Cathode Bias EL34 - Output Wattage?

[isaac]

Hello,
I am building a Marshall 1974x preamp into a el34 PP power section. How did you hook up your power tubes? I just started a week or so ago, and ran across this thread, funny we are doing basically the same thing? Are there any big things needed? Can I use the 1974x power supplies and everything, so far I've been ok, but just figured out a rectifier, because i hadnt been successful in getting enough voltage but I should be fine now. Is there a layout of anything that you used? Any help would be very helpful.
-Thanks,
-Isaac

[mikeboone]

Issac... What I ended up doing was copying the Power Supply architecture from a JCM800 to operate the EL34's. I am not an engineer so I stick to tried and true methods for most stuff and tinker with values to get were I need to be.

I dont think the PT for an 18W will source enough current and may not provide adequate plate voltage for the EL34's. The heater supply also has to be ale to source 5A or your PT will get really hot. I tried a PT from a Bassman Reissue and had the heat issues. Triode Electronics sells transformer sets for Marshall style builds at very reasonable prices. I built a Plexi kit from there and it is AMAZING!!! I Digress...

To configure your power tube section, you can copy the architecture of the Matchless Chieftain or Bad Cat Hot Cat 30.

For what its worth, here is a clip of the final product... PPIMV - YouTube

[isaac]

How did you do the biasing. in the jcm 800 it looks like it has no bias trim pot or anything. Did you need all the extra capcans. I have a transformer which should provide correct voltages. My dad builds amps, so eventually once he gets time he and I should be able to sit down and figure out everything. Was your soundclip only guitar and amp or did it have pedals, that was quite a bit of drive for bedroom levels.
Thanks.

[mikeboone]

I only adopted the power supply from the JCM 800 as I knew it would support the EL34's from the perspective of sourcing the current. The output tube section is cathode biased so there is no trim pot. The cathodes of the power tubes are not grounded as in a fixed biased amp. There is a resistor and capacitor between each cathode and ground. Its "self" biasing so to speak. I used the same values as Matchless.

This amp of mine is a mutt. It has the power supply of an 800, the preamp and PI of a Blackface Fender (with tweaked values to kill low end flub), a post PI master volume, a cut control similar to a Vox AC30 and the power tube section from Matchless Chieftain.

I screwed around with it and somehow backed into a really cool sounding amp thats not quite like anything particular, it just kinda kicks. Dumb luck!!!

The PPIMV is how I get the high gain at low volumes. However, the PPIMV (LarMar I think they call it) is thinner at low volumes, not bad but not as fat. The best feature of this amp is that I can get into a sweet spot in the power tubes at nearly the perfect volume for playing a club.

No pedals on the clip. Just the guitar straight in. The changes were pickups and guitar volume control. The guitar is an Ernie Ball Axis Super Sport.

[eschertron]

Nice tones on the audio clip. Is that recording through a 1x12" open cab, red fang?

[Gingertube]

On the weekend just gone, I finished rebuilding my "Junkbox Trainwreck".

Output is cathode biased 6CA7/EL34 as per:
- individual 270 Ohm with 200uF bypass on each cathode
- shared 470 Ohm % Watt screen resistor
- 3K4 Output Transformer (copy of an early Marshall 50W O.T. made by a local winder)
- B+ = 380V

For calculating bias resistor power, assume that cathod ecurrent (= anode + screen currents) will be 120mA per tube at maximum signal. For the 270 Ohm indivicual resistors that means 3.9 Watts, That was a bit too close to 5W for my liking so I used 10W resistors.

Screen current at max signal is 19.5 mA per tube, that has to be managed. That high screen current (compared to some other output tubes) is why EL34 have a reputation of having fragile screens. At the B+ of 380V which I use then the common 470 Ohm 5 Watt screen resistor is fine. The common screen resistor is an old Mullard trick, in a push pull amp as the screen current goes up on one side of teh oush pull it goes down on the other side (and vv) resulting in a toatl screen current which is reasonably constant. Using a common screen resistor results in a stable screen voltage.
If running higher B+ voltages then you neeed to start using individual screen resistors starting at 2 x that 470 Ohm value and gouing higher as you use higher B+

Just below the onset of clipping I measured 32 Watts RMS into 4 Ohm dummy load, datasheet says 35 Watts.

Cheers,
Ian

P.S.
Why did I rebuild the amp? It was a quick an dirty conversion of an old PA Amp., seriously bad hum (from random grounds) and hiss AND when volume control was turned passed about 7 of 10 it broke into oscillation. During the re-build I decided that may have been due to bad lead dress (the reason I decided to rebuild it) but was probably due to a lack of any power supply decoupling between the power amp phase splitter and the final gain stage in the preamp (added during the rebuild). Also identified a microphonic tube in the input stage and replaced that.

