Cathode biasing EL34’s

[dude]

Amp is a 6v6 Plexi, AO43 chassis, Hammond original PT, 630v and a 30 watt OT. I also installed a switch for fixed or cathode bias. I’m having an issue checking the dissipation on the EL34s in cathode bias mode. I’ve looked at the data sheet for EL34s and pin 1 is a grid3 but needs to be tied to the cathode, pin 8, which it is. When l try to calculate the values to determine what cathode R to use, (single R for both tubes), l get like 45 volts lower between cathode to plate on the EL34s, 377v (pin 8 to pin 3) than ground to plate, which is 420v. Using SS plug-in. For 6v6s, 6L6s those two voltages are usually the same.
So, do l use the 377v or 420v in figuring out what cathode R l need for about 24/25 watts dissipation ?
I have a 5w, 750 ohm on the screens rather than 1k? Right now l have a shared 250 R, and l’m coming up with 20 watts dissipation, l’m figuring this on the cathode to plate voltage, 377.
Seems l need to lower the cathode R to say 200 ohms to get to 25 watts. Am l going about this correctly?

[tubeswell]

When l try to calculate the values to determine what cathode R to use, (single R for both tubes), l get like 45 volts lower between cathode to plate on the EL34s, 377v (pin 8 to pin 3) than ground to plate, which is 420v.

So is there 45 volts across the cathode resistor? And 420 volts on the plate?
What (actual) resistance do you measure on the cathode resistor?

[sluckey]

According to YOUR voltages I calculate each tube is idling at 32.4W (ignoring screen current).

I would just copy a well-known proven circuit. Matchless Chieftain uses a 270Ω for each tube. If you want to use a shared cathode resistor make it half that value, 135Ω.

     https://el34world.com/charts/Schematics/files/Matchless/Matchless_chieftan.pdf

If you do the math using the voltages on page 7 of the above pdf you'll see each EL34 is idling at 34.75W (ignoring screen current). If you want to run your amp cooler just use a larger cathode resistor(s).

[pdf64]

… When l try to calculate the values to determine what cathode R to use, (single R for both tubes), l get like 45 volts lower between cathode to plate on the EL34s, 377v (pin 8 to pin 3) than ground to plate, which is 420v. … For 6v6s, 6L6s those two voltages are usually the same.
…

That seems impossible?
In cathode bias, the cathode voltage will always eat into the HT voltage available for the anode.

[66Strat]

Amp is a 6v6 Plexi, AO43 chassis, Hammond original PT, 630v and a 30 watt OT. I also installed a switch for fixed or cathode bias. I’m having an issue checking the dissipation on the EL34s in cathode bias mode. I’ve looked at the data sheet for EL34s and pin 1 is a grid3 but needs to be tied to the cathode, pin 8, which it is. When l try to calculate the values to determine what cathode R to use, (single R for both tubes), l get like 45 volts lower between cathode to plate on the EL34s, 377v (pin 8 to pin 3) than ground to plate, which is 420v. Using SS plug-in. For 6v6s, 6L6s those two voltages are usually the same.
So, do l use the 377v or 420v in figuring out what cathode R l need for about 24/25 watts dissipation ?
I have a 5w, 750 ohm on the screens rather than 1k? Right now l have a shared 250 R, and l’m coming up with 20 watts dissipation, l’m figuring this on the cathode to plate voltage, 377.
Seems l need to lower the cathode R to say 200 ohms to get to 25 watts. Am l going about this correctly?

There is no requirement that I'm aware of that requires the suppressor grid to be tied to cathode. In fixed bias amps it is done so as a convenient means of connecting both the suppressor grid and the cathode to ground. In a cathode biased amp doing so may impact the cathode voltage measurements that you are using to bias the amp. The Phillips operating points indicate the suppressor grid G3 voltage of 0 volts. (see page 6 of the attached)

 

[dude]

I'm a little confused, forgive my ignorance.


With Fixed bias and EL34s, tie pin 1 and 8 (grid 3 and cathode) is this correct, YES.  (correct me if I'm wrong).
With Cathode bias and EL 34s, tie pin 1 and 8..?  yes or no? If no, then difficult wiring to have a switch for either cathode or fixed.


I just used my best meter, voltages on EL 34s are the following with SS plug-in: B+ = 411v
Screens, pin 4 to pin 8 and 1 (tied together) = 364v  (voltage with 750 ohm screen R) (higher if I go to ground below)
Screens, pin 4 to Ground = 394v                              (voltage with 750 ohm screen R)


Plate, pin 3 to pin 8 and 1 (tied together) =375v
Plate, pin 3 to ground = 394v


Using a 244.2ohm cathode R, 10 watts
Voltage across the 244.2 cathode R, hot = 28.7
 
My math:  28.7/244.2=.11753       Single tube .11753/2=.0588
Plate voltage (using pin 3 to pin 8 and pin 1,(g3) tied together) = 375v - 28.7v=346v
346v x .0588=20.35 dissipation (watts) per tube if matched


If I use 394 plate voltage (pin 3 to Ground), instead of 375v 
28.7/244.2=.11753,   .11753/2=.0588    394v - 28.7v= 365.3v
365.3v x .0588 = 21.47 dissipation (watts)


Obviously, the g3, pin 1 on EL34 tied to cathode affects the tubes plate voltage..?
So where am I wrong here, and why the difference with pin 1 tied to pin 8.


