6m QRO Experts - Help!

[Dale KD7UO]

This is directed at you QRO experts that have built high power tube amps on 6m. I need some help with a 6m amp conversion I have been performing over the past couple of months. A friend gave me an old Henry 2K Classic 80-10m HF amp. This is a pair of 3-500Z triodes in grounded grid. I stripped the RF deck of all the 80-10m stuff and rebuilt it pretty much using the Heath SB220 6m conversion design that is well published on the web. The problem I think I am having is getting the pi-network output to resonate correctly. I cut down the Tune air variable cap so that I'm measuring 12-55pF on my LC meter. The coil is 3T of 1/4” copper tube, 1-1/4” ID, about 3” long. The load C is one of the two sections of the original HF amp, probably a few hundred pF at max mesh. The input network to the amp is a T-match per the SB220 conversion design. The rest of the amp is pretty much the original Henry amp, including the same cathode choke, plate choke, plate doorknob caps, parasitic suppressors, etc. The tubes were reportedly working fine when the amp was taken out of HF service, pumping out an easy kW on 20m.

I'm using a variac so that I can keep the HV fairly low as I commission. I've been running 1000-1500V on the plates and using about 10W drive. I've got a perfect 1:1 SWR from my exciter to the amp. Idle current when keyed without a signal is right on spec. Filament voltage is spot-on.  The amp appears stable. Here's my problem – I cannot seem to find the sweet spot for the pi network. I'm pulling 300ma on the plate at 1400V, 80ma grid current, putting in 10W and getting out...20W. This is at the best tuning, which is close to min tune capacitance and about 50% load capacitance. 1400V and 300mA is 420W input and yet I'm getting only 20W out, less than 5% efficiency. When I try to drive it harder the grid and plate current jump up but there is no increase of output power. What I want to know is, where is all the power going? It sure isn't coming out as RF. I've spent 2 weeks fiddling around with the output coil and now have a bench littered with all sorts of pieces of copper tubing. I've added turns, reduced turns, stretched the coil, compressed the coil – nothing gets any better and most trials make it worse. I've done the classic back-driving the amp (without power of course) using an MFJ antenna analyzer and a non-inductive resistor from plate to ground to simulate the plate resistance. I can usually find a spot where the SWR goes to 1:1 indicating resonance. The tuning is very sharp and always very close to minimum tune capacitance. But as soon as I put it all back together and power it up, I'm still at extremely low power out.

I wondered if the original Henry parasitic suppressors might be acting like RF chokes at 6m, so I removed them completely. No improvement at all. One thing I'm wondering about is the giant plate RF choke; it's still the original 6” long close-wound unit of the 80-10M design. I know at 6m I don't need nearly as big a choke, but I haven't bothered messing with it to date. Is it important? Should I wind a much smaller one? Can you have too much plate choke?

I had the same problem a couple of years ago when I built a completely homebrew 6m amp using a pair of Russian GI7B triodes. The amp worked but I never could get the efficiency anywhere near what it was supposed to be. I rebuilt it twice trying different pi output techniques and finally gave up on it. I was really hoping this 3-500 amp conversion would work out, but I appear cursed by the same inability to get any sort of decent efficiency. I really believe I'm just overlooking some basic principle but I sure haven't stumbled across it yet. There seem to be plenty of people running a pair of 3-500Z on 6m with an easy kW but I'm sure not finding it easy to get there. Any suggestions?

Thanks

dale kd7uo

[Brian Moran]

Here's how I determine whether the pi-network is going to work:

First, I calculate what the plate resistance is going to be at typical values I'm going to run the amp, using this formula for Class AB:

RP = EP / 1.8 X IP 

So if you're going to run the tubes at say 3300v, 550ma, thats' about 3.3K of plate resistance

To see if my network is going to work, I "run it in reverse" -- the output network normally transforms the plate impedance (3300 ohms) to the antenna impedance (50 ohms). It works the other way too. I UNPLUG the AMPLIFIER. I check that the AMPLIFIER IS UNPLUGGED. I short circuit the B+ to ground to make sure B+ is at ground. Then I remove the tubes. I take a resistor of the proper value (3.3K in this instance), and connect that between the plate cap and ground. That represents the plate resistance.

