Hi,
I just breadboarded the input 6J4 stage of a Pete Millett design, not for that project but just to copy his operating points as a start my Darling amp...
Here it is, only the 6J4 and its plate and cathode resistors pertain Pete has a lot of optional features for this board:
Push-pull driver board
He's running a 6J4 with B+ of 293V, plate resistor of 30K and a Rk of 1K. As you can see he measures 190V right at the plate. I did the same lash up with 293V B+ and came close at 185V at the plate (slight resistor differences). Pete measured 3.4ma, I'm pretty much the same measuring across the plate resistor I get about 3.7ma. The tube grid is grounded. OK all is good, we match.
Then I went to go read the data sheet for 6J4 and it says that the max plate voltage is 150V!
https://frank.pocnet.net/sheets/049/8/8532.pdf
Finally my question is... Is the data sheet max plate voltage a hard stop or does it pertain to something entirely different? I'm confused as to why Pete Millett (a tube god) has 190V on the plate when this tube says 150V max on all data sheets.
Its more likely I probably have a big knowledge gap learning 🙂
I just breadboarded the input 6J4 stage of a Pete Millett design, not for that project but just to copy his operating points as a start my Darling amp...
Here it is, only the 6J4 and its plate and cathode resistors pertain Pete has a lot of optional features for this board:
Push-pull driver board
He's running a 6J4 with B+ of 293V, plate resistor of 30K and a Rk of 1K. As you can see he measures 190V right at the plate. I did the same lash up with 293V B+ and came close at 185V at the plate (slight resistor differences). Pete measured 3.4ma, I'm pretty much the same measuring across the plate resistor I get about 3.7ma. The tube grid is grounded. OK all is good, we match.
Then I went to go read the data sheet for 6J4 and it says that the max plate voltage is 150V!
https://frank.pocnet.net/sheets/049/8/8532.pdf
Finally my question is... Is the data sheet max plate voltage a hard stop or does it pertain to something entirely different? I'm confused as to why Pete Millett (a tube god) has 190V on the plate when this tube says 150V max on all data sheets.
Its more likely I probably have a big knowledge gap learning 🙂
It's like your tires, you can run them at a higher pressure, but they may not last as long.
Also, some brands can behave differently from others.
Also, some brands can behave differently from others.
6J4 is designed for grounded grid configuration. I wonder if the short grid to plate distant has something to do with the conservative spec.
My grid is grounded just to get some accurate DC measurements now, it will be grounded Cathode like the schematic.
8532 and 7245 are the rugged mil version of 6J4 they all have the same plate voltage. Its a frame grid tube. Its been hooked to the audio generator so far about 11 hours at 190V plate voltage, same way as the Millett diagram. No problems, but maybe it would have a shorter life? I think I am going to go ahead and run it right at the max of 150 +10% in the end since it doesn't seem to mind 190V. Maybe Pete will see this and offer some insight!
Of all limiting values, plate voltage is the one whose excess is best tolerated. I saw a lot of commercial designs with Ua 20-30% in excess of data sheet rating. Just make sure that other things, most importantly plate dissipation, are well below limiting values.
An RF tube often is used in Class C.
A typical RF Class C amplifier tube is used to drive a very high Q resonant circuit.
If the tube plate is rated for 500V B+, in that kind of a circuit, the flywheel action of that high Q resonator may take the plate voltage to 1500V, 2000V, 2500V, or more, during the time the Class C tube is cut off.
This can be especially true when the US military ship radioman manually adjusts the transmitter and antenna tuning and loading on a 2-30MHz 500W transmitter.
If he uses to high of an impedance, the voltage may arc.
If he uses to low of an impedance, the tube may melt.
He may also melt the antenna tuner ribbon coil.
Radiomen have to be skilled to do the tuning and loading Quickly, or else the electronic technician has a 500 Watt transmitter that he needs to fix. And those parallel 4CX250 tubes are costly enough that an officer may notice the $$$ charges.
What is worse is when the ship is in the middle of the pacific, and all the 4CX250 tubes have been used up!
Or even more expensive, is a rebuild of an antenna tuner, and there are no spares on board.
A failed equipment like that requires a radiio communication (using another transmitter) to the fleet that the transmitter is out of operation (the captain of the ship looks bad, and he does not like to look bad). 'Nuf said.
Very few Audio amplifier tubes are designed to be used that way. And very few Audio output transformers are designed to be abused that way.
And a single ended class A amplifier with a tube that is in the cutoff region sounds really bad.
Another example of plate voltage ratings has to do with other ways that the tube will be used:
1. A TV Horizontal Flyback tube.
2.A TV Vertical Amplifier tube (that drives the Vertical CRT yoke).
A 6SN7 maximum plate voltage rating is 300V. But, the same tube in Vertical Amplifier Service is rated for 1200V peak (during the time the tube is cut off, and the Yoke voltage has Inductive kick-back.
