Triode connected Pentodes?

[leadfootdriver]

I have a Crate V3112 with 4-el84's. I was thinking about doing this for that amp. I've read that the plate volts can be bad for the screens, but I've measured only about 4 volts difference between them on this amp, so I don't see what the problem is.

I would connect them with the 100 ohm screen resistor in the middle.

What do you think?

 

[jim p]

screen grid should be fine on an EL84

I don’t think you need to worry about the voltage on the screen grid if you check the Mullard data sheet for the EL84 the screen grid and the plate have the same voltage ratings. You can find the data sheet at tubedata.org. What I do see on the second page of the data sheet is that under the same setup relative to plate voltage they have only half the plate current in triode connection. Also they have doubled the value of the cathode resistor in triode operation. I have set up EL84 tubes in single ended amps but have just made sure plate dissipation was below 12 watts in both pentode and triode mode on the data sheet they almost half the plate dissipation in the triode mode? What ever you decide you should check the power dissipation of the tubes in both modes to see that they are less the 12 watts.


Looking over the load lines for the tube in the triode mode it looks like 24mA would put at the center of linear operation of the tube for class A operation.

 

[leadfootdriver]

What a great site! Thanks for that.

I'll do a test, and check dissapation.

Somone told me that you could just take the screens out of the circuit. Any objections to this?

 

[jim p]

I Think the screen grid will be fine

I don’t think you should float the screens the spacing of the elements of the tube are based on having the screen grid and the gain of the pentode is dependent on the screen grid voltage. The majority of the current should be plate current the screen is a screen and in the shadow of the control grid so it’s current should not be too great. The big threat to the screen grid is when it is more positive then the plate and that can’t happen when triode connected. Another problem would be if its voltage is too high but by the data sheet that should be no problem.

I have a V33 head that I am going to get back too in the future is the output section in this amp the same? Also attached a load line of the EL84 triode connected with 300 volts on the plate and a 5k load looks like 24mA would put it in the center for best linear operating range.

 

[leadfootdriver]

It looks like you really know your tube theory! I'm very new to all of this. But everything I could find online pointed to a triode connected pentode. I was told you could float the screen by a guy that is trying to push his own brand of amps, so I gave it a minutes thought. It would be a lot easier to wire that's for sure.

I see in your schematic that there's 470 ohm screen resistors. In my V3112, there's a 470 that feeds 4-100ohm screen resistors, giving each tube a 570 ohm. http://schematicheaven.com/newamps/crate_v32_palomino.pdf

However, I've read that a common value is a 2.7k, but I really don't understand the effect on tone, or what determines the final value. But, I do know that they're there to control screen current .

I was thinking about jumping the 470, and replacing the 4-100ohms with a higher value like a 1k, or a 2.7k, and rebias to test for tones. This way I can have a safe value while switching between triode, and pentode. I'll use a 3PDT toggle so I can have switchable cathode resistors for different bias requirements.

 

[jim p]

ideas on triode/pentode mode

I don’t think you will need to change the cathode resistor between the two modes but you may need to increase its value. I would check the drop on the cathode resistor in the pentode mode then wire up as the triode mode. From the curves for the triode mode and the values to use in the tables in both the Mullard and Tung-Sol data sheets if you have 6.5 to 8 volts on the cathode resistor in the triode mode I think you will be good. Your cathode resistor R22 is 60 ohms while on my amp it is 82 ohms so there should be room to increase it in value as far as the pentode mode. A big concern will be the voltage across the switch used for the triode to pentode mode. In the pentode mode it is no big deal you have the same voltage on all the screen resistors. While in the triode mode there will be big swings on the voltage because the screen is tied to the plate so you need to worry about arcing in the switch and crosstalk between the two different sections. I would plan on using a switch with a large body and also 4 poles use the outside poles for the screen connections and ground the center sections. Then route wires well away from each other. You can pair up the screen grid resistor connections on each side so R59 and R24 being one pair while R25 and R58 are the other pair

As far as screen grid resistor values a larger value will increase the safety margin as far as screen grid current going to high, more voltage will be dropped across the resistor decreasing the tube gain more and the resistor will help to limit the current. The 470 ohm resistor R9 is also lowering the plate voltage for the triodes in the amp something to consider if you change it.

