Linear amp NOOB

[W7SOE]

OK, what am I missing?

At Winterfest I bought a linear from W7JHU that was built by K7LFE.

It is a single 813 in GG configuration.  There is a single meter for plate current.

I am feeding it with the ricebox in CW mode, it dumps into a dummy load.  I can see the output on a scope.

When I switch on the HV I see carrier on the output and about 80mA on the plate.  The tank has a roller inductor.  Turning it and the cap has no effect on the plate current. 

The 813 plate glows cherry red and the I switch it off.

What is this linear amp NOOB missing.


Rich

[W7SOE]

I seem to be over driving it.  Lowering the drive lowers the red glow on the plate.  There is still a glow though while only driving it with ~4 watts....

Still having trouble getting a good dip on the plate meter.


Rich

[w4bfs]

hi .... seems like resonance is hard to find .... do you have a grid dip osc or ant analyzer ? .... these could tell you what is going on .... a single 813 running grounded grid with 2000 plate volts needs a matching network of 5000 to 6000 ohms impedance ... looking at a hand book with pi-net values shows you a good preset value of Cin, Cout, and L that are reasonable starting values to work from ....is this more than you want to do ?  then maybe you want something a bit more commercial ... 73 ...John

[K1JJ]

Rich,

It appears that everything is working except the pi-network output or TR relay contacts.

Work with John's good ideas and if you're still stuck:

Get in there and look over the tank wiring carefully to make sure it is wired like a standard pi-network that you find in the handbooks.

Then use your VOM to be sure there is a DC connection between the final variable loading capacitor hot pin and the S0-239 output when the amp is keyed up.

You may have to disconnect the plate tuning cap from the coil to check that it is not shorted to ground.

Also be sure the plate choke is OK and not shorted out on itself, or RF will be bypassed into the power supply.

Let us know what you find and we'll go from there.  This should be an easy thing to solve.

Tom, K1JJ

[WA1GFZ]

You might be tuning up on a harmonic. Remove the high voltages from the rig and drive a little rf into the input. Look at the output on the scope and try to tune for max signal. this will get you close. Do this after you check the relay and controls.

[ka3zlr]

Well, A couple little things first... is this an all band amp..are we operating in the amps operating area....and do we know the toobie to be a good one...sometimes little things do the worst ....

[W7SOE]

Well, Tom et Al,

You were right, it was time to dig in there and check stuff.  Tuning up on the exciter with the HV off was a good idea.  While diddling with it for 13th time I noticed that the output would jump and stabilize if I pushed on the roller inductor knob.   I cleaned some contacts and now it behaves as one would expect.

Boy, it takes very little to drive this thing.  I get 130 watts out while driving it with 4 watts from the ricebox. 

While fiddling with the exciter I looked down to see the wattmeter at 350 Watts.  350 Watts!  Cool! That is a lot of carrier.  Oh, wait.  There is only a single 813 in there.
Sure enough I had increased the input power so that things were getting quite toasty in the amp!  Killed the power and waited for the heat and cooked electronics smell to dissipate.  It looks like this thing will take whatever you give it, right up until it melts down.....  OK, I've learned that lesson.

At 4W in I get ~135W out.  The plate in the 813 glows a very dull red.  Is this ok?

I suppose I could measure the plate voltage to see where I am running.  I am guessing that this amp was designed to run the tube flat out.  This 813 is of the steel plate variety, I wonder if a graphite version would fare better.

Thanks all for the help.

Rich

W7SOE

[W7SOE]

Another update:

Measure the plate voltage (Fluke meter with HV probe) with the amp loaded up.  I get 3200V (!).   The plate current is 50mA.

[W3RSW]

Wow, 3200 volts.   No wonder it's so 'efficient.' 
Glad you found the problem.

So two 813's might suck down your supply to 2800?   Graphite anodes with a fan and you'd stay cool and get a KW easy.  I think the graphite anode 813s are a lot more rugged; but then you run the risk of killing the filament by drawing too much current if you couldn't see the plate turn red     This calls for instrument flying... 

