• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

EL84 fixed bias PP Triode

Some time back I was the recipient of an amp that had seen better days but had good output and power trafos.

Given that I have well behaved Fostex single driver speakers of around the 95dB/W ballpark (and access to a local tube amp guru) I've decided to dip my toe into an el84 tube amp.

What we have on the cards is a silicon rectified B+ in the ballpark of 320v with less than 40mV of ripple, a separate bias supply for the output tubes enabling me to run them all in fixed bias. I'm wanting to run them in triode mode for starters, can always do the ultralinear thing later on if needed as the output trafo is sorted for that if I desire but the purist in me would rather not and it's not like I need the power.

Also planning to see how things sound with zero neg feedback, again an easy retrofit if necessary.

Input stages are an Alps 100k switched attenuator into a 12ax7 gain stage, cap coupled to a 12ax7 concertina phase splitter (very similar to the input stages on a Rogers Cadet but minus the neg feedback connection)

Comments? Brickbats?

For the moment I'm planning on limiting the silicon content to the diodes in the psu for the heaters, B+ and output bias, no CCS or LED arrays.

As much a learning exercise and first step as an end point. Gotta start somewhere I guess and if you already have the mains tranny, choke and outputs then you're most of the way there. (other bits on hand include a matched quad of el84's and the alps attenuator)

TIA, Drew
 

revintage

Member
2007-05-30 4:12 pm
Nice setup! With 30mA Iq and 8k primary you should be able to squeeze 4W A and 7W AB out of it. Add 10 ohm resistors between each catothde and ground. But use a better tube than 12AX7 as driver. 6DJ8 would be a far better choice. I would prefer paraphase but that is a matter of taste.
 

ray_moth

diyAudio Moderator Emeritus
2004-01-27 8:55 am
Jakarta
First off, do you know if your OP trannies are the right impedance, coupled with your speakers, for EL84 in triode? They need ot be 10k P-P. For UL, they need to be a bit lower at 8k P-P, so the load for triode is a singificantly higher impedance than for UL. If they suit UL, then you'll probably get more distortion than you really should in triode mode. I any case, the most you're likely to get from triode mode is about 3 to 4 w and since EL84s are especially good in UL, that's the way I'd go if there's any choice.

Secondly, EL84s behave so well (in both triode and UL) using cathode bias that I doubt you'll do any better with fixed bias, though it won't do any harm to try if that's what you want.

Thirdly, unless you're stuck with 12AX7 for the voltage amp and splitter, you could do better IMHO. It's OK for guitar amps but I don't think much of it for Hi-Fi use. I'm also unsure if a 12AX7 concertina splitter would do very well driving EL84s in triode mode. A better tube, with 60% of the gain, would be a 12AT7.
 

revintage

Member
2007-05-30 4:12 pm
ray_moth said:
They need ot be 10k P-P.

Secondly, EL84s behave so well (in both triode and UL) using cathode bias that I doubt you'll do any better with fixed bias, though it won't do any harm to try if that's what you want.

A better tube, with 60% of the gain, would be a 12AT7.

Hi Ray,
I do not agree with you about that 10k for triode unless we are talking about pure class A but then B+ should also be lowered to ca 250V. 8k will work well in triode AB-mode.

My example is based on AB, B+300V, really on the limit for the tubes.

About the two other matters I totally agree even if I consider 6DJ8 better than the 12AT7. But that might also be a matter of taste.

Cathode bias will drop the available voltage over the tubes so this will make everything a little more reliable.

Brgds
Lars
 

kevinkr

Administrator
Paid Member
I've used output transformers with primary Z of 7.8K - 10K PP with triode connected EL84 and have noted only very small differences in performance with real speaker loads and indeed on the bench as well. I would think most recycled EL84 UL/Pentode output transformers will provide more than adequate service in triode connection. I generally run in class A/B, but have built at least one amplifier that was class A.

You should expect no more than about 7W in AB1 and perhaps 5W in A.

I built a 7Wrms pp stereo amplifier on a Scott 299B (7189A) chassis nearly 18yrs ago, it is the only amp I still have from those days - and it still sounds great on the occasions where I use it.
 

revintage

Member
2007-05-30 4:12 pm
As I said in my first post you can expect 7W AB and 4W A with 320V and 8k.

If you want to try different Iq and Raa you can use my calculator:

www.revintage.se/PPABAMP.xls

Use it together with an Ua/Ia-diagram to find your specific working point. Just remember to plot Pamax in the curves before you begin. You can find an Ua/Ia for triode-connected EL84 in Mullards datasheets at www.duncanamps.com
 
ray_moth said:
First off, do you know if your OP trannies are the right impedance, coupled with your speakers, for EL84 in triode? They need ot be 10k P-P. For UL, they need to be a bit lower at 8k P-P, so the load for triode is a singificantly higher impedance than for UL. If they suit UL, then you'll probably get more distortion than you really should in triode mode. I any case, the most you're likely to get from triode mode is about 3 to 4 w and since EL84s are especially good in UL, that's the way I'd go if there's any choice.

Secondly, EL84s behave so well (in both triode and UL) using cathode bias that I doubt you'll do any better with fixed bias, though it won't do any harm to try if that's what you want.

