Bassman 100 help: screen resistors fried

[charlie_p]

I have an early 70's silverface bassman 100 that i am modding. I first changed the tube sockets to take el34's. Converting a Fender Bassman Head from 6L6's to EL34's Then i started to rework the balance into a bias control. Also changed out the value of some resistors in the PI section to blackface specs (Tube Amp Book) Bias conversions

Turned it on, let the tubes warm up. Took some measurments. Plate V is 446 so i set my bias to about 35mA. Started playing and everything sounded great. The amp use to start to break up around 6.5/7 now around 3.5/4 it starts to break up. Played for a few minutes and cranked it up to about 8 played one chord. Tube 3 fried and took the screen resistor with it. Turned it off, pulled the tubes turned it on, took some readings.. getting -37v through out the bias circuit and on pin 5. I replaced the screen resistor, looked through the changes i made to make sure i had everything right. Put in some tubes and turned it on. After warm up, started to play a bit at low volume. Everything sounded great. Turned it up to about 7 played for about 5 minutes, turned it up to about 8/9. Hit a chord and BAM. Tube screen resistor fried, this time tube 2 (counting from left to right). Same thing, replaced the resistor, cleaned the tube sockets out, double double checked resistors that i changed in the PI section. Inserted a set of tubes, Turned it on and straight up to 8, hit a chord, WHAM, same thing this time tube 1 fried.
Decided to put the bias mod back to the original balance, and when i went to unsolder the wires i found a that the wire that went to the center lug on the pot coming from the two 220k resister that go to pin 5 was broken and wasn't attached. Could that be all that was causing the problem? I re-soldered it but haven't tried the amp yet. I need to order a new set of tubes first, only have one more complete set here. I want to try it before going back to the balance pot.

At first i thought maybe the balance/bias mod won't work with the el34 mod. Then i was thinking maybe the el34 mod was the problem, but have read online that many people have had great success with both. I am at a loss here and input or suggestions would be helpful.

[Enzo]

Is this the sasme set of tubes each time? And are we paying attention to which tube is in which socket? If you only have the one set, at least number the tubes with a marker. You could be moving the bad tube from socket to socket. it certainly sounds like a bad tube.

[Steve Conner]

What value of screen resistors are you using?

What speaker load do you have connected?

This amp has a high screen voltage, and no impedance selector on the OT. It's really going to punish the screens on EL34s, unless you have the output loaded right down to 4 ohms.

The only cure I can suggest is: Don't ever use more than a 4 ohm load, and increase the screen resistors until the tubes stop frying. If you peek through the holes in the plates, you'll see the screens light up like lamp filaments before they actually fry, and that could save you some money.

Or if you want to run 8 ohms, pull one pair of tubes, or rig up a 4/8 ohm switch that disconnects them.

[charlie_p]

Enzo - No, not the same set of tubes, After the first set blew i numbered the tubes. And used the light bulb current limiter to identify which tube was bad. I replaced the bad tube. The first set was a set of jjel34's out of my carvin and where in working order. The second set was a set of svetlanas i had that were good (what i replaced the jj's with in the carvin). The set now is 2 jj's (inner two tubes) 2 Svets (2 outter tubes) just for testing. I have a brand new set of rubys here, but i don't want to put them in until i have this fixed.

I also have a set of 5881's and a set of 6l6's here if i need to use those for testing. I was going to put in a set of these to see if the problem was just with the el34's or in the PI/ Biasing section.

Steve - i am using 1000ohm 2watt resistors. Should i increase the wattage since they are burning up as the volume is increased? Maybe to 5watts? I have the head connected to the 4 ohm jack on my marshall 1960 cab (I ohmed out the jack to make sure it was 4 ohms)

[charlie_p]

After looking at some schematics i found online, most of the amps i have seen that take 6l6 and el34 tubes use anywhere from 700-2kohm resistors at 5w. I think i will up the screen resistors to 1k 5w and see if it holds with the volume increased.

[Steve Conner]

No, I meant to increase the resistance, not the wattage. The wattage rating of the resistor doesn't influence the tubes in any way, they'll still blow. 1.5k or more.

