811 Rig Upgrade

[K9ACT]

Never being happy when something just works well, I decided to soup up my 811 rig by adding another in parallel.

In addition to the tube stuff, I cut the grid leak in half and re-worked the PI output and fired it up on 80 meters.  As hoped for, it put out about twice the power (200W) with about twice the input power and worked just fine.

I then switched over to 40 Meters and got real humbled.

Doubling the input produces very little increase in output.  I can not get more than 120W out and the tubes get red when I do this.  I have spent hours trying different ideas and nothing helps.  I have 4 tubes and they all put out 90-100W in either socket at 900V and 150 ma but any two produce little more at 300 ma.

Schematic of single tube is at http://schmidling.com/811_am.png

Any ideas are welcome.

Jack

[W3RSW]

I guess you've done all the obvious but dumb as it sounds do you now have twice the fil. current, 8Amps at 6.3v.   Orig. fil. xfor might put out 200 watts on 80 meters at a sucked down 4 or 5 fil. volts but not on 40 meters.

Some interaction between parasitic generation on 40 but not on 80? Try a parisitic suppressor on the plate between plate and 155uh RFC. I see you have one on the grid.  Could it be close to blocking resonance on 40?  Or Or coupled with your neut. circuit could it be upsetting the main input tank on 40? Doesn't have to be right on, but anything close would reduce power to grid.

Might try another plate choke for that matter.  155uh might be sucking down 40mtrs in the usual strange partial winding mode.

I'll keep thinking.  works on 80, sucks down on 40.  oughtta be a big hint for us there.  Backwards from what you'd expect if the pi net loading cap was insufficient on 80.

[KB5MD]

Do you know for sure that the 40 meter tap (?) is in the right spot on the tank coil?

[W3RSW]

Right.  We assume you've changed both 80 and 40 meter pi net taps since power easily doubled on 80?  Assume you're working a dummy load for starters.

Plates turn red at ordinary load with reduced output. Sure sounds like parasitics. But it worked on 40 with one tube ok.  ....velly intellesting.  What with increased current would cause a droop on 40 but not on 80? Circuit is about as basic as you can get. Do you get 1:1 or so SWR on input link on 40 at the higher input (now 10 watts? )- 5 sounds way low for even one tube. 

 My book shows -125v on grid of 811's for C-P service at 1250 volts. But needs 11 watts drive for 120 output/tube.  Ip is 125ma, Ig is 50ma.

812's take only about half the power input and show half the grid current for same 120 watts output/tube.

As you know C_out, C_gp, etc. are changed so you've re-neutralized for the highest frreq. to be worked.

I'd still drop a parastic suppressor in the plate circut as a matter of course and 'modern' design.

Might want to put a freq. meter on the output with a 40 mtr. trap in the lead to see if any other wierd freq.s show up from self oscillation.

Hope your not insulted; just writing down everything I can think of. At my age I'll run out soon 

[Ralph W3GL]

Sounds like a "Q" problem to me...

[W3RSW]

could be. -
you think the windings of the final pi net coil are too small a guage for higher power, lower R_p? - or maybe the construction of the input coil?

Hmmm, might have something there. The plates turning red show dissapation over design. The power is obviously being radiated inside the tubes.... being bottled up by the bottles, heh, heh.  So it got to the tubes ok which shows the input circuit working.  A parasitic or triggered oscillation won't get through the output net at the wrong freq. 

[W3RSW]

Jack, any luck yet?  Is 40 mtr. drive sufficient for two tubes? i.e, 22 watts?
Maybe you should've used 812's, heh, heh.

You could be getting sufficient drive (grid current) on 80 CW to get full CW output, but not sufficient drive for AM service. - then shows up big time on 40 as not even enough for CW service.

Well, easy enough to change to 812's if you've done all the other stuff. Just add some bias, 1/2 coming from a VR75 or zener, other half from your grid resistor. Come to think of it 1k sounds awfully low. Was that a typo?  1k at 50 ma yields only -50 volts.  Two tubes require 100 ma. ( 811's) and the 1k still only yields -100 volts.

Interestingly 811A's require only 40 ma for CW service, 811's, 35ma CW instead of 45 and 50 respectively for AM service. (sounds backwards but just reading the tables.)

[K9ACT]

Jack, any luck yet?  Is 40 mtr. drive sufficient for two tubes? i.e, 22 watts?