Sluckey, yes I can copy the Matchless but that amp is cathode biased only, no switch for fixed, are the pins 1 and 8 tied on the Matchless..?  Not looking for a what works but is my math off? Do EL34s require separate cathode resistors in my set-up..?  Or, is what I'm trying to do can't be done with the switch for cathode and fixed bias..?


 


 






 

[66Strat]

What happens to cathode voltage with pin 1 connected to ground vs pin 1 connected to cathode?

[dude]

What happens to cathode voltage with pin 1 connected to ground vs pin 1 connected to cathode?

Both pin 1 and 8 are tied together AND go to ground through a 1 ohm R


UPDATE: I've been checking the cathode voltage to ground, across the R.  I have a switch to select either fixed or cathode biased. Cathode voltage to ground is 403V cathode to pin 1 and 8 is 375v 

[dude]

Sluckey are all those voltage taken on page 7 from ground? I assume the cathode tied to g3 of an EL34 is not used as the ground when checking voltages on that page. But rather using the star ground. Correct.


Also, using those voltages, each EL34 is 34.72 watts dissipation, correct.
Isn't an EL34 rated at 25 watts..?  Wouldn't 34.7 watts fry the pants off those EL34's or I guess that's a Matchless tone, burn'm hot..?


Just trying to figure if I'm figuring this correct.

[sluckey]

With Fixed bias and EL34s, tie pin 1 and 8 (grid 3 and cathode) is this correct, YES.  (correct me if I'm wrong).
With Cathode bias and EL 34s, tie pin 1 and 8..?  yes or no? If no, then difficult wiring to have a switch for either cathode or fixed.

Just connect pin 1 to pin 8 in either case. Matchless did. ***OR*** connect pin 1 to ground in either case. Either way works just fine and does not complicate cathode/fixed bias switching.

Screens, pin 4 to pin 8 and 1 (tied together) = 364v  (voltage with 750 ohm screen R) (higher if I go to ground below)
Screens, pin 4 to Ground = 394v                              (voltage with 750 ohm screen R)

ALWAYS MEASURE SCREEN VOLTAGE IN REFERENCE TO GROUND! NEVER SCREEN TO CATHODE. You don't even need screen voltage.

I highly recommend using a separate cathode resistor for each tube. Then instead of ***ASSUMING*** the tubes are perfectly matched with the same current is flowing through each tube, you'll be able to measure each tube individually.

You calculate the plate to cathode voltage by subtracting the measured cathode voltage from the measured plate voltage.
You calculate the cathode current by dividing the cathode voltage by the cathode resistance.
You calculate the screen current by dividing the screen resistor voltage drop (measure across the resistor) by the screen resistor value.
You calculate the plate current by subtracting the screen current from the cathode current.
You calculate the plate idle dissipation by multiplying the plate to cathode voltage times the plate current.

Repeat for the other tube.
Done.

Sluckey are all those voltage taken on page 7 from ground?

Yes. In fact, unless otherwise stated all voltage readings on any schematic are referenced to ground.

[dude]

ok, I've been figuring the dissipation correctly, at least I think..


Plate voltage to ground is 403, voltage across the cathode R is 29, and the cathode R is 244 ohms. All I need.
403 minus 29 = 374
29/244=.11885
374 x .11885 = 44.44 watts (two tubes divided by 2 = 22.22 watts dissipation


Is this example correct?


Why is my screen voltage at 394v so high as compared to the Matchless (I have 403v Matchless 415v on plates), I have 750 ohm screen R, Matchless has 1K and another 50W 1K, why such a big wattage? Is the filter choke dropping way more then my choke..? Only 1.3K difference from my 750 ohm screen. Are my screens too high for EL34s in cathode bias?
The difference in volts between screens and plates in my amp are about 10 volts, why is the Matchless 46volts less on the screens the plate?
I'm about a 1 compared to your knowledge of 10.  If I check the screen voltage from pin 4 to 1 and 8, 8 goes to ground, I get 375v much lower than if I check screen (pin 4) to star ground, 403v. You said all voltages are from ground unless specified, so I assume in the Matchless they got the 269v screen voltage from pin 4 to ground? 


Also, I used your schematic to wire up the dual bias switch for fixed or cathode.
Looking at the switch, I can see I can't use separate cathode R for biasing, do you have a schematic that shows how to wire separate cathode biasing R's, and fixed?

[sluckey]

Plate voltage to ground is 403, voltage across the cathode R is 29, and the cathode R is 244 ohms. All I need.
403 minus 29 = 374
29/244=.11885
374 x .11885 = 44.44 watts (two tubes divided by 2 = 22.22 watts dissipation


Is this example correct?

yes

Why is my screen voltage at 394v so high as compared to the Matchless

The Matchless is running much hotter than your amp, 89mA cathode current v. your 59mA cathode current. If you consider that up to 10% of the cathode current flows through the screen and it's resistor, you'll see that the Matchless screen resistor is passing more current than your amp. That means the Matchless screen resistor drops more voltage than your amp. More voltage drop means there will be less voltage at the screen.

