5F4 Tweed Super Mod Build--Shorting with Power Tubes

[ViperDoc]

I built up a 5F4 Tweed Super amp with "only a few tweaks". I'm in the light bulb limiter check phase and there's a short of some kind just when inserting the 6L6 tubes. All power supply, power string and plate, cathode preamp voltages appear in line noting the limiter.

Some useful data:
(All Tweed Super Iron)
PT: Classictone 40-18094
Choke: Hammond 159R 6H 200mA DC
OT: Heyboer APD-8088-H (with 2.6, 4 and 8 ohm secondaries)
Preamp tubes: EH 12AY7, 12AX7
Power Tubes: 6L6 EH
Rectifier Tube: 5U4GBEH

Mods from stock Fender SCH/Layout:
1) Switchable 100pF-250pF Bright cap on bright channel
2) 3-way Switchable LNFB on V2 grid between stock 4.7M, "no connection" and a Bourns 5M trimmer resistor
3) Switchable .02 uF--.001uF OD cap into the PI for handling 10 inch vs. 12-inch speakers
4) 470K PI grid stopper resistor added
5) .01 uF bypassable GNFB cap
6) 470R 1W resistor on PT screens
7) 1.5K 1W grid stopper resistors on PT grids
PT cathode leads to bias test jacks
9) Adjustable Bias pot
10) Manning bias mod
11) OT secondary output selector.
12) 500mA fuses on HT, 6.3V CTs
13) Shaffner AC module with internal fuses
14) Standby switch and 0.05 uF ground reference cap delete.

Schematic of this current build:

5F4 Mod Build SCH 4-18-2021.pdf

Chassis shots:











My limiter bulb is comatose when the chassis is powered on without tubes, more-but-mildly lit with the rectifier tube in, about the same with preamp tubes, then gets quite bright with the 6L6 tubes in, and there is no guitar sound. I get mild buzz through a speaker cabinet when probing the PI plate wires.

The OT center tap is connected to V4 pin 6 which is jumped to V5 pin 6 and the second 16 uF power supply cap as per the Fender 5F4 schematic. Incidentally, that schematic draws the tube heaters connected to preamp pins 4-5 and the pins 9 to ground; I connected my heaters conventionally with the pins 4-5/9 and 2/7 approach.

The tube heater string test intact with 6 VAC across each socket.

The Bias pot is connected to the blue wire coming from the 220K grid leak resistors and the white wire from the 10K 2W resistor in the bias circuit.

The red-green wire is the PT bias tap lead off the PT secondary. With the limiter in place (40W bulb) it measures 34 VAC.

The input jacks all appear to function normally.

Any recommendations would be much appreciated. I posted this in a new thread to avoid further hijacking the 5F4 thread posted by another user. Cheers.

EDITS:
1) The turret connected to the 8 uF cap and V3A plate has been re-soldered.
2) Input pic added.
3) OT primary DMM-tests intact with 51R on one side and 57R on the other side when disconnected.

[Stevem]

I would start off your trouble shooting by first unhooking the C12 side of R29 so that your just working on the output section.

What resistance do you read from pins 8 of the output tube sockets to ground, ground meaning right on the chassis, not some test point jack.

Please confirm that with the output tubes removed that you have V+ on pins 3 and 4 and at least a -47 volts on pins 5.

[pdf64]

Assuming no wiring errors (?), if the output valves are drawing too much current, there’s either insufficient bias voltage at their pin5 control grids, or they’re faulty.

With the output valves removed, I suggest to take VAC and VDC readings for each socket terminal of each output valve.

[ViperDoc]

The voltage on PT pins 5 are between -22 and -27 VDC on the bulb limiter from left to right on the bias pot. I know the target is -40 VDC or so. Off the limiter, the bias voltage will increase a bit (become more negative), right? Forgive the ignorance, I'm just having flashbacks from my last PR build. I don't want to cook the goose too soon. The bias tap voltage reads 34 VAC on the limiter. If that's too anemic, I can pull it from the secondary and adjust R36.

[sluckey]

The bias tap voltage reads 34 VAC on the limiter.

Probably not enough for 6L6s. You can safely experiment to get the bias voltage up to where it should be. I'd shoot for -40vdc to -60vdc range. Just pull the output tubes, remove the current limiter, and experiment until you are satisfied. Then set to max negative. Plug back into the limiter, plug in the 6L6s and start over.

[Stevem]

There's no way around it, as even 6V6 tubes no less 6L6 tubes would need like35 volts D.C. feeding the junction of the 220k resistors!

[xtian]

Brian! Really beautiful work on this build!

What wattage bulb is in your limiter?

Looks like you have a fused heater CT that goes to ground? No problem, just confirming.

[ViperDoc]

Thanks, man. Yes, I put the HT and 6.3V fuses on center taps to ground. They are grounded through the mounting lug of the tag board on the left.

I've been using a 40-watt bulb on the limiter.

I used your headphone trick on the iron so I can't wait to "not" hear that. The noise, that is.

[xtian]

Very good. Sounds like you need to sort your bias voltage first. Carry on!

[Stevem]

You should be on the safe side and test your output tubes for shorts with a ohm Meter.

Pins 3,4 and 5 should not test shorted to pins 2 ,7 or 8.

Pin 5 should not test shorted to 3 , 4 or 8.

[ViperDoc]

I pulled the 6L6 tubes and lit up the chassis without the limiter. I found the following:

Wall AC: 122 VAC
HT: 372-0-372 VAC
Rectifier pin 8: 500V!!!
Bias tap: 54.5 VAC
Reading after 1K8 Bias dropper: 51 VAC
Bias range on grids: -32 VDC to -40.5 VDC

I then put the 6L6s in and lit it up.

B+ settled out at 430 VDC. A 200R 5W dropper off of pin 8 of the rectifier socket gets me to 408 VDC, which is where I'll keep it for now.

The amp sounds good and clean! Maybe even too clean, but I don't want to push it until I get my bias voltage where it ought to be. More progress tomorrow.

[Stevem]

Side note.

The screens are pins 4, the bias voltage is found on pins 5, the grids.

Good too hear that all is likely well!

[ViperDoc]

I ended up bypassing the 1K8 dropping resistor on the bias tap and now I can get 18 mA of cathode current on the 6L6GCs on the coldest setting. This amp sounds really good. And it's (quiet) at idle. *VERY* quiet. Thanks, Aaron! Where do y'all like your 6L6s biased at for tweed amps?

My B+ is now back up to 440 VDC, so I'm going to knock that back down.

[Stevem]

With between 430 and 440 on the plates I would idle them at anywhere up to .047 amp that gives you what you like in regards to a good combo of clean and then a smooth change into clipping.

[ViperDoc]

I put the bias tap drop resistor back in, just as Martin designed and--what do you know--it's better that way. I biased the 6L6s at 44 mA and this amp can do quite a bit. The chime is everywhere on this thing. The LNFB values really change the amp dramatically; that 3-way switch is very useful. I'm still going to drop that B+ and try some 6L6G tubes. Those bottles are huge, I haven't been able to find retainers that will fit them, though.