Need help interpreting tube plate curve

[Solidus]

Normally in a tetrode or a pintode the bias requirements are directly proportional to the screen grid voltage. For example if the screen grid voltage of a 4-125 is 600 volts the the control grid DC bias should be around -94 volts. If the screen grid was 300 volts the the DC bias on the control grid should be about -47 volts. The reason for this has to do with space charges within the vacuum of the valve. The control grid characteristics has little to do with plate voltage. So my suggestion would be ,if you want to run lower voltages, is to reduce them all by the same ratio with the exception of the heater voltage. It must remain around 5 volts or damage to the valve will occur.

One thing about running tetrodes or pintodes as triodes. A tube like a 12AU7 has a mu or AC gain of about 70 if I remember right and a 12AX7 has a mu of about 100. These valves are purposely designed to have lots of gain. I don't know what a 6V6 would give you if hooked up as a triode, maybe 10 or 20. I don't know but certainly would be much worse than most triodes designed to be voltage amplifiers.

Okay, now, does this mean that I can run the B+ at 800V and still run screen at 600 and grid to -94?

Well, seeing that it's late at night and my mind no longer is in its prime to think about circuit logistics, I figure we can talk a little about tubes for fun.

That makes sense what you're saying - triode connecting power valves doesn't give near the same voltage efficiency, which is why SET (single-ended triode) amplifiers using directly heated triodes like the 300B and 2A3 are known to be some of the worst in the efficiency category.

Which perplexes me, because I always thought that the whole upside to using tetrodes and pentodes was not only their increased capacity for voltage gain and efficiency, but also their linearity of amplification. After all, I thought that the whole craze with audiophiles and Hi-Fi category amplifiers was to generate more "raw" audio, that is, undistorted and more true to the recording.

12A series tubes are downright dirty when it comes to THD as I said earlier when driven hard - a lot of higher-fidelity amps use instead the 6SN7/6SL7 variations. Dual triode, but different structure and tolerances means a lot more linear levelling comes out.

The downside of triodes isn't just noticed in audio spectrums, but in everything else as well. The most popular series of Eimac tubes is their tetrodes, which is produced in as little as a 65W tube (4-65A) and up to 35kW dissipation, the 4CX350000D - be thankful I didn't post asking how to run that!

I have been able to track down, with a lot of effort, Eimac RF triodes, such as the 3-500 and 3-1000Z, but aside from the datasheets these tubes are rarer than the Yeti. They had really specialized uses, even in radio circles, and if I'm ever able to collect either of those, I'll consider myself lucky and turn it into a nightlight. Even rarer than that is Eimac pentodes - as far as I've been able to find, only the 5-500 exists.

Makes me wonder why most of the audiophile world is hellbent on directly-heated triodes for their end stages - even a bargain-bin brand 300B runs $300 for a matched pair nowadays, and I've seen precision-matched audiophile-grade 300Bs online for$1200. Don't care how rich I get in later life, if I ever try to buy those feel free to find me and beat the common sense into me.

That being said, I think most of that culture thrives on the fact that a lot of these are subjective notions. We can look at datasheets and curves and see linearity, but what does that mean exactly? This is food for thought that if I make a bunch of money making amps (hey, a kid can hope, right?), maybe I'll be able to afford $10,000 of analyzers and a modern oscilloscope unlike my ancient S51B Berkeley stock and possibly quantify this madness.

Until that day arrives, I still would rather judge an amp by my ear.

In practicality though, I want this build to have a good mix of sounding good and being efficient. No super-audiophile-grade $100/piece capacitors, but at the same time no "Class A all the way!" and burn out my output trafos after a day of use.

 

[john monks]

I'm not much of an audiofile but I do go through a lot of audio amplifiers. I go to a recycling center and pull them out of the trash and if they're working fine, which is about half the time, I give them away. The other ones I take abort for the parts, and sometimes I will repair them. All of them are solid state and I do believe that the modern solid state audio amplifier is simply outstanding. It was the earlier amplifiers that you noticed crossover distortion and other weird anomalies in the sound. The tube type amplifiers alway sounded good even from the 1930's.
Now the pintode and the tetrode are inherently non-linear and need negative feedback in order to get rid of distortion. Triodes are inherently very linear. So I can understand that. And if you stay towards class A amplifiers you won't get much distortion.
I've been experimenting with class B push-pull amplifiers with lots of negative feedback and noticed that even though I could see no distortion with an oscilloscope I could hear the crossover distortion and it simply sounded terrible even though my total harmonic distortion is only about 1%. So my conclusion is that the human ear is frequently much more sensitive to distortion than what can be measured using an oscilloscope.

If you use 800 volts for the B+ and 600 volts for the screen grid then you will need about -94 volts bias on the control grid. The plate current is not affect very much by the plate voltage within reasonable limits. This is because the accelerating current is controlled by the grids closed to the cathode. The space charge set up just beneath the plate is set up more by the screen grid than the plate simply because of the proximity of the screen grid.

By the way in a triode the mu or gain of the triode is set up by the ratio of the control grid to the cathode to the plate to the cathode. How finely the screen is wired also has some effect. So if you ever take upper division physics these concepts will be helpful.