One more example:
An 807 Beam Power tube (or Beam Tetrode tube as some call it), was made for RF service, including class C, and also for Modulator service.
In Beam Power mode, the screen maximum voltage is 300V.
But in Triode wired mode (screen and plate tied together), the screen maximum voltage is 400V.
The 807 was a very Robust tube. Some are better than others. But if you abuse the tube enough, Robust is converted to Busted.
As with all tubes, and high rev reciprocating gas engines . . .
Your Mileage May Vary.
A typical RF Class C amplifier tube is used to drive a very high Q resonant circuit.
If the tube plate is rated for 500V B+, in that kind of a circuit, the flywheel action of that high Q resonator may take the plate voltage to 1500V, 2000V, 2500V, or more, during the time the Class C tube is cut off.
This can be especially true when the US military ship radioman manually adjusts the transmitter and antenna tuning and loading on a 2-30MHz 500W transmitter.
If he uses to high of an impedance, the voltage may arc.
If he uses to low of an impedance, the tube may melt.
He may also melt the antenna tuner ribbon coil.
Radiomen have to be skilled to do the tuning and loading Quickly, or else the electronic technician has a 500 Watt transmitter that he needs to fix. And those parallel 4CX250 tubes are costly enough that an officer may notice the $$$ charges.
What is worse is when the ship is in the middle of the pacific, and all the 4CX250 tubes have been used up!
Or even more expensive, is a rebuild of an antenna tuner, and there are no spares on board.
A failed equipment like that requires a radiio communication (using another transmitter) to the fleet that the transmitter is out of operation (the captain of the ship looks bad, and he does not like to look bad). 'Nuf said.
Very few Audio amplifier tubes are designed to be used that way. And very few Audio output transformers are designed to be abused that way.
And a single ended class A amplifier with a tube that is in the cutoff region sounds really bad.
Another example of plate voltage ratings has to do with other ways that the tube will be used:
1. A TV Horizontal Flyback tube.
2.A TV Vertical Amplifier tube (that drives the Vertical CRT yoke).
A 6SN7 maximum plate voltage rating is 300V. But, the same tube in Vertical Amplifier Service is rated for 1200V peak (during the time the tube is cut off, and the Yoke voltage has Inductive kick-back.
One more example:
An 807 Beam Power tube (or Beam Tetrode tube as some call it), was made for RF service, including class C, and also for Modulator service.
In Beam Power mode, the screen maximum voltage is 300V.
But in Triode wired mode (screen and plate tied together), the screen maximum voltage is 400V.
The 807 was a very Robust tube. Some are better than others. But if you abuse the tube enough, Robust is converted to Busted.
As with all tubes, and high rev reciprocating gas engines . . .
Your Mileage May Vary.
Last edited:
Interesting ratings data
John Atwood's archive includes RCA 1958 application note AN-174 that explains the different rating systems: Absolute-Maximum, Design-Center and Design-Maximum here:
http://www.one-electron.com/Archive...-Maximum System for Rating Electron Tubes.pdf
The original 1944 registration for the 6J4 showed a Design-Center maximum of 150V. Usually this is the most conservative rating. In the instances where a Design-Maximum rating can also be derived, the plate voltage is around 25% higher than the Design-Center rating, or 187.5V for the 6J4. On reading the RCA application note, I'd be comfortable with Pete Millet's voltage.
What is really interesting in Windcrest77's snip of 6J4 data is that it shows an Absolute-Maximum rating of 150V which seems to apply to operating at altitudes up to 80,000 ft and frequencies up to 500 Mc/s. This condition wasn't stipulated in the 1944 6J4 registration - it corresponds to the manufacturer's 1964 registration data provided for the 8532/6J4, a premium version of the 6J4 that appears to be equivalent to the 6J4WA and 6J4WB.
John Atwood's archive includes RCA 1958 application note AN-174 that explains the different rating systems: Absolute-Maximum, Design-Center and Design-Maximum here:
http://www.one-electron.com/Archive...-Maximum System for Rating Electron Tubes.pdf
The original 1944 registration for the 6J4 showed a Design-Center maximum of 150V. Usually this is the most conservative rating. In the instances where a Design-Maximum rating can also be derived, the plate voltage is around 25% higher than the Design-Center rating, or 187.5V for the 6J4. On reading the RCA application note, I'd be comfortable with Pete Millet's voltage.
What is really interesting in Windcrest77's snip of 6J4 data is that it shows an Absolute-Maximum rating of 150V which seems to apply to operating at altitudes up to 80,000 ft and frequencies up to 500 Mc/s. This condition wasn't stipulated in the 1944 6J4 registration - it corresponds to the manufacturer's 1964 registration data provided for the 8532/6J4, a premium version of the 6J4 that appears to be equivalent to the 6J4WA and 6J4WB.