The way I would do this is to get a good feel for how the amp is now in the pentode mode. Then just lift the screen grid resistors and wire them to there plates and check the amp out in the triode mode to see if it is worth adding the switch. With the pentode mode your wattage will drop by approx 60% the amp will have a lot of inherent feedback by now being a triode output so a bit more like a solid state amp. The bass response should be improved by almost a factor of 6. Also if doing this is good and works for you when you add the switch it will give you an idea if something is wrong with the switch wiring and routing.

 

[leadfootdriver]

I don’t think you will need to change the cathode resistor between the two modes but you may need to increase its value. I would check the drop on the cathode resistor in the pentode mode then wire up as the triode mode. From the curves for the triode mode and the values to use in the tables in both the Mullard and Tung-Sol data sheets if you have 6.5 to 8 volts on the cathode resistor in the triode mode I think you will be good. Your cathode resistor R22 is 60 ohms while on my amp it is 82 ohms so there should be room to increase it in value as far as the pentode mode. A big concern will be the voltage across the switch used for the triode to pentode mode. In the pentode mode it is no big deal you have the same voltage on all the screen resistors. While in the triode mode there will be big swings on the voltage because the screen is tied to the plate so you need to worry about arcing in the switch and crosstalk between the two different sections. I would plan on using a switch with a large body and also 4 poles use the outside poles for the screen connections and ground the center sections. Then route wires well away from each other. You can pair up the screen grid resistor connections on each side so R59 and R24 being one pair while R25 and R58 are the other pair

As far as screen grid resistor values a larger value will increase the safety margin as far as screen grid current going to high, more voltage will be dropped across the resistor decreasing the tube gain more and the resistor will help to limit the current. The 470 ohm resistor R9 is also lowering the plate voltage for the triodes in the amp something to consider if you change it.

The way I would do this is to get a good feel for how the amp in now in the pentode mode. Then just lift the screen grid resistors and wire them to there plates and check the amp out in the triode mode to see if it is worth adding the switch. With the pentode mode your wattage will drop by approx 60% the amp will have a lot of inherent feedback by now being a triode output so a bit more like a solid state amp. The bass response should be improved by almost a factor of 20. Also if doing this is good and works for you when you add the switch it will give you an idea if something is wrong with the switch wiring and routing.

Well I went ahead and connected the plate and screen. There's a very slight but audible buzz/hum for about a second when I flip the standby, and then everything stablizes and away goes the hum. The sound I get is nice, but I didn't crank it. So I don't know how it will respond to the higher output demand

Before I made this post, I already installed an 82ohm cathode resistor. :beer: I calculated 9.982 watts dissapation in pentode per tube, calculating my formula with the 100ohm screen resistor value and a 1 volt drop. I did another calculation with the combined 570ohm resistance, and an 8 volt drop from B+ to the screen side, and came up with 9.51 watts. I'm not clear on which formula to use, but the outcome is not that much different.

However, I checked my cathode volts in triode, and it stayed the same at 11 volts.

 

[jim p]

Looking good

I guess you don’t have a scope and waveform generator to check things out? I think I was looking at the table values wrong if you go to page D2 on the Mullard data sheet they have a cathode resistor value of 570 ohms per tube so in this case that would be divided by four for approx 140 ohms for your cathode resistor. Also it looks like the better curves to determine the operating point are C6 on the Mullard data sheet but they are with 250 volts on the plate ( tube gain will be 20% higher with 50 volts more on the plate). Even though the plate voltage is off by 50 volts you can see with changing control grid voltage (Vg1) you should be in a linear operating point on the tubes. So where you are now you are further into the linear operating point of the tube you could increase the cathode resistor which will give you a bit more power out of the amp, but if you want power you can just go back to the pentode mode for a factor of 3 to 4 times more. Increasing the cathode resistor will just lower the quiescent current (zero signal) in the amp which is canceled out in the output transformer. Reducing it will give you greater peak to peak current with maximum signal but may also introduce crossover distortion. If you don’t care about maximum power and the amp sounds good to you, your tube dissipation is fine so you could just go with what you have.
If you were going to change cathode resistor value between modes I would parallel the higher value resistor with another resistor using a mosfet switch for the lower value.