I hook up an MFJ-259B to the OUTPUT of the amplifier (the amplifier is unplugged, right?). I trip the output relay (usually they're open frame, and can just be depressed with a screwdriver), so the MFJ is feeding into the PI-network. Setting the MFJ-259b to the desired frequency, I then adjust plate and load capacitors (and any fixed coils if it's that far off) for 1:1 SWR; 1:1 a this point means that your plate resistance is being transformed to the 50 ohm antenna resistance at the frequency of interest.

-Brian N9ADG

--
~The Voice of the Pacific NorthWest VHF Society~
You are subscribed to the Google Groups "PNWVHFS" group.
To post to this group, send email to PNW...@googlegroups.com
To read message history, visit http://groups.google.com/group/PNWVHFS
---
You received this message because you are subscribed to the Google Groups "PNWVHFS" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pnwvhfs+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Tree]

Dale - on trick I have done is to remove the connection to the tubes - and replace it with a resistor.  For a pair of 3-500Zs with 4K on them - the resistor should be about 5K ohms.  Then play with the values of your network until you see a nice SWR looking at the output connector of your amp.  This should be pretty close to what you need.

That's how I got mine figured out. 

Tree N6TR

--

[VE3AX/7]

Dale:

 

I feel your pain!

 

Years ago I did up an SB-220 for 6m and for 3 weeks I had nothing but a huge oscillator around 120 Mhz!

 

Couple of recommendations....

 

I think your plate tune cap is still too high at minimum.  I used to shoot for 3 pf at the low end and 30 pf or so at the high end.  That will require a different cap – wide spacing – 3 or 4 plates max.  You can remove every second plate from a surplus cap sometimes to get that.

 

I moved the original plate tube cap and used it as the load cap.

 

I did not change the Plate Choke and did not have issues with the original.

 

But, the parasitics were a huge problem.  A friend sent me an article about converting (I believe) an SB-200 and the author used 1/4” wide silver plated brass strips about 3” long (I’ll have to look up the actual dimensions) which are then formed to make a hairpin loop about the 3 resistors.  I’ll see if I can find the article I wrote up on the conversion as the details are in there, including pictures.  Once the new ones went in, my “oscillator” became an amplifier.  I used it for years.  Even though you say the amp is stable I would change to this design. Prior to the change, a grid dip oscillator near the tank circuit showed a huge “suck the life out of the GDO” dip around 120 Mhz.  After installing the new chokes, the GDO would show a very shallow broad dip up around 150 Mhz.

 

Your 3 turn coil sounds about right.

 

I also found the original T/R relay SWR was too high at 6m and pulled it out, replacing it with a DowKey relay mounted on the back of the amp.

 

Best of luck.

 

73

 

Peter

VE3AX/7

--

[Terry W7AMI]

Brian,

Wouldn't it be better to leave the tubes in place?   That way the inter-electode capacitances of the tube would also be present when  tuning the output stage.

Terry W7AMI

[Brian Moran]

Maybe, but I've not done it that way before. Might be interesting to compare doing it both ways.

[Terry]

Stephen, 

  According to the data sheet a pair of 3-500Z tubes should have approximately 10 pF of inter-electrode capacitance on the output.    I think you could get closer to the correct match with the tubes in place but in practice the difference at 50 MHz is small.   Besides,  the purpose of the experiment is to get close to the correct values in the output matching circuit,  not to be perfect.

Terry

[Tree]

I agree with Terry's answer on this.  If you find that you are getting a good SWR - but you are near minimum capacitance with your plate capacitor - then you need to work on changing things.  You want to have plenty of adjustment available to deal with different loading situations. 

Tree N6TR