more headroom possible?

[Helmholtz]

But measuring across the 1R cathode resistors as i always do will not show negative voltage.

Cathode voltage cannot be negative.

[daz]

Cathode voltage cannot be negative.

ok, as u can see i am clueless. Bias voltage is a neg v but not at the cathode. I coulda sworn the meter was in - range when i'd set bias but i guess i just assumes since bias is a - voltage. But at the grids, not cathode. I should have thought of that. Duh...oh well, i never said i know jack about this stuff. So now i have to figure out where that set resistor has to be. Maybe back to 10k where it was b4 the 6v6 change.

[Helmholtz]

Plug your black meter lead into the black meter socket and connect other end to amp ground.
Plug your red meter lead into the red meter socket and connect other end to grids.
You should see a negative bias voltage.

[daz]

That i realize, but i doesn't matter now that i realize i was looking at + voltage all along. It was just in my mind that bias is negative, but at the grid. What is still odd is now i put the el84s back in and can't get ANY voltage. Tried everything from a 4.7k to a 100k for the set resistor and it just sits at zero no matter where i set the bias pot. Maybe another bran fart. I got a million of em....

[Helmholtz]

Tubes will suffer when operated wihout sufficient bias, Pull them until you measure at least - 10V at the socket grid pins.

[Tom Phillips]

ok, as u can see i am clueless....

We all started out clueless. We gain knowledge by study and experience. At this point, I think it would be good for you to do a little study. You will then enjoy your circuit tweaking much more because you would better understand how the circuit functions and know what to do to improve/fix your design. Do you have any reference material. If so, please tell us what you have available.

...Bias voltage is a neg v but not at the cathode. I coulda sworn the meter was in - range when i'd set bias but i guess i just assumes since bias is a - voltage. But at the grids, not cathode. I should have thought of that. Duh...oh well, i never said i know jack about this stuff. So now i have to figure out where that set resistor has to be. Maybe back to 10k where it was b4 the 6v6 change.

It's confusing at first However, once you understand what's going on it's simple.

Bias is similar to the throttle of an engine. The bias throttles the tube and prevents it from going to red line. The bias sets the current flowing in the tube. The voltage you measure at the cathode in a fixed bias circuit is only there if you have installed the resistor from cathode to ground. That voltage measurement allows you to calculate the current flowing through the tube but we aren't done yet. You also need to measure the plate voltage under the same conditions. Once you have the plate voltage and the current values you can then calculate the no signal power dissipation in the tube. Knowing that measured power dissipation figure and the maximum power dissipation specification of the tube (from the tube data sheet) you can calculate the % of max power dissipation. It has become internet lore that 70% is the magic number. However, I find that 50% is a good starting point and blind sound tests with users rarely result in them picking vales as high as 70%. Reminder - this description applies to fixed bias circuits only. For cathode bias you would use the plate to cathode voltage vice the plate to ground voltage before you calculate the power dissipation.

Disclaimer: The numbers are approximate but very very close to reality because we are not adjusting for screen current of the small voltage across the one Ohm cathode resistor.

[daz]

O, i figured it out. One of the wires came off again in one of the converters i made. I took old octal tube, broke the glass and cut all the tube parts out leaving just the base then soldered wires from 9 pin sockets to the shortswires where i cut the internals out of the octals. So those pins in the octals don't flow solder well and they keep popping off. Just want to use them long enough to decide whether to replace the octal sockets in the amp and go permanently with el84s. I really like the sound with fixed bias more but theres one problem. At around 10-11v bias the plates are 410. I can't afford to replace the tubes 3 or 4 times a year so it seems i will have to use cathode bias. On the other hand, i read somewhere that somebody cited an amp, maybe a traynor IIRC, but they said it runs them at more then 400v but uses huge screen R's and they last and are reliable like that. Is that possible and if so how high a value would make the tubes reliable at 410v on the plates?

EDIT...checked out a bias calculator and i see 11v is very low so i entered the plate v and it gave me 17v at 60% so tried that and now i'm VERY sure i like these more then 6v6 ! But again, will large screen values keep the tubes safe? I'd really hate to lose this tone. Oh,and right now i have 2k on each screen. Voltage is almost the same as the plates. (which are now down to 400v with 20ma bias) Might be 5v lower.

