813’s, Headroom, and IMD?

[KC9LKE]

There is some discussion about IMD, K1JJ’s HPSDR post, and designing for headroom, KB1HIS’s QRO post.
So I was curious about typical IMD numbers for a pair of 813’s in class C using a three element (pi net), and, lets say, 500W carrier into a 50 ohm dummy load?.
I know this is overkill but it should be cleaner at 375W. Yeh it’s a broad question, Just wondering what to truly expect out of a pair, and since the 813 is underrated on the data sheets.

Thanks

Ted / KC9LKE

[WA1GFZ]

class c amplifier has poor imd. As a linear you need a higher conduction angle Class AB to A. If you plan to modulate a class C stage then there are many factors to consider.

[K1JJ]

Hi Ted,

It's good you're thinking in terms of designing a rig for cleanliness.

Generally, IMD figures are done with tubes in linear service. The 813 run in GG linear has some NFB, so is usually clean. When run in grid driven linear, they are just "OK," no better or worse than other tubes like this that are designed more for class C service.

But for class C plate modulated service, as you desire, the overall on-air IMD figures will GREATLY depend upon how clean your modulation system is. That will be the overriding factor.  

From experience, I've found a pair of 813's modulated by a pair of 813's in triode connection to be at least -30db 3rd. (estimate) This assumes the audio driver to the modulators is clean. If you run the recommended voltages and currents to the 813 finals, then don't worry about the finals - the audio deck is where to put the major effort via NFB and other good engr practices.


What do I know about good engr practices? Evidently, not much. Yesterday I  dumped 200 watts into my Mercury SDR receiver board and fried the front end. Picked the wrong cable and it became the dummy load... [sigh]

[KC9LKE]

Thanks for the replies guys!

That’s what I needed to know.

I’m a long way from the modulator, but I was considering 810’s with the mosfet driver, and 1.3:1 mod iron. I have some options for the NFB but more on the mod deck many months from now.

Good luck with the HPSDR!

Thanks again

Ted KC9LKE

[k4kyv]

I think what you are looking for is modulation linearity.  The best indication is to connect an oscilloscope for trapezoid pattern.  Feed the horizontal plates with a sample of the audio coming directly from the modulator, and vertical plates with a sample of the rf output from the transmitter. If the class C final is modulating linearly, the edges of the triangle at 100% modulation should be perfectly straight.  If significant curvature appears at the edges or abrupt breaks appear in the line, then the final has poor modulation linearity.  This is likely to cause distortion and a broad signal, for the same reasons that a "nonlinear" linear amplifier does. It can also explain why the positive peaks on the envelope pattern may appear less than what you think they should be.

Look in any older handbook to see how to take the audio sample from the modulator.  Normally, a resistive voltage divider and DC blocking capacitor are used. The greatest problem is likely to be undesirable audio phase shift that muddles up the edges of the trapezoid. If this cannot be corrected using a capacitor somewhere in the voltage dividing network, run a sweep with a signal generator, and you should be able to find a single frequency tone that exhibits little or no phase shift, and use that tone to run the test.  The result will still indicate modulation linearity/non-linearity.

Avoid the bogus trapezoid set-up I have seen described that derives the audio sample by rectifying some of the rf output.  That method might work for checking modulation percentage, but the result is totally meaningless for determining modulation linearity.

[w1vtp]

But if you really mean distortion products from modulating a class "C" 813 final and want to quantify the distortion then you are looking for THD of the modulated AM carrier.  If that's the case, I refer you to my post on doing THD of my EICO 720.

http://amfone.net/Amforum/index.php?topic=16771.msg115775#msg115775

I used lab equipment but the same thing can be done with a calibrated receiver such as a Flex 5000.  Be advised that this means long duration key down transmissions into a dummy load.  I think it's more practical to consider Don's suggestion on using a Trap.. scope pattern.

On another note, Stu AB2EZ had an excellent method of capturing faithful  modulation by superimposing the modulating AF with the actual RF envelope  - dual trace scope looking at the modulating AF over the RF envelope but when I went to his site it was gone.  Stu!  Where did it go? I should have downloaded the page when I had a chance.

Al

[KM1H]

The class C final bias and grid drive is also important as the tube(s) require to be run into deeper Class C than for CW.

An 813, and other multi rid tubes, also likes to have the screen receive a bit of audio and there are many discussions on ways to do that.

Its fairly easy to obtain good audio, its another big leap to get great audio and I dont mean just the hi fi bandwidth that some seem to require for whatever reasons.

Carl

[KC9LKE]

OK Thanks for all the direction, I just want a good clean setup, so I appreciate the input. I'll print this off and use it when time allows.
I have an old scope, "like" the millen, that I plan on using full time.

Thanks
Ted / KC9LKE