Use a tuned input or not

[W1RKW]

I'm trying to determine if I need to use a tuned input to the PA on my 813 rig.  My thinking is (and I don't know if I'm right or wrong) if I use an exciter that has a tuned output would I need a tuned input at the PA?  If I should use a tuned input what's a good, effective and simple configuration to use?   I've seen several configurations and they seem complex and take up alot of real-estate and my setup is tight and I'm hoping I can get away with not having to use one if  the exciter has a tuned output network.

BDubLU

[W3SLK]

Bob,
    Although my HN-500 originally had a tuned input. It was designed to be driven with a Ranger. When I replaced the tank circuit with a B&W 850-160 coil set, I attempted to make a coil that would oscillate the grid on 160. Unfortunately, I was met with miserable results. So I said the hell with it and drove the grid with brute force capacitance. No problems on 160 but I have yet to try it on anyother band. Of course I don't really care to get on 75m.

[W1RKW]

Hi Mike,
Thanks for the tip.  I think I will just drive the 813 rig with my Viking II and see what happens.  I got to thinking I can build a tuning network for the input of the 813 rig on a separate chassis if necessary and just route the RF from there to the PA.  Hopefully I can tune it bruteforce with the network of the Viking II. 

[flintstone mop]

A lazy way to do this is to use an antenna tuner between your exciter and the input of 813's.
Fred

[K1JJ]

Hi Bob,

If it's a linear amp, GG, you definately want a tuned L/C circuit right at the cathode input.  Within a few inches of the tube pin.

If it's a class C amplifier, I would also recommend an L/C circuit there too. It will be easier to drive, even if you swamp it later with a non-inductive resistor for more stability.
I always use tuned inputs right at the grid pin for all rigs here.  You should also neutralize it, and feeding the negative feedback thru the bottom of the L/C circuit back into the grid works FB. Refer to the 813's X 813's rig on HUZ's page.

You can get away without a tuned input with both amplifiers types, but, it's so easy to do, why not do it right....

73,

T

[W1RKW]

Fred,
Antenna tuner, excellent idea.  Never thought of that.  It's worth a go.


Tom,
This is for the 813x813 rig I've been working on for eons.  So yes, it would be for a class C amp.  I've thought about it for awhile and I'm going to do a tuned input to the amp. Might as well do it right the first time. I don't have enough room on the amp chassis so I'm going to put it on a separate chassis and feed the amp from that.  The chassis will mount on the rear side the rack since the rack is deep enough.  I didn't want to build another chassis but so be it.  I was looking at the 813 rig schematic on HUZ's website.  Where do you pick up and feed for the negative feedback in this schematic? Is NFB part of the neutralization circuit? Also, how critical is the configuration of tuning network.  Would a pi type network work?  I'm looking for simply yet effective.  I'm figuring once this is set not much tweaking will be necessary unless operating frequency changes, one of the reasons why I want to mount this in the rear.  Thanks.
Bob

[K3ZS]

I found out a few years ago when I bought my first solid state ricebox, that you need some sort of tuning to get full drive.  I have an old Gonset GSB-101, grounded griid linear.  It has some sort of loading coil on  the input, but no bandswitched input tuning.  The input impedance of four 811's driving the cathodes are easily within range of an older tube rig with pi-net outpur tuning.  The ricebox was lowering the power due to the mismatch, and the linear wasn't doing much in that configuration.  I resorted to using a low power tuner between the ricebox and the linear.  In my case, I also have a tuned antenna, so to simplifiy it I got one of the inexpensive autotuners to use when driving the llinear.  If you are using a ricebox on AM you want to drive this type of linear to a maximum of 200W carrier on the output.  In this case the automatic power reduction of the ricebox may be about right without using a tuner between the rig and amp.

[K1JJ]

  Where do you pick up and feed for the negative feedback in this schematic? Is NFB part of the neutralization circuit? Also, how critical is the configuration of tuning network.  Would a pi type network work?  I'm looking for simply yet effective.  I'm figuring once this is set not much tweaking will be necessary unless operating frequency changes, one of the reasons why I want to mount this in the rear.  Thanks.
Bob

Bob,

The neutralization, which is a form of NFB, but is used here to cancel the G-P feedthru capacitance, is derived via a  1X1" plate facing the tube plate. It can be mounted on a standoff and put about 1/2" away from the tube glass, looking at the plate. This forms a capacitor.

The input tuning network is not critical, but needs to be accessible if you QSY more than 50kc or so on 75M.

A reverse pi network would work, but you'd need to find a way to introduce the neut signal back in, just like the parallel LC circuit does at it's bottom in that circuit.  I'd stick with that proven circuit.

T

[W1RKW]

Hi Tom,
Thanks for the advice. OK on the circuit.  I'll give it a go.  Actually, after lookiing at it for a while it wouldn't be to hard to build than any other matching network. 

For the coil, what is the diameter of a miniductor or that particular coil that you used, and turns per inch?  I don't have any miniductor coil stock so I'll have to make a coil.

Also, where do you find solid teflon wire? What wire size did you use?  Home Depot sell anything like that?

