• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Calculating expected output power.

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Is it possible to calculate the approximate output power of a P-P amp, with 811A's as O/P valves, set to run with idle around 75mA, (AB1 ish) at 400 odd volts into 6.6K A-A load Tx, while using MOSFET drive?

What is the ideal Output Tx Impedance for this tube, under voltages of around 400v Va?--Ive seen designs on the 'Net with as low as 2K and as high as 7K anode loads, with voltages around 4-500v Va, and cant find the plate curves for it. (Not that Im much good at reading them anyway!)

Im looking to designing OP TX, but need to have some kind of figure to base its sizing on and a more definate anode load figure.--The current breadboard is using an 'El Cheapo' Danbury Maplin 20W Tx with suprisingly good results, sound wise--Ive done no proper testing apart from listening at the mo.!--Sounds like a lot more than 20W at full o/p, before the Tx saturates and distortion is apparent.

Would I be right in saying that as the Max Anode current of the 811A is 175 mA, and as Ive got P-P, Ive got 200 mA 'Usable Current' at 400V--Around 80w Input, not including bias, to the output stage. Assuming a close to perfect output Tx, and not taking into consideration anode dissipation would be exceeded at full output and all things being equal, could I expect around 50W output?--Or am I completely off-it this morning!
 
Here is a link to the PLATE CURVES......
http://www.triodeel.com/811a_p6.gif

Due to the nature of these curves.... It is not your normal TRIODE design procedure....
It is "almost" like PENTODE operation.......
The plate load you choose here is really all about the PEAK current your able to tolerate.... The lower the plate load the higher the AC peak current....and the higher the drive signal needed.... I will crunch the numbers and get back to you...

Chris
 
WIth 400V at the plates.... and a 6.6K plate load...
You would drive up to +50V on the 811-A grids...with about 70mA of grid current...which roughly looks like driving at input resistance as low as 714 ohms..... so figure you need a follower to drive this with pretty low output Z.....
This would be a power Output of roughly 38.4 Watts....
Figure in Class AB2 you would bias above 0V...figure +20V would give you 60mA each valve....

Chris
 
OK...
You have 400V and a 4K Plate-to-Plate load.....
You make 57 Watts of power output....
Drive signal to +67 volts AC at the grid...
Grid current will be about 100mA .....
This will result in a 670 ohm input grid resistance at it's lowest...so you need to appropriately drive this.... I would drive this with a 50 ohm source minimum.... which is a 7% voltage loss...which constitutes a -.63dB loss at full steam...
Figure you can bias at +20v which is 60mA per valve.....
You need to establish what B+ you intend on using....
In this case you have round up and say 60W on the output transformer at 4K ...this will result in a AC voltage of 490V AC across the full winding....
For the transformer...Figure you want 13,000 Gauss at full power output at 20HZ .... That would work out to be a E-I 150 lamination...square core being 1.5" stack....
SO that would be 3260 Turns.... Using M6 core material, which is grain oriented...... Figure to use #29 or #30 AWG primary wire for about 900 to 1000 CM/Amp...
I leave it to you to figure the secondary turns and winding geometries to best fit your application....
 

PRR

Member
Joined 2003
Paid Member
> Is it possible to calculate the approximate output power of a P-P amp

Yes.

> with 811A's as O/P valves

Ugh. Antique brutes. Shame we don't have anything better.

> idle around 75mA .... while using MOSFET drive?

If you are looking for, or just curious about, POWER, then bias and drive should come out of the power calculations, not fixed in advance. And driver flavor should not be an issue: either you have enough driver to whack the costly output tubes to full power, or you don't.

What is even easier than calculation is to steal the suggested conditions from the tube data sheet. The tube makers had people paid to figure out how best to use the tube (and increase sales). The suggested conditions may not be exactly perfect for you: you compromise with available supply voltage, load impedances, and the changing winds of fashion in THD and distortion spectrum. But they are always a good starting place.

> at 400 odd volts

Here is the first red flag. The lowest suggested condition for 811 is 750V, running up to 1,500V. This suggests that the 811 is "too big" for a 400V supply.

> anode dissipation would be exceeded at full output

Another red flag: if you plot the plot and curve the curves, you will find that you CAN'T get near the plate rating of 811 for any sane push-pull condition at 400V supply. Again, the tube is "too big", and rationally must "cost too much".

RCA probably had another, smaller, less-cathode-heat tube for 400V chores. AAMOF, the 807/6L6 would probably be the go-to tube, when you "think you need" a 811 at 400V. But let's see....

If the 811 in P-P at 750V makes 178 watts, then half the B+ would tend to give 1/4 the power or 44 watts. (Unlike transistors, we can't increase current to infinity.) 44W is essentially the numbers Chris is getting the hard way (38W-57W). And indeed, roughly the number you got from a process not based on actual tube parameters (sure, sometimes you can get the right answer without detailed measurements).

From the plate curves, ALL useful 811 current requires positive grid operation, and grid current, and grid power, and driver power. At 400V on the plate, zero-grid, you only get 10mA! According to the sheet, the 178W condition needs 10 watts of drive, and remember this is a very lumpy load not a pure resistor. The grid drive power for a lower plate voltage won't be less, and may be more (note that the 235W and 340W conditions only need ~4W drive). As an extreme, it would be perfectly possible to reduce plate voltage so you got 20 Watts out and needed 20 Watts grid drive to do it! Even 40 Watts out for 10 Watts in is pretty sad from an engineering point of view.

Optimum load? Over a wide range, it hardly matters: damping will be low, distortion non-negligible. Low-Z load increases output power but needs more grid drive power. How big a driver do you want to build? MOSFETs are a useful cheat; doing it the old way, 2 to 20 watts of driver is hard work.