 

[Solidus]

I'm not much of an audiofile but I do go through a lot of audio amplifiers. I go to a recycling center and pull them out of the trash and if they're working fine, which is about half the time, I give them away. The other ones I take abort for the parts, and sometimes I will repair them. All of them are solid state and I do believe that the modern solid state audio amplifier is simply outstanding. It was the earlier amplifiers that you noticed crossover distortion and other weird anomalies in the sound. The tube type amplifiers alway sounded good even from the 1930's.
Now the pintode and the tetrode are inherently non-linear and need negative feedback in order to get rid of distortion. Triodes are inherently very linear. So I can understand that. And if you stay towards class A amplifiers you won't get much distortion.
I've been experimenting with class B push-pull amplifiers with lots of negative feedback and noticed that even though I could see no distortion with an oscilloscope I could hear the crossover distortion and it simply sounded terrible even though my total harmonic distortion is only about 1%. So my conclusion is that the human ear is frequently much more sensitive to distortion than what can be measured using an oscilloscope.

If you use 800 volts for the B+ and 600 volts for the screen grid then you will need about -94 volts bias on the control grid. The plate current is not affect very much by the plate voltage within reasonable limits. This is because the accelerating current is controlled by the grids closed to the cathode. The space charge set up just beneath the plate is set up more by the screen grid than the plate simply because of the proximity of the screen grid.

By the way in a triode the mu or gain of the triode is set up by the ratio of the control grid to the cathode to the plate to the cathode. How finely the screen is wired also has some effect. So if you ever take upper division physics these concepts will be helpful.

So the gain is proportional to (g1:k) : (p:k)?

Do you know much about the idea of the critical distance effect? That is, the notion that led to the invention of kinkless tetrodes - the specific multiple of grid-screen-plate distance that allowed for maximum power transfer?

What would the approximate current flow at the plate be at those operating points for the 4-125A?

 

[john monks]

In a triode the gain or the mu is approximately proportional to the distance of the control grid to the cathode and the distance of the plate to the cathode. However other factors affect gain also such as how fine the grid is wired.

In a tetrode kink is the secondary emission of electrons coming from the plate to the screen grid. This has the affect of causing the voltage if the plate to be pulled toward the screen voltage when the plate voltage dips below the screen voltage. This can cause parasitic oscillations and sometimes provisions need to be made in transmitters to snub this oscillations because this robs the amplifier of power. I suspect that this kink occurs when the plate is hotter that the screen grid so that the electrons can flow in this reverse direction but I never studied.

To study this would be interesting because we are only looking at current flow and space charges and the effects of different temperatures of elements in a vacuum in 3-D space.

I never analyzed thus very

 

[Solidus]

In a triode the gain or the mu is approximately proportional to the distance of the control grid to the cathode and the distance of the plate to the cathode. However other factors affect gain also such as how fine the grid is wired.

In a tetrode kink is the secondary emission of electrons coming from the plate to the screen grid. This has the affect of causing the voltage if the plate to be pulled toward the screen voltage when the plate voltage dips below the screen voltage. This can cause parasitic oscillations and sometimes provisions need to be made in transmitters to snub this oscillations because this robs the amplifier of power. I suspect that this kink occurs when the plate is hotter that the screen grid so that the electrons can flow in this reverse direction but I never studied.

To study this would be interesting because we are only looking at current flow and space charges and the effects of different temperatures of elements in a vacuum in 3-D space.

I never analyzed thus very

I looked into the tetrode kink one time.

The logic being that in the very low plate voltages, the screen grid is driven harder than the plate, and because of the low charge, electrons aren't accelerated highly enough and tend to create secondary emission or "bounce" off the plate where instead they are captured by the screen, thus meaning at certain plate voltages the screen can carry a high enough dissipation to endanger the ratings.

That secondary emission is, from what I've read, related to the photoelectric effect - it's the basis of photomultipliers, as well as Geiger counters. Idea being in a nutshell, the electrons must strike the atoms of the plate, and when they do, they create occasionally an avalanche effect resulting in emission of electrons not related to current flow.

At very low plate voltages below the kink, electrons don't have enough accelerant energy to cause this to a significant degree, and above the kink they do, but the plate differential is enough to recapture the electron almost instantaneously. This is what I've garnered.

I understand that this creates a small area in the tetrode characteristics where there is negative resistance, and thus altering transfer adversely.

The pentode (and later kinkless tetrode/beam tetrode) was a fix to this, using a grid at reference potential between the screen and plate to "trap" electrons at the plate. Downside to this is that because the negative screen grid is effectively also a modulating grid, the screen grid has to be driven harder.

Just my 2 cents from what I've studied.

 

[john monks]

And what you say tends to line up with my personal experience because I remember hooking up tetrodes 6V6's and allowing the plate voltage to drop to 0 volts and I noticed that the screen grid became red hot and in some cases the screen grid simply melted away. And I never had that experience with pintodes. So I think your analysis is correct.

 

[Solidus]

I'm working on the final touches for the schematic, will post with questions in a few hours when I have them done.

Was looking on Digikey and found a Hammond output transformer around 10-15k primary for $20...started thinking, good deal, better than$172 that I was planning on spending.

I looked at the datasheet...3.5W output. 4-125A-driven smoke machine will have to come later.

What is the expected screen current draw at 600V? The datasheet only provides a fixed current number for 350V, and I need the exact number so I can calculate resistor values for the screen supply.

I'm looking at using a 47nF coupling capacitor at the 4-125 grid. Does that line up or should I move up in capacitance?

 

[john monks]

From looking at the datasheet I would expect somewhere around +/- 1ma screen current.
47nF coubling capacitor looks good to me with about a 100kohm resistor betwen the control grid and the bias supply.