Low voltage on cathodes of EL84 output tubes

Ronno25

Tele-Meister
Joined
Dec 31, 2020
Posts
238
Location
Oregon
I'm hoping some of ya'll experts can help me with this. I just built an 18w Marshall clone. It's a two el84 cathode biased circuit. My power transformer puts out a bit lower voltage at 250-0-250 than what's called for and has resulted in lower voltages all around my amp, about 75% of what's called for on the schematic. This doesn't surprise me.

However, my output tube cathode voltages are at about 40% of what's called for. This strikes me as odd and I don't understand why this is happening. It's also made it difficult to bias the output tubes. It would also require me to use a 50r instead of a 130r (what's installed in the circuit now) to get anywhere near the 100% plate dissipation range.

Any idea what's going on with my output tube cathode voltages? Is it a problem and can it be fixed? Any issue with using a 50r or so cathode resistor to bias the tubes?

Using a solid state rectifier my voltages are:

Heaters 6.3

B+1 270
B+2 246
B+3 227

V1
P1 111
P2 .94
P6 112
P8 .8

V2
P1 178
P2 35
P3 47
P6 172
P7 33
P8 47

V3
P3 4.75
P7 260
P9 245

V4
P3 4.75
P7 260
P9 245

V5
p1 254
P3 272
P7 254

Expected voltages:
Screen Shot 2022-06-21 at 5.19.21 AM.png
18wattLite2b.gif
 

D'tar

Friend of Leo's
Joined
Jan 11, 2013
Posts
3,783
Location
WNY
Im not an el84 guy...... however... with already lower than desired voltages, any attempt to heat things up will further decrease your voltages and depending on the current (mA) capabilities of the PT the voltages may start dropping exponentially from here. How does it sound?
 

2L man

Tele-Afflicted
Joined
Nov 23, 2020
Posts
1,537
Age
62
Location
Finland
EL84 control grid voltage is quite low. It is designed to be driven smaller signal.

Bias current is only about 16mA each and idles about only 4W. You should make cathode voltage even lower changing resistor value smaller which increase cathode current.

When plate voltage is low tubes do not operate hard and sound might suffer. You can increase load using lower impedance speaker or put current soeaker to higher impedance output.
 
Last edited:

andrewRneumann

Tele-Afflicted
Joined
Mar 22, 2020
Posts
1,992
Location
Cincinnati, OH, USA
Any idea what's going on with my output tube cathode voltages? Is it a problem and can it be fixed? Any issue with using a 50r or so cathode resistor to bias the tubes?

It appears you just have an undersized PT. That's going to limit the amount of voltage and current you can use... and that means you may not be able to bias up to 100%. Even if you could bias to 100%, the input headroom of the EL84's would be so small that you'd be overdriving them near all the time. Biasing it cold might be the only realistic way to go. And if it sounds good--who cares? The power tubes will last forever and you'll be a happy camper.

I don't know what you did with the dropping resistors in the power supply, but at these voltages, you could use the 1K5 for the screen dropper (labelled 10) and probable be fine with no screen grid stoppers (6 and 7). That should raise up the current a little. YMMV.

And if we are really scrimping, get rid of the 220K bleeder resistor as it's just wasted current.
 

Ronno25

Tele-Meister
Joined
Dec 31, 2020
Posts
238
Location
Oregon
Im not an el84 guy...... however... with already lower than desired voltages, any attempt to heat things up will further decrease your voltages and depending on the current (mA) capabilities of the PT the voltages may start dropping exponentially from here. How does it sound?
It sounds good. I didn't realize there was a problem until I checked voltages. The cleans aren't that great but the distortion is fantastic. I tacked in a few extra cathode resistors to bring the bias up to ~90% or so. This way the sound was the opposite; amazing cleans but not so great break up.

Would it hurt anything to lower the cathode resistor to something like 50r to get 100% plate dissipation? I mean, aside from lowering voltages all around.

It appears you just have an undersized PT...
This is probably true. I converted a Hammond AO44 to make this. Very small PT. Initially I made it with a tube rectifier but the PT was getting too hot.
And if we are really scrimping, get rid of the 220K bleeder resistor as it's just wasted current.
Already done!


Here is a layout of what my amp actually is. Except I've changed to a solid state rectifier and added a 100ohm sag resistor.

Screen Shot 2022-06-21 at 10.44.50 AM.png IMG_5769.jpg
 
Last edited:

printer2

Poster Extraordinaire
Joined
May 24, 2010
Posts
8,488
Location
Canada
The 10 ohm will not do much, any sag will be coming from the PT already. If it sounds good button it up and call it a day. Or you could convert it to fixed bias...
 