Instead of "optimum load", we should consider optimum supply voltage. For a given amount of cathode-stuff and heater-power, a tube's maximum current is fixed. Maximum power implies increasing the plate voltage until something bursts: plate is damaged by excess heat or excess electron velocity. Or in the real world, until rising plate voltage at same plate current means a load impedance too high to wind a transformer for. RCA aimed the 811 for 1,000-1,500V and a few hundred watts and several K load, all reasonable conditions.

Another red flag is seen in Chris' calculation. He found a condition where grid current is greater than plate current. If you are not using a step-down transformer, or a separate power winding and a very fat driver tube, then in the process of getting 67V 100mA of drive from a 400V rail the drive power consumed (in grids plus driver) will be MORE than output power. That's doing it the hard way.

What would we do in 1949? If we have 400V and can live with ~40 watts, we would skip over the 200W-300W 811 and find something smaller. How about 807, better known in its low-rated form as 6L6? A couple of these will make 32 watts with 0.001 Watts drive, or 47 watts with 0.1 Watts drive, at around 400V supply. To get the 47 Watt level, the driver could be transformer coupled '50; to make 32 Watts any darn tube will do for a driver. No heroics. And the heater power is 1/4 of the 811! On engineering grounds, this is how you would "use 811 at 400V for 40W": change to 807/6L6 and pocket the difference in cost!

There is another answer but it is wasteful. Bias the 811 at zero volts. That gives, at least for transformer drive, a much more balanced linear input impedance. Then let peak current be about 4X the idle current, or about 40mA peak. This leads to 40K p-p load and less than 8 Watts output, and GROSS inefficiency in heater power, but linearity is much improved over the more conventional condition.
 

PRR

Member
Joined 2003
Paid Member
> using an 'El Cheapo' Danbury Maplin 20W Tx with suprisingly good results, sound wise

OTOH, if you already built it this way, then I say drop on any good 8KCT 50W transformer. But...

> more than 20W at full o/p, before the Tx saturates and distortion is apparent.

A "20W" transformer is 20W only at a specific low frequency and distortion. It will take more power at higher frequencies. If it is good for 20W at 50Hz, it will do 80W at 100Hz at the same THD. Unless you are specifically fuzzing-up on bass tones, you are likely not bending the iron, you are running out of drive or just asking for more than 811 can do at 400V. And in ear-only tests, 20W and 40W are "the same".
 
Thanks for the info

Great Info Cerrem! Im grateful for your input, and have started the transformer design based on your excellent info!

Now, Mr PRR,

Firstly, Im sorry, I have obviously irritated you, and I grant your knowlege is obviously superior to my own. I didnt have the luxury of being in the Industry when valves were the norm, When I started, the best you got was Hybrid sets, maybe with a restoration of a Vox AC30 occasionally!

I Quote--

'Ugh. Antique brutes. Shame we don't have anything better'

What do you consider better?--The Boring more like! 300B anyone?? Maybe PX25, Equally Antique!!

If I wanted to build a valve amp with something as boring as a 6L6, 807, EL34(Ugh, I would never use an EL34, Vile things!) 300B, or a million other well known and equally boring output valves, I would have done so, and followed the croud as many have done before, and Im sure with excellent results..............!

However--

The idea was to make something slightly different out of the very cheap, available, and DIFFERENT 811A I assume when you were referring to Cost, It was component and not energy efficiency, If we were making amps for efficiency, we would be making Tripath based amps! (Been there, Done that!)

You can get at least 10 811A's for the price of a second-hand well used balloon PX25!--I have two NOS specimens of these antquated beasts, which I would never ever dream of making a quality amplifier with! I would rather sell them to the 'easily brainwashed' and NOT be drawn into all that 'This valve sounds better than that one' crap

About the 'never able to exceed the Max Anode dissipation rating with only 400V on the anode--Wrong!

My Bias arrangement follows the rather excellent 'Power-Drive' devised by Tubelabs and is MOSFET based, (Maybe my use of silicon has been the source! of your irritation) the preceeding stages are Mu stage then simple phase-splitter (both Valve Based) During my testing, I was able easily to exceed the anode diss of the 811A, by winding the bias to end-stop, and applying 38V to its grid, I had 135mA Anode current per triode which at 400V is I believe 54W, some 9W over its maximum rating!
The power-drive stage has its own dedicated +150v/-50v supplies, thus the issue of power wastage from the main +b is not an issue

Conserning Output Transformers, Im well aware of the efficiency at different frequencies, as I have had quite some success designing and winding my own. Thus in my statement--

more than 20W at full o/p, before the Tx saturates and distortion is apparent'

I meant that bass lines are distorted at high outputs, which must be exceeding the LF power pass of the Tx Im using, which is a cheap cruddy part! I assumed this would have been clear, evidently not!--The idea of this post was to assess the size of a trans. where this would not become a problem--I would rather other factors limit the the amplifier than the iron, which is dirt-cheap if you design and wind yourself, and gives a much greater sense of achievement!

I will not post the schematic, as Im sure it would cause you even greater irritation.................

So, in ending, I apologise if I upset/irritated you, it was never intended that I do this, with my very stupid questions, But Remember-- No One Knows Everything!
:whazzat:
 
Man, I must be the apex of insensitivity; I didn't pick up irritation at all. In fact, I thought PRR's response was a fine read (as usual) and quite an enlightening exegesis. There are some things in this world that just aren't great ideas, and there's nothing wrong with knocking them down early in the process.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.