EF86-->Paraphase PI-->EL84: Power Tube Imbalance (Not the Tubes)

[Commodore 64]

So I've built an amp here, kinda inspired by a few things, including my never having built (or heard) a paraphase PI. Taking a page from Dr. Z (and fred mertz, where are you fred?), I used ultralinear taps on my OT, an Edcor GXPP15 and an EF86 as a preamp tube.

I've attached the schematic.


1. I added the 470k between C7 and volume pot. Much better volume sweep with clean headroom.
2. There's 60Hz hum, doesn't change in volume with the knobs. More than any other amp I've built.
3. Suspecting power tube mismatch (all my 6BQ5 are used/tested). I did some tube rolling. Seems like whatever I do, there's always 6-7 mA difference. And that difference increases, the longer I leave the amp on. This behavior follows the tube socket, not the tube.
4. Amp sounds great, except for the 60hz hum, BTW.
5. Best I can tell V3 is drawing some grid current. It got up to 1.8V before I turned it off.. It's not leaking coupling caps (swapped out and tested 'em). So maybe something to do with the UL taps or OT?
6. Is there any other way this behavior could happen? Would this be a sign that my PI isn't balanced?

Here's some voltages:

V1 (EF86)
1: 173
3: 3.2
6: 105

V2 (12AX7)
1: 176
3: 1.26
6: 179
8: 1.26

V3 (6BQ5)
2: 1.8v (and climbing)
3: 10.8
7: 311
9: 308 (UL tap to OT)

V4 (6BQ5)
2: 60mV
3: 10.8
7: 311
9: 308 (UL Tap to OT)

Still debating on whether to add the MV. The Pro jr. trick of diming the amp and using the guitar volume seems to work pretty well.

Build pics:

Built some cabs. Speaker from an old pine shelf, the other a dovetailed flat-pack I had made at an amish "drawer" building company. An amp cab can be a glorified drawer in their eyes. And I can't buy non-cupped pine boards for the price of the poplar "drawer" ('bout $19). Plus, unlike my cabs, they are always square!

And a little closer to done, with some tolex.

 

[peteb]

Edit delete:Isn't 1.8 more right than 0.06 for screen voltage?> pin 2 is the grid.


Have you measured the resistance of the OT primary's?

Theres supposed to be a little different but not much.


Which is the higher current draw spot? V3 or v4?


Nice work on all parts.



I wasn't able to view the schematic, about the PI and balance, but the PI sends signal, so I don't see the idle current being affected.


You've got balance at the plates, what affects idle current at one spot or another? only the idle voltages of the plate , cathode, and screen? Right?


How is it biased? Edit: must be cathode biased.

 

[Commodore 64]

The schematic is a PDF.

V3 is higher current draw spot. Power tubes are cathode biased.

OT was tested with an RG Keen Ne bulb short tester, and checks out.

I suspect grid current because there's a voltage drop across the 8.2k grid resistor. 1.8 at the screen and 1.3V after the resistor, at the junction of the 0.022uF off the PI.

 

[peteb]

i see the schematic now.


I think you are right in tracking down the source of the 1.8 v on the v3 grid because its making the bias hotter.

v3:1.8 - 10.8 = - 9 vdc of bias

v4: 0.06- 10.8 = - 10.74 vdc of bias


I would still measure the resistance of the OT primarys for future reference, and to calculate the bias.



it would be interesting to ground out the grids on v3 and v4, and check the idle current. the sockets should be identical then.

 

[robrob]

You could remove plate or heater voltage from V3 and see if the grid voltage goes away. If it remains you know the grid voltage is caused by grid current.

 

[Commodore 64]

Poor troubleshooting job by me. The grid voltage wasn't following tubes. I just had 2 out 3 tubes being bad and didn't keep good track of which ones I was swapping in. Both USA made 6BQ5--bought Used/Tested. Not doing that anymore. Not worth pulling my hair out troubleshooting this crap.

A couple times a year, someone gives me a box of their father's old tubes. TV repairmen, etc, seem to never throw out tubes. They always seem put the old one back in the box when they replace it. Just causes headaches when I try to use 'em, especially on new builds.

New production for me, henceforth (because NOS prices are ridiculous).

Or at least, I ain't keeping tubes that aren't the same brand as the box when I sort through "donations".

 

[FenderLover]

When you get the output sorted out, come back and let us know what you think of the circuit. I'm kicking around something similar now that will be more like the Z-28 with UL output (Hammond 1608) and Princeton power.

 

[Commodore 64]

Yeah, it's time to button it up and put some hours on it.

 

[Commodore 64]

Everyone in the band gives this amp a huge thumbs up. Sweet tone that isn't overly bright (using a Dano with Rose P90s in it). Might put a bright cap on a toggle. Lot more power than the ~12W (cathode biased el84) power section would suggest. Though that could be due to the UL taps on the screens. Very touch sensitive, easy to drive the EL84 with picking dynamics.

I'd like to sort the hum issue. It's not really loud, and isn't really noticeable when playing. But I want to figure out the source.

It's 120hz hum (pretty sure). I recorded it. Volume pot has no effect on the hum. Towards the end, I'm rolling the tone pot back and forth--which causes the hum to wobble. The tone knob is kinda in a funky spot (see schematic in post 1)...



This being ultralinear, the screens and plate voltage are referenced to the OT...so the screens don't have the extra filter stage (choke or resistor). I'm wondering if this is a symptom...

 

[tubeswell]

How are the heater pins wired on the EL84s? (In synch or out of synch?)

 

[Commodore 64]

I was looking at those today...and realizing I have no idea. And dreading that I probably need to redo them. I didn't pay attention when I did the heaters... Actually, I never have. I didn't know it was a thing...until recently when I was googling hum issues. Probably useful to use 2 different colors of wire for that reason, huh?

Is it as easy as just swapping the heater wires on one of those EL84 and seeing if it makes a difference? Edit: I just tried that...no change.

In other news...Just swapped in 12AY7, Sino 12Ax7, JJ 5751, and an RCA 12AT7. The higher gain the tube, the louder the hum. I guess that's not surprising. Hum is least with the 12AY7, and the 12AY7 gives a much more useful volume sweep. So that's a keeper just for that standpoint.

 

[Commodore 64]

OK, i got the noise floor of the hum down pretty low. I ended up with a 47uF first filter cap, and a 47-ohm resistor between the rectifier and that first filter cap. LOVE this amp. The UL taps give it just that little bit more of clean headroom, I think. And the power section is so sensitive, responds so nicely to picking dynamics.
12AY7 is perfect amount of PI gain for this EL84 power section, with standard Hi/Lo jacks, can accommodate humbuckers or lipsticks.

Made a trade last night for some various crap-one piece of which is some ported Peavey 1x10 with a Emi blue alnico. Great match for this amp!

I reckon the noise was due to lack of power filtering on the screen, given the UL taps? Not as prevalent on the Bassman 10 (also UL) because that has 2, 220uF filter caps in the first node, being diode rectified and all.