[Chuck H]

You will always read somewhere between zero and a positive voltage at the cathode across the 1R resistor. With fixed bias the negative voltage will only be on the grid pin of the tube. The tubes "bias" is the voltage relationship between the cathode and the grid. The grid must be negative with respect to the cathode. So... With cathode bias the cathode has a standing voltage and the grid is referenced to zero volts. Since zero is less than the voltage on the cathode the gris has a "negative relation" to the cathode.

In fixed bias the cathode is at zero volts and the grid is made negative.

Both methods make the grid negative with respect to the cathode.

Get it?

So... You will only be able to measure negative voltage at the grids in a fixed bias circuit. Not the cathode. And...

The positive voltage on the cathode in a cathode biased circuit (with the grid referenced to zero volts) would be the amount that the amount that the grid is negative with respect to the cathode. Even though it is a positive voltage.

[daz]

You will always read somewhere between zero and a positive voltage at the cathode across the 1R resistor. With fixed bias the negative voltage will only be on the grid pin of the tube. The tubes "bias" is the voltage relationship between the cathode and the grid. The grid must be negative with respect to the cathode. So... With cathode bias the cathode has a standing voltage and the grid is referenced to zero volts. Since zero is less than the voltage on the cathode the gris has a "negative relation" to the cathode.

In fixed bias the cathode is at zero volts and the grid is made negative.

Both methods make the grid negative with respect to the cathode.

Get it?

So... You will only be able to measure negative voltage at the grids in a fixed bias circuit. Not the cathode. And...

The positive voltage on the cathode in a cathode biased circuit (with the grid referenced to zero volts) would be the amount that the amount that the grid is negative with respect to the cathode. Even though it is a positive voltage.

yeah, i get it. I basically was having a brain fart, but thats all over and done with at this point and i've come to begin questioning my decision to go with el84s. I now have tried it both fixed, CB, and with different screens and tried different bias points. Once i had a chance to play it for a long while after getting past all the technical stuff, i have come to feel the 84's are not very dynamic. At least not in any config i have tried. Higher screens dull are needed with high plates from what i have read ad 2k just dulls the tone too much. In most configurations i liked the tone in the lows a lot more then the 6v6 but it felt almost solid state compared to the 6v6. So i went back to 5v5 and lowered the screens to 500R and will see how that goes. It does give it better definition but i haven't spent enough time to know if it's much better. The 6v6 feel much better but i prefer the sound of the el84s. If i could get some dynamics from the 84s i'd likely stick with them.

[Chuck H]

Higher screens dull are needed with high plates from what i have read ad 2k just dulls the tone too much.

For my own tastes this has been my experience also. My own amp uses a 2.2k dropping resistor between the plate and screen nodes in the HV rail and then 470R individual screen grid resistors. I tried 1k screen grid resistors because it's the more acceptable (conservative) value in this case but I didn't like the change in tone and feel so I went back to 470R. Damn the torpedoes.

The 6v6 feel much better but i prefer the sound of the el84s.

I get it. It may be that being a higher mu tube the el84 is giving you more power tube clipping. Which is a enough different from preamp clipping to hear a difference. In truth clipping is clipping. But there are some things that happen when the power tubes clip that change the overall tone. Not the least of it is how the frequency balance gets leveled out. There is a frequency compression as well as a dynamic compression.

If i could get some dynamics from the 84s i'd likely stick with them.

This was the trick for me designing with el84's. At higher voltages they do cooperate better making dynamics. Fixed bias is important for this. I mentioned earlier that I'm using a coarse, hybrid bias circuit. I think you could do the same with your higher Vp. It's best to use a scope to dial it in but you could probably do well enough without as long as you don't mind soldering and ear testing. AND...

A more efficient speaker is a must! With the el84's at fixed bias when clipping and a good efficient speaker my own amp (at 365Vp with 470R screen grids and burning through tubes in half the time of other amps) won't do a solid clean tone for even club gigs. But it will get a pseudo clean tone, clean and loud enough to be useful and the overdrive is just great. It should be noted that my own amp does not have a master volume of any kind. In other words, loud is a tall order for a pair of el84's in any circumstance. And what's been "loud enough" for me might not be for others.

[Chuck H]

On another note...

I'm using a PT that is rated for more current than yours and that could surely have an affect on dynamic response.

EDIT: I just looked up some specifications and my PT is rated 144mA while yours is 120mA. I do think your PT is up to the job considering the circuit in question. And mine only ever gets a little warm to the touch when the amp is cranked full blast for over an hour.