Thirdly, unless you're stuck with 12AX7 for the voltage amp and splitter, you could do better IMHO. It's OK for guitar amps but I don't think much of it for Hi-Fi use. I'm also unsure if a 12AX7 concertina splitter would do very well driving EL84s in triode mode. A better tube, with 60% of the gain, would be a 12AT7.


As the amp was UL to begin with I expect the trannies are 8K, not 10k. I'll have a play and see which I prefer. Not chasing anything above 3w.

Concertina phase splitter has a gain of less than 1 anyhow, irrespective of the tube used if I read the theory correct. It's the input tube that gives the voltage gain to drive the outputs.

As I haven't purchsed the input/splitter tubes yet and my power trafo has dual heater wingings it's open season for input gain stage and splitter duties, 12a?7, 6sn7 or whatever I want really.

Same, I'm not fixed with any particular B+ aside from the fact that my secondaries will only give 120 and 240 AC (well, I guess that means that if was going to use a voltage doubler then anything from 600 odd volts downwards is achievable) point is, B+ certainly doesn't need to be at 320.

The output section of Sorenj's amp is pretty much precisely what I was planning on doing, solves all those issues with how cathode bypass caps and cathode resistors of different types "sound" by just biasing the grid instead. But, as people have said, cathode bypass is simple, has been widely used, plus it gives an element of failsafe as the tubes age.

With a room of 3.5m x 6m and a listening distance of about 2.5m with a pair of 95dB/w efficient single driver speakers, more than a couple of watts will be plenty sufficient to aggravate my wife, cats, 2 year old... Not chasing megawatts.
 

ray_moth

diyAudio Moderator Emeritus
2004-01-27 8:55 am
Jakarta
Not chasing megawatts, true, but it doesn't hurt to have some headroom so that transients don't get distorted. That's my main reson for suggesting you use UL, plus the fact that I doubt if you'll be able to hear any difference between triode PP and UL PP a low levels.
 
ray_moth said:
Not chasing megawatts, true, but it doesn't hurt to have some headroom so that transients don't get distorted. That's my main reson for suggesting you use UL, plus the fact that I doubt if you'll be able to hear any difference between triode PP and UL PP a low levels.


Umm, sorry to disagree Ray since we are usually on exactly the same page :D but to my ears at least triode connection does sound a bit better at all power levels with the 6BQ5 whether in SE or PP. (Dyna trannies in PP)

DrewP said:
Do fixed bias amps recover from overload faster than cathode biased ones (where the cathode caps charge up)?

Yes, there is no large time constant to discharge provided of course that you have not also severely clipped the driver stage. (Coupling cap) Avoiding the use of a large, and probably mediocre cathode bypass cap is a huge benefit.

Sorenj's design is a reasonable choice.


:D
 
MAybe off topic question

sorenj07 said:
here's an amp I built recently that uses EL84's in triode, with a negative voltage supply. I like the way it sounds :)

[IMGDEAD]http://img90.imageshack.us/img90/6626/el84triodeppsignalmarkifj6.th.jpg[/IMGDEAD]


Is there any reason a design like this couldn't be modified by separating the inputs to eliminate the phase splitter and use each half of the 6922 as a driver for a balanced input directly (already phase inverted) to drive the EL84 PP?
 
Thank you! I have been trying to find out if this was reasonable because, even though I am new to this, I have good balanced outputs from DAC and figure it would make sense to use them. How come there is almost zero information on this approach? I can't be the only one with the option of balanced inputs. Is there a reason I should know about?
 
Some people do used balanced signal paths with tubes, it's just not that common. You can feed a balanced signal to the LTP based phase splitter, gain will be 6dB higher than for single ended drive, the change is trivial, just use an xlr instead of the rca jack. Replicate what sorenj has done on the currently used 6922 on the unused input. XLR pins 2, 3 are signal, and 1 is ground IIRC.. Check for the correct input pin polarity and be sure to wire both channels the same.

If you need volume control use a shunt connected volume control across the + - signals with series resistors at the input end. Details you'll have to work out.

Details of balanced dac, pre-amp and other tidbits here:

http://www.triodeguy.com/vacuum_tube_audio.htm
 
Update

Okay, after some testing, the parts on hand are:

Power Tranny: 293v CT 293v @ 150mA, 6.3v @4A, 5.4v @ 2A

Filter choke: 1.5H @ 200mA

OPT Trannys: 21% UL taps (plan to run triode instead) 11k primary for a 4 ohm secondary connection

100k 24 position stepped attenuator

4 x EL84's

Looks like I'll either have to bleed off and resistor down a B- supply for the fixed bias, or else fit an additional little power tranny.

For some reason (wonky aesthetics aside) I'm attracted to the idea of 3 6sn7's for input and LTP phase splitting duties.

Notions from the audience?
 

Svein_B

Member
2005-10-24 7:34 pm
For the negative bias voltage, and negative for LTP if you need it, use the 5V winding and a voltage doubler. Should give you -12V.
Use schotkey diodes for minimum diode voltage drop.

With 293V CT transformer your B+ is going to be around 400V. A little high?

SveinB.