On your power dissipation I am getting plenty low numbers figure 11 volts divided by 82 ohms for 132 mA that divided by four for 33.5mA minus screen grid current of approx 4mA for 29.5mA times 300volts(plate voltage – cathode voltage?) for 8.86 watts. Then the screen grid is 4mA times 300 volts for 1.2 watts
Triode mode you have 33.5mA times 300 volts for 10 watts so that is OK

 

[leadfootdriver]

I guess you don’t have a scope and waveform generator to check things out? I think I was looking at the table values wrong if you go to page D2 on the Mullard data sheet they have a cathode resistor value of 570 ohms per tube so in this case that would be divided by four for approx 140 ohms for your cathode resistor. Also it looks like the better curves to determine the operating point are C6 on the Mullard data sheet but they are with 250 volts on the plate ( tube gain will be 20% higher with 50 volts more on the plate). Even though the plate voltage is off by 50 volts you can see with changing control grid voltage (Vg1) you should be in a linear operating point on the tubes. So where you are now you are further into the linear operating point of the tube you could increase the cathode resistor which will give you a bit more power out of the amp, but if you want power you can just go back to the pentode mode for a factor of 3 to 4 times more. Increasing the cathode resistor will just lower the quiescent current (zero signal) in the amp which is canceled out in the output transformer. Reducing it will give you greater peak to peak current with maximum signal but may also introduce crossover distortion. If you don’t care about maximum power and the amp sounds good to you, your tube dissipation is fine so you could just go with what you have.
If you were going to change cathode resistor value between modes I would parallel the higher value resistor with another resistor using a mosfet switch for the lower value.

On your power dissipation I am getting plenty low numbers figure 11 volts divided by 82 ohms for 132 mA that divided by four for 33.5mA minus screen grid current of approx 4mA for 29.5mA times 300volts(plate voltage – cathode voltage?) for 8.86 watts. Then the screen grid is 4mA times 300 volts for 1.2 watts
Triode mode you have 33.5mA times 300 volts for 10 watts so that is OK

No scope over here! I'm super new at this stuff. But I went ahead and installed the switches. When my hammy OT finally arrives, I'll be able to run Ultra Linear Pentode as well, all with toggle switches.

I used the hole that runs the reverb lines through the chassis for one on the switches. I driled one hole so I could move the sub pcb over. That way I would have enough room to put the toggles. But I had to mod the sub pcb by running new traces, and trimming it for clearance at the spade connection for the OT. But it was no big deal. I used a Dremel cutoff wheel, and grinding stone to round the corners.

Triode connected is a different color. It seems to add more midrange, and is a little richer sounding, and feeling on the strings. It makes the cleans warmer, and rounder too.

Nothing's blown up yet, so as of now, the amp is running well. The fan I have on there is a worthwhile addition. I expect the new OT to really wake up this amp. I'll post more pics when I install it.

Thanks for your help! Really!

 

[jim p]

Nice mods

I tried ultra-linear with a valve jr that I installed an Edcor output transformer on and didn’t notice much difference in the sound. Both ultra-linear and triode mode are basically feedback through the screen grid of the output tube. What is a plus for the ultra-linear mode is the fact the power of the amp is close to the same as the pentode mode. The mods look clean a bit neater then the Peavey Windsor I am carving up. I see the speaker is made in Italy is that the stock speaker?

 

[leadfootdriver]

I tried ultra-linear with a valve jr that I installed an Edcor output transformer on and didn’t notice much difference in the sound. Both ultra-linear and triode mode are basically feedback through the screen grid of the output tube. What is a plus for the ultra-linear mode is the fact the power of the amp is close to the same as the pentode mode. The mods look clean a bit neater then the Peavey Windsor I am carving up. I see the speaker is made in Italy is that the stock speaker?

Thanks for the complements! I would have never gotten this far without the help of people that actually know what they're doing. :master:

I got the amp off' eBay. The guy put a Jensen C12K which is a 100 watt speaker. I received the amp with some intermittant crackling, static. I think that's why he sold it.oke: So I lifted the board and found a lot of metal shavings, presumably from testing a lot of tubes. The shavings were making connections in the wrong places! I cleaned up the board, and haven't had the problem since.

 

[leadfootdriver]

'Got the Hammond installed. Haven't played it that much, but it's good! The Ultra Linear mode is just like Triode mode, but louder from what I can tell.
I can run JJ ECC83S's, and they sound right. You get the presence roll off this amp needs, but the midz cut.



New clip, different settings.

http://www.netmusicians.org/files/64-Veteran 30.mp3

 

[leadfootdriver]

New speaker, different settings. I think this sounds good.

http://www.netmusicians.org/files/64-Veteran 30.mp3