Grounded Grid Amp feeding problem

[Ed/KB1HYS]

I have a homebrew dual 4-400 amp in grounded grid mode.  I'm feeding the cathodes through capacitors in the usual way, there are no tuned circuits, just coax from the rear panel to the caps to the cathodes/filaments.  The amp seems to work fine, BUT (why is there always a BUT?)  The rigs feeding it seem to be getting some kind of feedback. 
I generally tune the rigs into a dummy load (home made "cantenna") and then plug into the amp. Seems that the plate current of the rigs gets kinda squirrelly when feeding the amp.  No problems on the dummy load or antenna with out the amp. 

The RF deck in the AMP appears to be ok, like wise the components underneath. Just flummoxed about the drivers plate current.  This happens both with the TS-520 and the Valiant.

Could the Coupling Caps be getting overheated?  I used several large ceramic disks in parallel.

[WA1GFZ]

The reflected impedance of the amp back to the exciter changes over the cycle. When the drive swings positive the tube is cut off reflecting a high Z back at the exciter. When the drive waveform swings negative the reflected z back to the driver is low since the final tube is turning on. The exciter is not happy with the changing reflected Z. You could hang a resistor across the input of the linear if you have extra drive to burn. Without a tuned input you don't have the flywheel action of the input network integrating this swing in input Z.

[KM1H]

Whats the input VSWR?

Have you tried bonding the chassis together and then to a good ground?

What are you using for a filament choke and is it adequately bypassed?

How are you feedng the bias and what is the idle current?

[KM1H]

That shouldnt make much if any difference to a tube exciter with a pi network Frank. OTOH its a major factor with SS.

[WA1GFZ]

Thur if the distance between exciter and amp is short.

[W2PFY]

Are you using the same coax from both drivers to the amp? Change out that coax if you are using the same one to test this amplifier. I had a problem just like you describe and again, it turned out to be a bad connection on one end of the coax.

[kb3ouk]

i have a similar problem with my quad 572B amp. input circuit is supposed to be tuned for 50 ohms, but it has a high SWR on it. the input circuit looks like some arrangement of caps and resistors feeding the filaments. on 75 it is up to like 5:1 to 10:1, slightly less on 40, then on 20 its only about 2:1 or 3:1. and i've discovered that the output tuning of the amp affects the SWR on the input too.

[Detroit47]

Why dont you just make a tuned input circut? The driver will be happy and the output cleaner. Just my .02$

John N8QPC

[W2VW]

Retune the driver for rated plate current or build a tuned input.

[Steve - K4HX]

If all else is good, the lack of a tuned input should not matter if you are using a tube rig as a driver. I've been doing that for about 15 years with no problems.

[Ed/KB1HYS]

If all else is good, the lack of a tuned input should not matter if you are using a tube rig as a driver. I've been doing that for about 15 years with no problems.

That is what I thought.  I will try retuning into the amp for best plate current.  Could be the impedance presented by the amp is varying more than I thought it would. 

[Steve - K4HX]

I would check the things Carl posted.

Yes, the input impedance will vary but not so much that the flywheel effect of tuned circuit of your driver won't handle it.

[K1JJ]

Ed,

Hopefully the suggestions already made will fix the drive instability problem.


As for using an input tuned circuit, I suppose it depends on how particular we wish to be.  You can run the amp either way and  probably nobody will notice the difference. But I think there is something lost when we have the tuned circuit in the form of a pi-network located 10' away through a coaxial cable - compared to an L/C right at the cathode pin of the amplifier.  

I have run tests (using an FT-102 tube driver rig) with the proceeding linear amp tuned circuit directly connected to the cathode -  and  then tried the tuned circuit disconnected.  Even with a Q as low as 2 to 3, there is a noticable difference in ease of drive with the cathode circuit connected and peaked. I also believe there is a difference in better IMD with the circuit as conventional info suggests.

That said, sure, no input circuit will work "OK" but why be lazy? Add it anyway.     The biggest point may be that  4-400A's in GG are not as clean as 3-500Z's in GG.  It may be in the order of a 5 db degrade in IMD numbers.  Maybe the tuned circuit will help you recover say, 3 db of this figure in better IMD. Hard to say without careful meaurements.

BTW, I have several homebrew amplifers here and they all use tuned input circuits.

Good luck with the project, OM!


T

[ke7trp]

Just a thought.

Recently, I have seen a bunch of hams use the insides of the cheap Vectronics antenna tuner. This is a small antenna tuner for low power and mobile use. It has nice range. You can remove the insides, bolt it right into the face of your amp and wire it in. Then you have a tuned input right to the tubes.

These cost next to nothing used and about $100 new.
C

[DMOD]

What Clark said.

The Vectronics VC-300  makes a good low cost input matching circuit and you get a power meter circuit as well.

From personal experience, I would not run more than 80-100W carrier through the VC-300 series. 

They're great little tuners but use a toroid transformer in an auto-transformer arrangement where the windings are close to each other at the center of the toriod, and can flashover and carbon track.

Good luck with the feeding of your GG Amp.

Phil - AC0OB

[KM1H]

The 4-400 is a decent linear tube in grid driven and its about 5-6dB better IMD in GG so I wouldnt lose any sleep over it; the TS-520 will provide the flywheel as well as be much cleaner than most riceboxes. The difference between the input at the tube or 10' away is the C of the coax which may have to be shortened or lengthened. Many old amps had a required cable length in the manual for that reason.

Carl

[W2PFY]

I found this picture of the V-300. Looks like it has a built in crummy load. What is that round thing with wires that looks somewhat like a variac at the top of the picture?

I looked a little closer and it looks like a series taped inductor?

[DMOD]

What is that round thing with wires that looks somewhat like a variac at the top of the picture?

That's the tuning inductor for the C-L-C tuning network:


"...but use a toroid transformer in an auto-transformer arrangement where the windings are close to each other at the center of the toriod, and can flashover and carbon track."

The tuning inductor is basically an RF variac.

Phil - AC0OB

[KA2QFX]

Like most antenna tuners the Vectronics is a C-L-C T-match, which is a high pass filter for stuff you'd like to attenuate. Using a high pass T-match is NOT a good idea for matching the input of an amp.   It WILL match the exciter to the amp at the fundamental, but will NOT provide all the other benefits of an overall better performing amp.

A low pass Pi-Match is preferred to attenuate the reflection of higher order products from the non-linear input Z of the tubes back into the exciter. 

I have to go with Frank on this one.

Mark