[charlie_p]

Steve - sounds good i ordered some 1.5k and some 2k resistors. I'll let you know what happens in a few days. thanks

[charlie_p]

Well, i put in the 1.5k 5w resistors and i played for an hour or so. but figured i would put in the 2k 5w resistors.. played for about another hour or so (longer than i have been able to). And everything seemed fine. Played for about 20 mins today and the same thing, fuse blows and with it a tube.

should i keep upping the resistor. should i change the resistor on pin5 from 1500 ohm to 5600ohm as on some marshalls running el34's?

I am getting the feeling this isn't going to work...

i know it can be done though, i have seen videos online of other fender BM100s modded to take el34's

what am i missing?

[stokes]

The resistor on pin 5 would have nothing to do with it.I would check that you have all the wiring connected right.Do you have the screen resistor connected between pin 4 and 6 as the instructions state?Now,do you have the B+ to the screen connected to the right side of that resistor?Maybe you have the feed connected directly to pin 4 instead of pin 6? Go over everything and make sure something isnt wired wrong.

[charlie_p]

Stokes - I have the resistors across pins 4 & 6. Pin 6 is jumped to the other pin 6's. Pin 6 also goes up to one side of the choke and a white wire in a bundle of wires i guess to the filter caps (maybe?)

In the phase inverter section the (2) 220k resistors form a v. Where they meet i have a wire going to the center lug on the bias pot. The right lug of the pot has a 15k resistor to ground and the left lug goes to a 50uf 50v cap and a diode.

[charlie_p]

could this be a transformer issue and not a circuit issue? I did have the 6.3v pilot lamp blow not long before i did these mods.

[txstrat]

I'd keep an eye on the bias, double check all connections and everything around it (resistors, wires, cap etc.).
To me it seems like the bias runs away (maybe vibration related) when you play the amp hard and the more sturdy tube lives on while the other goes down and fries the screen resistor. I'd bet if you wait long enough the other tube would go down too (not suggested ).

[charlie_p]

txstrat - is it just because of the age and the parts may be failing. I guess it wouldn't hurt to upgrade everything while i am inside the chassis. If the bias is all over the place, would replacing with new components take case of the issue you think?

Maybe i can clip my leads to the 1ohms on pins 1/8 to ground while i play and see what is going on over a period of time.

[charlie_p]

Some voltages. These are all with the power tubes removed from the amp

Power Secondary
Red - Red = 335 - 335vAC
Grn - Grn = 6.93vAC
Bias (Blue/red) = 61.8vAC (this is right off the transformer before the resistor and diode, schematic says it should be 50v)

Coming off the rectifier i have 445vDC

-68vDC From resistor/diode tied to bias lead from secondary
-68vDC @ left leg on bias pot
-38vDC @ center leg of bias pot
-25vDC @ right leg of bias pot (w/resistor tied to ground)
-37vDC top of 220k resistors that form "V" in PI section
-38vDC Bottom of 220k resistor "V"
280vDC pin 6 PI
44vDC pin 7 PI
70vDC pin 8 PI
284vDC pin 1 PI
41vDC pin 2 PI
446vDC pin 3 Power Tubes (plates)
441vDC pin 4 Power Tubes
441vDC pin 6 Power Tubes
-37vDC Pin 5 Power Tubes

Here are the readings with the TUBES IN:

Everything stayed pretty much the same except these

430/431vDC pin 3 Power Tubes (Plates)
429vDC pin 4 Power Tubes
Bias is 24mV,23mV,22mV,22mV (which seems "cold" but these were brand new tubes i just put in for this test these should be biased around 34/35mV)

434vDC off the rectifier
276vDC Pin 6 PI
42vDC pin 7 PI
66vDC Pin8 PI
279vDC Pin1 PI
39vDC Pin 2 PI
66vDC Pin 3 PI

[txstrat]

The voltages seem OK, +/- 10% from a schematic is no problem.

I wouldn't let the power tubes dissipation go much over around 45mA per tube.
35 to 40mA would still be enough IMHO.

The bias voltage could drift cause of old/failing parts or just because something is wired wrong (or bad solder joints). Double check everything and if you're not sure use new parts for the bias section.