Also, I used your schematic to wire up the dual bias switch for fixed or cathode.
Looking at the switch, I can see I can't use separate cathode R for biasing, do you have a schematic that shows how to wire separate cathode biasing R's, and fixed?

Don't have one at the moment. Let me think about it for a while.

[PRR]

....up to 10% of the cathode current flows through the screen and its resistor....

In EL34, can be as much as 17%.

[dude]

Sluckey thanks again for your help and clarification, much appreciated.
I ended up putting in a 215 ohm 10 watt cathode R, l got 28.07v across it and 402v on the plates. Screens at pin 4, 389v with a 750 ohm R. Tubes 65mA’s, 24.3 watts dissipation, minus the screens, so probably 23.3 watts area.  EL34s sound good but the cathode R gets hot. I’d like to cool things down but the switching l used for fixed to cathode just lifts the ground for bias, used a DPST, 4 lugs. I’d need a different switch and the wiring looks a little tricky to get two separate cathode R’s in cathode mode without interfering with the fixed mode..?
So, l was thinking either get get a 215 ohm 25 watt R or use two 10 watt, 430 ohms in parallel for 215… Should l brother, the 10 watt alone isn’t discolored at the moment, but hot to touch when l turn amp off after a minute.
I keep thinking about the 1k/50 watt R from the choke in the Matchless to the screens, you posted but that amp had 269v on the screens, l have  390v. I’m not going to blow that 14H hammond choke, am l?

[sluckey]

So, l was thinking either get get a 215 ohm 25 watt R or use two 10 watt, 430 ohms in parallel for 215… Should l brother, the 10 watt alone isn’t discolored at the moment, but hot to touch when l turn amp off after a minute.

Power resistors are supposed to get hot. You have all the info you need to calculate the power being dissipated by that resistor. Do the math and then tell me if a 10W resistor is good enough. 

I keep thinking about the 1k/50 watt R from the choke in the Matchless to the screens, you posted but that amp had 269v on the screens, l have  390v. I’m not going to blow that 14H hammond choke, am l?

You've danced all around this issue. It's high time for you to measure the voltage drop across the screen resistor and calculate the current flowing through the resistor. Repeat for the other screen resistor. Add the two currents then you'll know how much current is flowing through that tiny choke. Knowing the screen current will allow you to accurately calculate plate dissipation too.

All the EL34 amps that I'm familiar with use a pretty hefty choke, if they use one. That little Hammond choke was not used in the power supply. It was the plate load for the reverb driver.

I'm surprised that you want to use EL34s with a PT and OT that was designed to work with EL84s. I'd keep a backup amp handy.   

[dude]

I=V/R   Current divided by resistance
            389v/750=.5187 mA screen current..?


            Times 2=1.02 Amps    Will that tiny choke handle 1 Amp? or is it 1 mA..?


V=IR     27.07 x .750= 6.02 volts (is this watts?)   


Never had a handle on current, voltage, wattage don't know the difference. I know about 120v, 240v and amperage but that's all electrician knowledge, I have no education an EE tech.


That Hammond AO-43 PT, seems to handle the EL34's fine, played it for a few hours and it's just a little warm, that tiny choke I used  might be too small, if I had the knowledge to figure out what current rating I needed, I wouldn't be asking. I'm not using that Hammond 8K OT, I have a 30Watt OT with two taps, 16 and 8 so I can get the correct impedance for EL34s. Just never got a grasp on OHMS law.
BTW, this addendum is after a few shots of Tequila.       
 

[tubeswell]

Never had a handle on current, voltage, wattage don't know the difference. I know about 120v, 240v and amperage but that's all electrician knowledge, I have no education an EE tech.   

Watts is power (in electrical terms: Voltage x Current)
A unit of charge is a Coulomb.
Current is amount of charge per unit of time (e.g. coulombs per second)
Voltage is energy per unit of charge (joules per coulomb). Voltage = Current x Resistance, which can be re-written to solve for either Current or Resistance, hence Voltage/Resistance = Current, or Voltage/Current = Resistance.

[sluckey]

READ MY LIPS!

measure the voltage drop across the screen resistor

That means put one meter probe on one side of the resistor and put the other probe on the other side of the resistor. IOW, across the resistor.

[dude]

Thanks Tubeswell.

measure the voltage drop across the screen resistor

Voltage across each screen R, 6.33v and 6.15v.

[sluckey]

Voltage across each screen R, 6.33v and 6.15v.

So 8.4mA and 8.2mA screen current. Plus maybe 5mA for the little tubes. Total of about 22mA. That Hammond choke is probably OK.

[dude]

Total ignorance on my part, all this stuff is on Akien’s page on ohm’s law. Using P=I(squared)*R that 10k, 215ohm cathode R is fine, about 4.5watts at idle. Just need to get my decibels correct. It’s starting to sink in, it’s a wonder l got this far. Thanks