The amp is tight and punchy and when you pull harmonics on the guitar they just hang in there for ever - I love it. Testing the power amp by itself showed that it was VERY HiFi'ish -3dB at abot 25Hz and somewhere around 45kHz.

[woodyc]

...Will I be cutting it too close with a single 50W speaker?...

Yes. Unless you play really conservatively, use two...

- With a distortion pedal, your output can be close to double the amp power rating.

- The speaker will fail some time after you reach the low frequency excursion limit, which occurs at WAY LESS power than the "power rating" manufacturers provide (the voice coil heating limit.)

- Besides distortion, other risk factors for speakers include open or vented cabs, drop tunings, pentode output stages, tubes with little or no feedback, and transistors with current feedback.

[kleuck]

Well, AC30 with blues are 33 watts amps equipped with a pair of 15 watts speakers....

[woodyc]

Well, AC30 with blues are 33 watts amps equipped with a pair of 15 watts speakers....

The recone kit is currently available from CEDist for $78.12.

Actually, for speakers with paper voice coils the excursion limit and heat limit are not as different as they are for kapton, but the power rating is still bogus for the other reasons I mentioned. In any event, if you think AC15s and AC30s don't blow speakers, you are mistaken.

[mikeboone]

Nice tones on the audio clip. Is that recording through a 1x12" open cab, red fang?

Thanks... That amp is built in a 1x12 combo. Its built in the gutted chassis of a ragged out Fender Blues Deluxe. The speaker is not the Red Fang I was going to use. First the Red Fang wont fit in that cabinet, too deep. Second the Red Fang has way too much high end respnse for a high gain amp. In the recording is actually a 150W Black Shadow 12 that come out of my Mesa Mark II C+ just to test it. The high gain brings about some unwanted high end artifacts that seem to be filtered best though this speaker. I tried a few others but this one works best. Its a pretty dark speaker.

[jbefumo]

Can anyone comment on the configuration described in the above attachment? It claims 40W, but those current figures seem kind of high to me.

Thanks

[Chuck H]

The end of a necrotic thread probably isn't a good place to put this post. Probably better to start a new thread. I'll bite though...

According to the numbers the total dissipation for each tube in the 40W output model is 24W. So class A, pretty much. The tubes would likely operate in class A until very near their maximum output. Note how the screen current jumps a lot more in the 40W model than it does in the 35W model. A compromise had to me made at 450Vp, nudging to the cooler side of class A to avoid over stressing the tubes. 60mA doesn't seem unusual for class A with a big bottle. The 40W example does seem a little odd because of the high-ish distortion figure compared to the 35W model. Still lower than one could expect for AB1 though and that is probably why the 40W class A model was offered. It gives designers the opportunity to choose their preferences for power and/or distortion at 450V.

[jbefumo]

Thanks SO much. I've reached the point where I'm pretty comfortable with everything up to the phase inverter, but from there back, kind resort to a bit of hand waving.

Haven't found any similar published configurations for the 6L6GC, but (guessing again) that substituting would offer a bit more of a safety margin?

It's going into a minuscule 1X8" combo, so the power figure is mainly a bragging-rights thing for punching above it's weight class. (The prototype turns out 24W using a pair of 5881s, and is remarkably capable.)

Thanks again.

The end of a necrotic thread probably isn't a good place to put this post. Probably better to start a new thread. I'll bite though...

According to the numbers the total dissipation for each tube in the 40W output model is 24W. So class A, pretty much. The tubes would likely operate in class A until very near their maximum output. Note how the screen current jumps a lot more in the 40W model than it does in the 35W model. A compromise had to me made at 450Vp, nudging to the cooler side of class A to avoid over stressing the tubes. 60mA doesn't seem unusual for class A with a big bottle. The 40W example does seem a little odd because of the high-ish distortion figure compared to the 35W model. Still lower than one could expect for AB1 though and that is probably why the 40W class A model was offered. It gives designers the opportunity to choose their preferences for power and/or distortion at 450V.

[Chuck H]

Since you're building a very small combo, keep in mind that class AB1 will create a lot less heat at idle. Which is where the tubes will be most of the time. Since you want disproportionate power for the amps size I would skip class A models and go AB1 for sure. Though the example above doesn't expressly say "class A", maybe because it's not quite strictly, it pretty much is. 24W from a pair of 5881's may also indicate close to class A depending on you're plate voltage. I might consider that a pair of 6V6's in AB1 at 400+ volts would give 20W (which is pretty much the same volume) with less idle current (heat) and give you a little more room in the cabinet (to make room for the speaker) because of their smaller size.

[jbefumo]

D'OH! Thermodynamics 101. I knew that .... made a mental note to consider it early on, then forgot amidst all the sexy electronic stuff! Well, was getting ready to stress-test it into a dummy load, so will have to include some temperature probing as part of that ... THANKS!