No joy so far.  I want to thank everyone for the ideas, most of which I have checked out and some still pondering.

Where does the 22 watts come from?   I see 6 per tube.  Is it possible that leaving off the A as in 811A is a non trivial detail?

>Maybe you should've used 812's, heh, heh.

Well, I have the extra 811A but I had another brainstorm but it does not work any better.

I happen to have an SB200 collecting dust so I pulled one of the 872B's and replaced the 811A with this.

This seemed like an alternative way to get more poop.  The tuning is totally erratic.  Like when I dip the plate, the grid current goes to zero and neutralizing produces similar nonsense.

>You could be getting sufficient drive (grid current) on 80 CW to get full CW output, but not sufficient drive for AM service.

Not sure I understand.  It's not AM till I talk.  I am just trying to get RF out of it.

>Well, easy enough to change to 812's if you've done all the other stuff. Just add some bias, 1/2 coming from a VR75 or zener, other half from your grid resistor. Come to think of it 1k sounds awfully low. Was that a typo?  1k at 50 ma yields only -50 volts. 

The single tube version has a 1K grid leak and the two tube is 500.  The book says 1200 but I didn't happen to have one and I am not sure how running at only 900V effects all this.

js

[W3RSW]

oh, ok you are using 811A's.  Orig. post just said "811 rig" so I assumed both.

Be that as it may, only in class B audio service does my book say 4.4 watts drive for two 811A tubes in PP.
In class C, phone mod. plate voltage is recommended at 1250. In class B audio and 'C', CW service book shows 1500. You can see up to 1800 in class B audio, i. e. Johnson 500 at 1800.

Now as for drive requirements, per tube 811A's at 1500 class C, CW req's 7.1 watts per tube.
At 1250 volts  Class C, AM req's 10.0 watts/tube.  (So 20 for two.)  For a little headroom above 100% pos. modulation, circuit losses will need somewhat more.  Heck maybe even 25% for circuit losses alone.

Now in linear service,()diff. chapter in '57 ARRL book,) 1500 volts plate, -4.5 grid, 2.2watts max. sig. driving power.   Possibly confusing  class B or audio SSB service with class C, CW or AM power.? -or just grabbed the wrong table in haste?

Yeah, you get AM when "speaking into the tube"  but your looking at lower plate voltages to allow dissapation headroom from superimposed audio as well as far more negative bias, etc. than class B service.  Don't want to write a ton but you can force air cool, operate shorter times, etc. and get a whole lot more out of an 811A, even much more out of a 572B but you need plate voltages way higher than 900 to do so efficiently, as you know.

I think 900 is way under-rating the efficiency possibilites of the 811A.  Anyway, that's moot at this point.
 The 572B' giving you really squirrily readings sounds like something really upset in the parasitic or self -oscillation area. Weird. I'll keep thinking.  Check all the bypass cap's ; maybe just replace with new ones having a max. voltage value twice or three times what you have just to eliminate possibilites there. They can be a real possibility of something being squirrly as a direct function of freq. and power.

[K9ACT]

Now for some good news to report.

I switched from my HB exciter to the Kenwood 430 and it more or less worked right off.  200W 300 ma just as planned.  The only problem is that I had already tried that and it didn't seem to help.

I put a power meter between the Kenwood and the 811's and measured 25W for 80 ma grid current.

There is no way I can get that much power out of my exciter but I have no problem getting 80 ma grid current with it.  The most I can get out of the exciter is 12W and I have no problem driven my pair of 8000's with it or this thing on 80 meters.

On the other hand, my 811A data sheet says 6.1W to drive one at 1000V.  That is only 115 ma plate current and 88W out and I do not know how all this things relate.  I never could make any sense out of tube curves.

js 

[W3RSW]

Funny, I beat up all kinds of possibilities of what might be wrong in the transmitter but other than power totally missed the possibility of a crappy signal from the a driver  .

Just 'shows to go ya...'

Glad you got it working.

[K9ACT]

Funny, I beat up all kinds of possibilities of what might be wrong in the transmitter but other than power totally missed the possibility of a crappy signal from the a driver

Don't feel like the Lone Ranger.  Why would one suspect the driver if it works just fine on 80 and produces the same grid current on 40 but does not work?

It still makes no sense.  Just a problem solved but not the mystery.

js