[K1JJ]

For the coil, what is the diameter of a miniductor or that particular coil that you used, and turns per inch?  I don't have any miniductor coil stock so I'll have to make a coil.

Also, where do you find solid teflon wire? What wire size did you use?  Home Depot sell anything like that?

Bob,

The coil is not critical at all... just put anything in there that will resonate the capacitor when it's at about 100pf for 3.8 mhz.   You could even test the L/C circuit by putting a generator across it and peaking for 3.8 mhz on the scope.   The Q of the coil can be anything that works out.

Teflon wire is not sold at Home Depot as far as I know. You can get away with regular insulated wire for the 50 ohm link since there is not much over 500V there anyway.  Worse case, put a spaghetti sleeve over it or whatever.  The wire size can be #22 or whatever. Adjust the turns (5 -6 turns is usually right) for a good <1.5:1 swr input.   Unless you swamp it with a resistor, the swr will not be accurate until you draw normal grid current in the operating tube to load the L/C circuit..

T

[W1RKW]

Tom,

OK on the coil dimensioning. Thanks. 

Would I really need to worry about the SWR?  If I peak for max output I would assume that I have a good match between the two stages, yes/no?

Bob

[K1JJ]

Bob,

You want a decent match, like 50 ohms into the rig.  This means maybe 5-6 turns for the link coil. This way it's universal for even a solid state driver later on. Do it right.

You can simulate the tube grid load with the rig off. Put a  resistor across the grid pin to ground that equals the E/I of the grid when running. Look at the operating tube specs for class C.  Then use the MFJ-259B into the SO-239 jack [with the T/R relayed keyed, of course] and you will be able to see the resonance peak and the swr reading.

You can cold neutralize it by using a scope probe on the tube's plate and feeding in a little RF drive to the grid. Tune the neut cap for min feedthru.  No voltages on the tube needed for any of this stuff.

BTW, you can use regular HV wire for the link turns for voltage isolation.

Use ANY miniductor that will fit and be convenient for tapping its turns.  5-10W is all that you will deliver thru it.

T

[W1RKW]

Hey Tom,
Sounds good. I was thinking about what you said in your last post while driving home from work.  So, I get a good match, I'm assuming now once the match is made and SWR is good, I will still need to worry about grid current when driving the 813's.  So in an effort to not exceed the grid current rating when driving the 813's I can simply adjust the drive level to within the proper level?  I'm sure I may have to tweak the resonant points again but I should be dialed in pretty much at that point.

[K1JJ]

Yep, you have it correct, Bob.

The input tank and link will be in the ballpark after the intial cold MFJ 259 setting. When you fire it up, you may need to add a turn or subtract one  to get the 50 ohm link swr right using an swr meter between the driver and 813 rig.  You might even find you need to tap the mini-ductor main grid coil in or out to get the tuning cap to be messed in the middle or so.

Yes, once everything is running - fixed bias, the grid leak resistor is working right and the rig is running normally, you then adjust the final tweaking of the grid current by the amount of drive into the grid. It will go from zero to more than you need just by adjusting the drive.

Also, there will be a slight interaction between the neut adjustment and the grid tuning cap. Once the neut is set. Just rock back and forth until you get a compromise for zero feedthru and good grid peaking at the same time.

I usually bring the neut cap shaft out to the front panel to allow minor tweaking on the higher bands. But then again, many guys don't even bother to neutralize tetrodes, so you will not have any problems if you set it and forget it on 75M or 40M.

T

[W1RKW]

Tom,
Thanks for your input on this.  This has been a very slow project but I am making progress. Like you said, do it right, which is what I'd want to do. But I'm just getting itching to get it finished so I can move on with something else.  I can't wait to get the rig on the air.  I'm hoping with the detail that I've put into it that it works good. I've got a homebrewed solid state audio amp to drive the modulator that exhibits excellent frequency response but the acid test will be in the RF world.  I hope that audio amp will cut the mustard when exposed to RF.  And if it does I should have some good hifi AM with plenty of RF BA's.  Just need a good antenna, but that's a whole 'nother story.

[K1JJ]

Yep, I'll look forward to hearing it.

Coincidently, I'm finishing up my single "blown"  813 X 813's now too.  160-20M bandswitched. It's one of six big rigs in the works, so it's getting done at 1/6th the normal pace...

I hope that rig will be the main "get on the air quick" AM rig, with a nominal 300W or so. For the audio driver I plan to try the Fet driver board designed by Tom/KLR and the guys. Haven't built it yet, but will do so once everything is finshed.

See ya and good luck with getting the bugs out.

T

[WA1GFZ]

Tom,
I did some simulation on Tom's modification of my circuit and that is the way to go.
I think the gain will change with drive level without using the op amps and feedback circuit. My circuit needed help. I started thinking the drain Z would be stable but in reality the tube is acting like a variable resistor in the drain circuit. The op amp with feedback sets the tube cathode voltage based on the modulation drive not effected by the resistance of the tube. Good Job Tom KLR it would have sounded weird.