It's a $4 tube. Why should you care about over-volting it? Buy a bucket-full and keep them by your chair.
Figured it out, but why?
OK,
I think I figured this out, Pete Millett can safely run the 6J4 at 190V plate voltage! It's all in the conditions of my measurements. Because that measurement must be a static one with the grid grounded. When the tube has a signal the plate voltage drops considerably, but when the grid is grounded the plate voltage jumps to over its max rating. In real life you wouldnt be running it that way.
Here is what I observed but dont fully understand yet... The data sheet said 150V is the max plate V as created with the grid grounded.
At Pete Millets 293V B+ and 30K I get 190V on the plate exceeding the max. This was seen with grid grounded. This tells me that Pete grounded the grid to to get that value too.
But if I simply disconnect the grid from ground and apply a signal, that plate voltage drops to 120V with the same 293V B+! And I have to increase the B+ to over 240V before the plate gets to 150V. So plenty of safe room here.
I think this explains it, Petes operating point is just fine and well under the max under normal dynamic operation. I was just measuring it under the conditions that the data sheet said, which was to ground the grid, that would be the static operation I guess. So with a signal that tube is well under its max.
Also observed is when I lower the B+ to a measly 100V (its on my bench supply). I can visually see the sine wave flatten out on one side as it goes more non-linear. But when I crank the B+ back up to 293 (or even higher), the sine wave peaks look very equal on each side. I assume this is because the tube is operating more linearly at that higher voltage and has a wider available clean swing on the load line?
Learning every day.
OK,
I think I figured this out, Pete Millett can safely run the 6J4 at 190V plate voltage! It's all in the conditions of my measurements. Because that measurement must be a static one with the grid grounded. When the tube has a signal the plate voltage drops considerably, but when the grid is grounded the plate voltage jumps to over its max rating. In real life you wouldnt be running it that way.
Here is what I observed but dont fully understand yet... The data sheet said 150V is the max plate V as created with the grid grounded.
At Pete Millets 293V B+ and 30K I get 190V on the plate exceeding the max. This was seen with grid grounded. This tells me that Pete grounded the grid to to get that value too.
But if I simply disconnect the grid from ground and apply a signal, that plate voltage drops to 120V with the same 293V B+! And I have to increase the B+ to over 240V before the plate gets to 150V. So plenty of safe room here.
I think this explains it, Petes operating point is just fine and well under the max under normal dynamic operation. I was just measuring it under the conditions that the data sheet said, which was to ground the grid, that would be the static operation I guess. So with a signal that tube is well under its max.
Also observed is when I lower the B+ to a measly 100V (its on my bench supply). I can visually see the sine wave flatten out on one side as it goes more non-linear. But when I crank the B+ back up to 293 (or even higher), the sine wave peaks look very equal on each side. I assume this is because the tube is operating more linearly at that higher voltage and has a wider available clean swing on the load line?
Learning every day.
Last edited:
> When the tube has a signal the plate voltage drops considerably
Why would it do that? Changing DC with AC level is another way to say "rectification" or "2nd harmonic distortion". 20+% 2nd harmonic is a LOT for speech/music signals.
Why would it do that? Changing DC with AC level is another way to say "rectification" or "2nd harmonic distortion". 20+% 2nd harmonic is a LOT for speech/music signals.
Look at the plate curves in the data sheet. They extend out to 300 volts.....if the tube is only rated for 150 plate volts, why would the manufacturer test it to 300 volts, then publish the data?
Because they EXPECT the peak plate voltage to reach 300 volts, as it will in an RF amp with an inductive load. The plate voltage in a single ended amp with an inductive load will swing upwards to nearly twice the B+ voltage in a normally operating circuit without being driven to clipping. It can actually go higher than that if resonance is involved, and the circuit is driven to clipping.
If you run the tube at 293 volts with a resistive load, the plate voltage may idle at 190 volts with plenty of room to swing above and below that 190 volt quiescent point. As long as the plate dissipation is kept below about 2 watts, nothing bad is going to happen in this situation.
Because they EXPECT the peak plate voltage to reach 300 volts, as it will in an RF amp with an inductive load. The plate voltage in a single ended amp with an inductive load will swing upwards to nearly twice the B+ voltage in a normally operating circuit without being driven to clipping. It can actually go higher than that if resonance is involved, and the circuit is driven to clipping.
If you run the tube at 293 volts with a resistive load, the plate voltage may idle at 190 volts with plenty of room to swing above and below that 190 volt quiescent point. As long as the plate dissipation is kept below about 2 watts, nothing bad is going to happen in this situation.
I routinely run plate voltages in the 600 volt range, and about double the plate dissipation as listed on data sheets....Even on the small signal tubes...... Been doing this for about 12 years now.. No tube ever complained.......The only thing I keep in check is the filament voltage. I like to see them running at abouts 6.0 - to 6.1 volts rms.