[daz]

My experience has been a bit different. I actually like the 84's for clean, in fact that was the thing i liked most. BUT, not as in turning the gain down and master up for a clean amp sound. I get my cleans via the guitar volume and in that way the 84s were making me very happy. But the overdrive was just too un dynamic. And i also experienced a difference from your experience in that i felt the 84's stayed cleaner louder. That was a bit surprising and maybe it's also a clue as to why it's not feeling very dynamic.

[daz]

I got to thinking and it occurred to e that the stiff lack of touch with the 84's might be countered with bigger dropping resistors in the preamp. Because they WERE quite large till i changed them out a few weeks ago because the 6V6's were a bit mushy. Now with the 84s i thought maybe they are too small and creating too stiff a feel. Worked pretty good. Gonna experiment more tomorrow and maybe go even lower to where they first were when i started messing with this thing again. It definately helped, how much remains to be seen once i can crank it up a bit more and play it long enough.

[Mick Bailey]

I've also noticed that EL84s are prone to compression with 2k2 screen resistors, depending on the amp in question. I consider there's more of a change in dynamics rather than tone, but that's just my ears. The main issue I had as a repairman was comebacks if a tube failed, especially during a gig. The usual failure mode I noted with some amps was a shorted tube and a screen resistor burnt up, almost exclusively with amps running a high plate voltage and insufficient screen protection. Even with stock 2k2 3w screen resistors the Dr Z Maz 18s I looked after would get through tubes at an alarming rate until the B+ was reduced.

The EL84 is fairly easy to drive into distortion, so there's an additional impact on dynamics at higher volume. One of the best amps I've had was a Traynor YGM3 and despite the short tube life on modern production tubes, was quite impressive both clean and driven hard. I did try it very briefly with a pair of Telefunken cylindrical plate EL84s and the volume increase over a pair of JJs was remarkable. I couldn't justify leaving those in the amp just to get eaten up in a few months. The plate voltage in that amp was well over 400v and it always sounded best just as Traynor built it. Sometimes the sacrifice of a pair of tubes is worth it for the sound. Back when I had that though the price of a pair of regular tubes was cheap enough not to worry. Oddly with that amp a tube failure never resulted in any other damage - it would crackle to give an advance warning that a tube was about to fail.

[daz]

I've also noticed that EL84s are prone to compression with 2k2 screen resistors, depending on the amp in question. I consider there's more of a change in dynamics rather than tone, but that's just my ears. The main issue I had as a repairman was comebacks if a tube failed, especially during a gig. The usual failure mode I noted with some amps was a shorted tube and a screen resistor burnt up, almost exclusively with amps running a high plate voltage and insufficient screen protection. Even with stock 2k2 3w screen resistors the Dr Z Maz 18s I looked after would get through tubes at an alarming rate until the B+ was reduced.

The EL84 is fairly easy to drive into distortion, so there's an additional impact on dynamics at higher volume. One of the best amps I've had was a Traynor YGM3 and despite the short tube life on modern production tubes, was quite impressive both clean and driven hard. I did try it very briefly with a pair of Telefunken cylindrical plate EL84s and the volume increase over a pair of JJs was remarkable. I couldn't justify leaving those in the amp just to get eaten up in a few months. The plate voltage in that amp was well over 400v and it always sounded best just as Traynor built it. Sometimes the sacrifice of a pair of tubes is worth it for the sound. Back when I had that though the price of a pair of regular tubes was cheap enough not to worry. Oddly with that amp a tube failure never resulted in any other damage - it would crackle to give an advance warning that a tube was about to fail.

Thanks Mick, that probably is enough to push me to pass on the el84s. Maybe i'll try them with 2k screens one more time after lowering the preamp screens which really makes the amp feel much better. That plus the 2k screen combined with that will likely be overkill but it will give it more squish in the attack (different than PA squish caused by the 2k's) and maybe i can counter the PA dullness with my adjustable NFB and treble/trebleshift/presence. But from your post thats just a unlikely last resort because 2k on the screens will probably be impossible to get past. When i lowered the preamp voltage as i said above i forgot i left the screens at 500R from (tried that with 6V6) so the 84s sounded really good like that but no way i'm running it like that. I can't afford a set of tubes every couple months. So chances are the 6V6's will stay but that pretty much will relegate the amp to garage queen or home use only. Maybe in time if i come to feel it's worth a try i'll get a high SPL speaker and see how it then cuts in in band.