I wonder what voltage the LC ratios in the Pi network were 'designed' for.?   Sounds like it was always intended to by 3000 or so.

Hey, keep it up with the red plates, and you'll put out the big byrd watts, ca'mon.

[K1JJ]

While fiddling with the exciter I looked down to see the wattmeter at 350 Watts.  350 Watts!  Cool! That is a lot of carrier.  Oh, wait.  There is only a single 813 in there.

Rich

W7SOE

Rich,

Is 350w the maximum output?   If so, then that is your pep and need to  set the AM carrier at about 1/5 of that to have headroom for audio. This means a 70 watt carrier as it now stands.

Once you get the L/C ratio adjusted right, you shud see little to no color on the plates. I've been told that 813 graphite plates shud show no color, unlike the metal plates.  For linear use, consider adding a chimney and air to increase this rating.

BTW, normally as a rule of thumb, take the tube's dissipation and divide by 2 to get the AM carrier maximum.  So, 125 w dissipation /2 =62.5 watt carrier, about where you are now.  With 3kv, tuned right and the tank L/C ratio adjusted right, a single 813 shud do about 90w of carrier at the MAX limit. (450w pep) The plate color is when you are at the limit. Air will help a lot. The 813 is an especially rugged tube compared to others.

In AM, class C plate modulated pissbeat service, that same 813 could do 500w carrier at 2kw pep... (not to be confused with linear service)

Good job getting it going!  Those roller inductor intermitants have plagued hams forever.

Later-

Tom, K1JJ

[W7SOE]

I have always looked for the 813's with the graphite plates, just bought a couple.  After being able to "modulate" the glow in the steel version I am starting to see the benefit of seeing that glow.

I have thought that I would add a second 813 as I think it would be as easy as fitting it in there and making sure the filament transformer was big enough.  Since the tubes would lay horizontally one on top of the other, maybe I would leave the steel plate version on top as the glow indicator!


Boy the dip when tuning the roller inductor in the final is very easy to miss.  No I know why the builder actually used pen to mark the roller for each of the bands!   (You can see the innards easily, it is in an old Collins flip top cabinet.

Oh, one more question.  In tuning the antenna load (80m) I get the most output when it is all the way to the stop, maximum capacitance.  Does this mean I should add a parallel cap to extend its range?

Thanks again to all for the help, gotta love this forum!
Cheers

Rich

[W7SOE]

Tom,
     Thanks for the insight, I am over driving it then with my 150W carrier.  Time to study the spec eh?

I don't think a chimney is feasible given the layout but there is a boxer fan blowing directly on the 813.

Rich

[Opcom]

It should stand the 3KV level OK.

As a plate modulated amp/carrier conditions:
CCS Eb=1500V
ICAS Eb=2000V

One tube should be able to produce 360W PEP at 3200V. Can the 813 be reliable running color? My elmer told me never run color on a graphite plate. I've never gotten more than 40W carrier per each as an AM linear amp without some redness creeping in. (a pair of 813's subbed into an HT-41 due to no more 7094's, where they ran about 150-200mA @ 1800V with carrier condition)

[Ed-VA3ES]

Wow, 3200 volts.   No wonder it's so 'efficient.' 

       lol!   

[W7SOE]

I tried reducing the input power to see when the glow in the plate went away, never did.

I still see a dull glow at 25mA on the plate with no input.

Rich

[ka3zlr]

Wow, 3200 volts.   No wonder it's so 'efficient.' 

       lol!   

  Crack me up....


Glad it's only a little Cleaning and service for this one....Easy Fix...just remember One hand in Pocket always around the HV supply...be Careful over there...

[W7SOE]

Maybe I need to get another 813 in there to get the plate voltage down.   

[w4bfs]

I think you might want to check your plate current metering just to make sure it is accurate ...plate color at 80 W dissipation dooesnt sound quite right ... if I recollect rightly, the RCA xmtr toob manual does not spec color at max (125 W) dissipation ... 73 ...John

[W7SOE]

John,
     You are right, the meter is not reading correctly.

Thanks

Rich