Ronno25

Tele-Meister
Joined
Dec 31, 2020
Posts
238
Location
Oregon
The 10 ohm will not do much, any sag will be coming from the PT already. If it sounds good button it up and call it a day. Or you could convert it to fixed bias...
Whoops, it's supposed to say 100ohm. Fixed.
 

printer2

Poster Extraordinaire
Joined
May 24, 2010
Posts
8,488
Location
Canada
Whoops, it's supposed to say 100ohm. Fixed.
Measure the secondary resistance. If it is well over 100 ohms I would not bother with it, mind you roughly 35 mA x 100 ohms = 3.5V, might as well leave it in.
 

andrewRneumann

Tele-Afflicted
Joined
Mar 22, 2020
Posts
1,992
Location
Cincinnati, OH, USA
Initially I made it with a tube rectifier but the PT was getting too hot.

Now that is a surprising result. I would assume the valve rectifier would take some load off the PT, not cause more current and heat.

The 10 ohm will not do much, any sag will be coming from the PT already. If it sounds good button it up and call it a day. Or you could convert it to fixed bias...

Put another way... the sag you would get from a valve rectifier is provided by the high resistance of this tiny PT. Consider removing the sag resistor. Monitor the heat in the PT. With low voltages and high currents, you end up more in Class A for more of the cycle--so there might not be much global sag going on anyway. Hook up a voltmeter to B+ while at full drive and find out!

I don't see any trouble in experimenting with 50R on the cathodes. Monitor the heat build up in the PT though. Definitely pushing that thing to its max.
 
Last edited:

Ronno25

Tele-Meister
Joined
Dec 31, 2020
Posts
238
Location
Oregon
Measure the secondary resistance. If it is well over 100 ohms I would not bother with it, mind you roughly 35 mA x 100 ohms = 3.5V, might as well leave it in.
I'm getting ~250ohms resistance across the hv secondary. I have jumpered it and the 100ohm resistor is dropping about 15v. I do think I prefer the feel of the amp with the 100ohm in. Irrespective, the cathode voltages on the output tubes are very close with or without the sag resistor.
 

Ronno25

Tele-Meister
Joined
Dec 31, 2020
Posts
238
Location
Oregon
I'm still wondering if it is okay (not bad for the tubes/amp/PT) to drastically lower the cathode resistor to, say, 50r. I really don't understand how this works.
 

dan40

Friend of Leo's
Joined
Aug 19, 2015
Posts
2,784
Location
Richmond Va
I'm still wondering if it is okay (not bad for the tubes/amp/PT) to drastically lower the cathode resistor to, say, 50r. I really don't understand how this works.


Lowering the resistor value to 50 ohm will cause the power tubes to draw more current. This will be fine but keep in mind that running them at 100% dissipation will wear them down quicker than running them cooler. Most EL84 amps run their tubes hard so they often need replacing much sooner than larger tubes. Also remember that biasing for more dissipation will cause the tubes to draw more current from the PT. This will naturally cause the B+ voltage to drop a bit more as the PT has to supply extra current. Check the PT after several minutes of playing to be sure that it's not being over worked and running hotter than it should.
 

jjlemon

TDPRI Member
Joined
May 11, 2014
Posts
94
Location
Abertawe
I'm still wondering if it is okay (not bad for the tubes/amp/PT) to drastically lower the cathode resistor to, say, 50r. I really don't understand how this works.
I think that would be pushing your luck too far. If the power transformer is being stressed, and it looks as if it is, I would lower the cathode resistor in increments and measure voltages each time to see the effect it has on the supply and bias current.

Say, drop to 120, then 100, and all being well to 82.

As an alternative to desoldering each time, you could try tagging a resistor in parallel to gradually approach lower cathode values. I.E. 680r para 130 is approx 110, 470r ll 130r ~ 100r, then 330r for 93r, then 270r for 88r and so on. Once you know it's all within safe margins, substitute the value for a single resistor of approximate value, or leave it paralleled if a decent power resistor is used.

Are you sure the tubes are healthy?
 

2L man

Tele-Afflicted
Joined
Nov 23, 2020
Posts
1,537
Age
62
Location
Finland
If you are not satisfied to sound you can test making it more raw shorting 1k screen dropper resistor using a piece of wire. Screen voltage is quite a lot lower than plate voltage and no loss over 1k will rise screen voltage slightly.

Possibly higher screen voltage will increase bias voltage too which might again needs to be made hotter. When you install more parallel cathode resistors which are 1k neighbourhood they can be only 1/2W.

Also when plate voltage is low you can manipulate OT impedance to half if OT has higher impedance output or you have lower impedance speaker or you parallel two speakers. Then tubes work harder when they produce actual output power and cooler bias might be fine.

Now when plate voltage is low power tubes might run cooler when they make more output power?
 

Lowerleftcoast

Poster Extraordinaire
Joined
Dec 30, 2019
Posts
5,636
Location
california
I do think I prefer the feel of the amp with the 100ohm in. Irrespective, the cathode voltages on the output tubes are very close with or without the sag resistor.
There *should not* be much sag in a cathode biased amp. Please be careful and try blind tests to determine if you actually feel a difference.

I can appreciate a 15 volt difference in B+ may not change the cathode voltage much.
 




Top