Vk = 112 seems all wrong for 6L6

[Phil_S]

You may remember that someone gave me a couple of old PA system amps. I started with the RCA, what I judged to be the lesser of the two amps, but I thought I liked the iron. It was running a 5U4 and a pair of 6L6. There is a choke. The circuit as built [1] was incomprehensible to me. In any case, most of the wiring had sticky stuff on it. I decided it was to be gutted and converted.

[1] I believe someone may have modified it from the original and not done it successfully.

Initial unloaded voltage on the PT was ~420-0-420, which I figured was high for a pair of 6L6, still in the relevant range for a 5U4. Even with a 5U4, I estimated B+ around 480. I reasoned the PT must be a little soft and that I should build out the power supply and power section to see what it might actually show.

There was some smoke (not fatal) which prevented me from taking the time needed to record accurate voltages, so we'll go with what I can remember.

I used a 5U4GB which appeared to be up to the task and a pair of Russain 6L6 wannabes (6p3c or 6p3s depending on translation). Under load, I saw the HT secondary drop to around 810VAC, B+ at the 1st cap 376V, and at the 2nd cap 373V. Vk=112...I shut if off quickly because of this. Things may have dropped more if I had be able to leave it powered for more time.

For testing, I decided to go with cathode bias, as it's much easier to build. I put 282z (measured) on the cathodes with a 25u 25V cap. In short order the 25u cap blew up. I figured correctly that the cathode was above 25V, see above.

The choke (pair of red wires entering the chassis lower left) is wired between the plates and the screens. Screens have 470z stoppers. (The pictures show a wiring error -- screens are connected to the wrong can cap terminal - thin red wire - and that's been corrected.)

The can cap is recycled 40-40-40. Yeah, I know I shouldn't be using it, but it is rated at 525V for all 3 sections and the only thing I had around that I felt would be up to the job. I didn't want to use 20u 450V caps, figuring that would be an endless headache. Indeed, on power up, B+ jumped to ~528V initially.

There are three pictures, an overview and a couple more close up.

The PT terminals are:
1-3 Primary (black, white)
2-4-6 HT Secondary (red-black-red; black to chassis ground - hard to see at lower left)
8-10 Filament 5V (green-green)
5-7-9 Filament 6.3V (white/orange-white/orange/black-white/brown)

The OT terminals are:
7-10-9 Primary, B+ to #10 (#7 blue, #9 white). The terminal strip near #10 is not grounded though it looks as if it might be. Both lugs are connected.

1-Com
3-70V/164z
5-500z

2-Com
4-4z
6-8z
8-16z

I don't understand how 410-0-410 into a 5U4GB results in B+ 376.
I don't understand Vk = 112.

Are the unconnected grids on the 6L6's a problem?

I must have done something wrong. I don't see it.

I'd be grateful for some help! Thank you.

Phil

[ampgeek]

Hey Phil,
The control grids on the power tubes need a ground reference to bias up correctly. Looks like there was over ~500 ma flowing through the cathode on each tube. Did you see any redplating?

I would stick a ~220K resistor between each control grid and ground and give it another shot.

Good luck!
Dave O.

[Phil_S]

Dave O, thanks! I had a feeling that was the problem. It's doing much better. BTW, with that pre-burn smell and the voltage readings on two meters, I never had time to look for red plates. Power was on for a very short amount of time.

HT supply and B+ stabilized quickly.
HT supply 819VAC.
B+1 449V, Plates 448V
B+2 449V, Screens 447 or 448 ... the meter was waffling
Vk 37.7V
Calculations give me:
Ia 132mA
Ik 134mA
Plate dissipation 54W

I was hoping for something a little less high voltage, but, as they say, it is what it is.

Without spending an arm and a leg, any suggestions for a pair of power tubes? I'm doubtful the 6p3c (not 6p3c-e) can do very well at such high voltage. Tubes can include something off the beaten path. The amp came with two 6J7 preamps from which I've cut out and kept the top caps, so anything goes.

The OT is ~6600 primary.

[old tube]

I'm a bit of novice so if I'm talking garbage, please go easy. I just built a 5C5 circuit using an old British PA chassis and iron.

