiring
??? Not sure what you mean by a checklist.  Good procedure is to post a Schematic of your amp above your bench.  Trace each segment of circuitry in the amp with a chopstick; then use a highlighter pen on that segment of the schematic to verify that it is correct.

[ululufut]

Bias voltage is (-), not (+), like B+, for example.  (-) circuits require reversed cap polarity; hence the (+) side of the cap is grounded. 

Makes sense.

[ululufut]

Some limited success here -

The wiring of the 15k resistor has been fixed, and I pulled out the 100Ω filament resistors, then grounded the bias section.


A little bit of guitar sound arrived, albeit weak and fluctuating - no, that's not a tremolo:

https://www.dropbox.com/s/u6rf4hkewhrvm8p/5F6A%2B5E5%20-%20test.mov?dl=0


I twisted the bias knob just enough to see if it clarified the signal, which it did, but then the V4 tube suddenly went nuts and turned white hot, so I immediately shut it off. Maybe that wasn't the best idea - either that and+or it exposed some other issue. But I was happy it at least went live for a minute.

[ululufut]

Normal channel not functioning at all though, so I need to figure that out as well -

[shooter]

I twisted the bias knob just enough

since you like the burn n learn method of learning 
a tip, do that test NEXT time WITHOUT the PA tubes in, measure the -(NEG)VDC and adjust the pot to test.  THEN set it to MAX  - (NEG).  ADD 1 ohm resistors from EACH cathode to ground.  NOW you can add the tubes, monitor the 1 ohm R (VDC = bias current in mA), set it to some cold value for the tubes you're using. FIX all your errors till you have constant, reliable guitar sound, THEN do the math and set bias to ~~~ 70% of max plate and evaluate.

[ululufut]

I twisted the bias knob just enough

since you like the burn n learn method of learning 

LOL, you got me - it's either that or endless avoidance techniques sometimes. no burn --> no learn! 

So just to make sure - in avoidance of a li'l bit o' Chernobyl - given my V4 light show (which was spectacular, I might add..), is it safe to follow the regimen you cited, or is it likely that there's another issue that caused/exaggerated it?

[ululufut]

I believe it's the latter, with respect to my last reply. I just shut down the amp after getting a nasty arc while measuring the V4 plate voltage. Then, I took a wire lead to discharge the filter caps, and before I made contact with the first cap, the bare end of the wire accidentally brushed against the OT, and the OT produced a spark. Super frustrating.

[ululufut]

Success! I let it sit for a week or so then came back to it. The amp is up and running - the only thing left to figure out is why the bias adjustment doesn't seem to want to let me increase the current to an acceptable level. With EL34's, I am only able to get about 22mA, and with 6L6's, it's maxing out at about 34mA, right at about 500V on the plates. I'll have to do some research on that.

I'm happy that the treble+bass tone stack seems to work. The overall tone is too bright, although aside from the tone stack, I reverted the cathode and coupling caps back to stock. The break up is a little anemic sounding, although that might be related to the low plate current.

Overall, though I'm very pleased that it's working - thanks for all the help with it!

https://www.dropbox.com/s/obbw685syc85r63/FARAD%20Ultra%2B%20TC.mp4?dl=0

[sluckey]

The amp is up and running - the only thing left to figure out is why the bias adjustment doesn't seem to want to let me increase the current to an acceptable level. With EL34's, I am only able to get about 22mA, and with 6L6's, it's maxing out at about 34mA, right at about 500V on the plates.

The bias circuit was designed for 6L6s. To use EL34s you will need to INCREASE the value of R36 (aka Bias Range on the layout). Start with 2K and go up/down to find the proper range for EL34s. You may not find a resistor value that will work for 6L6s and EL34s.

Or, you could DECREASE the size of R35 to work with EL34s. Might have better luck finding a resistor that will work with both tubes.

[ululufut]

The bias circuit was designed for 6L6s. To use EL34s you will need to INCREASE the value of R36 (aka Bias Range on the layout). Start with 2K and go up/down to find the proper range for EL34s. You may not find a resistor value that will work for 6L6s and EL34s.

Or, you could DECREASE the size of R35 to work with EL34s. Might have better luck finding a resistor that will work with both tubes.

Thanks - yeah I was thinking about decreasing R35. The range was too low for the 6L6's as well - I'll experiment with it and see if I can find a balance. As long as it works for one of the tube types, I'm fine.