Shouldn't there be a dropping (10K?) resistor across the two sections of the can with pin 4 being fed from the lower voltage after the resistor? That's what I needed to do for the 5C5 circuit.

Also would the choke go across both sections of the can? I used the choke in the amp I built even though the circuit doesn't call for it, but like you, I was worried about the B+ being too high and I included it firstly because it was there and secondly because it took the B+ down about 20v in my case. I used the method fender used in the later tweeds and used it to "parallel" the first two 16uF cans in the chain.

[Phil_S]

As noted above...my verboseness causes people to skip stuff...the picture shows the choke miswired. I addressed that. So, no dropping resistor, used the choke instead, but it doesn't drop much voltage There is a really big 23K resistor that I may put in series after the choke.

You are right about the caps. I didn't have a pair of appropriate value. From what I see, I'm OK with the 525V rated can cap as long as it doesn't give out. If I actually build an amp in there, I'll either use 600V rated or a pair of 350V rated.

Remember, this was just a power section test to determine what I've got to work with, which turns out to be more voltage than I really want.

[old tube]

Sorry for not reading your post properly. I must have been writing my post when you were posting that you'd fixed it.

A question is do you need grid resistor in place because it's only a partial build? I don't have anything like that on the 5C5 and it's fine.

I was surprised while building mine how much the preamp section (3 x 6SC7's) pulled the B+ down once I had it all wired up. I was really lucky and ended up with 390V on the anodes of the 6L6's and rest of the amp is pretty much exact spec for voltage.

[Phil_S]

A question is do you need grid resistor in place because it's only a partial build? I don't have anything like that on the 5C5 and it's fine.

Some kind of reference is needed by the grids. The 5C5 uses a paraphase inverter. Look at where you see a pair of 250K with a 6800 between them. Those are the grid leak resistors, just not grounded they way we'd do it when fed from a LTP inverter.

[old tube]

Many thanks, I think I understand that too!

Good luck with the amp.

[ampgeek]

Great! Glad that it helped!!

Rock on,
Dave O.

[Phil_S]

Update to the testing...I decided that large 20K resistor would be of good use if it put in series after the choke. I wasn't happy that the plate and screen voltage were both so high...just a hair under 450V and depending on line voltage on any given day could easily be above that. It's an arbitrary mark, but it does affect the choice of filter caps.

With the 20K in the circuit, Va = 460, screens = 379, Vk = 31.75.

In any case, suspect the power tubes are much happier at the new voltages and I'm certainly relieved to see plate dissipation go from 54W to 47W. Remember, this cathode bias, so 100% would be 20x2 = 40W for these 6p3s tubes. The spec sheet states VAmax = 375 and VG2max = 300, so I'm still over spec. I think I'm going to have to find a pair of real 6L6GC for these operating conditions.

[gingertube]

I see you have 282 Ohms for teh common cathode resistor for your cathode (auto) bias.

I would put separate resistors on each power tube such that matched pairs of output tubes are not required.

282 Ohm common is the same as if we had 2 x 282 = 564 Ohms on each tube.
However you want to bias about ( er 40/47 = 85%) 15% colder.
564 x 1.15 = 648 Ohms.

Try 680 Ohms on each cathode each with a bypass cap. If you think that is too cold then you can try say 2 x 1K2 in parallel = 600 Ohms on each cathode.

Remember that while we normally bias fixed biased amps at 70% max anode dissipation, with cathode (auto) bias you can safely go much higher than that and bias at 90 to 95% of max dissipation if you want.

Cheers,
Ian

[Phil_S]

Ian...thanks for the thoughts. This is just set up as a testbed for the PT as I wanted to determine the relevant range for B+. I had no sure way of knowing what it was going to produce. The testbed circuit produced all the information I was hoping for and more.

I'll be building a fixed bias amp in this chassis, probably a slightly modified blackface Bassman. I'm thinking I really need to find a pair of real 6L6GC that are built to take 500V. I have several singles on the bone pile and I might build it with two bias supplies to allow for balancing.

I'm not looking to invest too much in this amp. As it is, I'm finding I have to order parts that I don't have. It's going to be too loud (maybe 50W) for me to use. I don't play out. This one will probably get sold on eBay with proceeds going to feed the habit.