Quick help!

[N4LTA]

I have the dual 807 transmitter about ready but I have one problem.

I need some help.   I can only drive it to 130 MA (both tubes and I would suspect that 100 mA per tube would be about the correct drive. I am running lower voltage than some  - 363 volts at 130 mA at about 1 watt drive. When I increase the drive - the plate current and power out actually drops and that has me confused.

Any ideas - The turns ratio at the grid tank is 15:1 for 50 ohms in.  Too high or low.

I had not idea what it should be so I took the plate bias voltage  -80 volts and divided by the grid current for two tubes 7 mA and came up with 11,500 ohms and divided it by 50  for an impedance ratio of 228 and then took the square root to get the turns ratio of 15. I have no idea if this is the proper methode and I did not take the tube capacitance into account.

I am getting a little better than 70 % efficiency so the plate circuit seems OK??

Any help would be appreciated


Pat
N4LTA

[k4kyv]

I never could get a pair of 807s to load up to full plate current as rated in the tube charts years ago when I used them in a transmitter, even with 750 volts on the plate.  So I just settled for about 100 watts input.  But the pair in the RF driver of my Gates BC1-T easily load up to 225 ma with only about 650 volts on the plates.  Not sure what the magic is, but it  has been years since I have used 807s in a homebrew transmitter.

[KC2ZFA]

from your previous description of the grid tank it looks like
you're grounding the bottom of the LC circuit and feeding the
top to the grids through a coupling cap. Also, the bias/leak
appears to be right at the grids (after this coupling cap).

Why not feed the bias/leak to the bottom of the LC and then feed
the top to the grids without a coupling cap ?

In the just-above-described circuit the input impedance seen by the
grid is (Pin*622*10³)/(grid current in mA)².

Peter

[N4LTA]

The grid tank is a coil and variable cap in parallel which is AC grounded at the bottom with a capacitor - I did this so I could neutralize if necessary. The 50 ohm input is fed with a 4 turn link. The top of the tank is coupled with a 100 pF cap

The bias is fed to the grid via a 2.5 mH choke - both grids have a 47 ohm carbon resistor from the grid to the choke to serve as grid stoppers. The 100pF cap, both 47 ohm resistor and the choke tie together at one point.

Plate current rises as grid drive rises up to 130 mA and the as the drive is raised - th eplate current drops which is a puzzle (to me )


I agree - The cap can be removed and maybe the bias shoul dbe injected at the botom of the tank?


Pat
N4LTA

[KC2ZFA]

I agree - The cap can be removed and maybe the bias shoul dbe injected at the botom of the tank?

that would turn it into a standard circuit.

[The Slab Bacon]

Pat,
      First, make sure your screen voltage is sufficient and not sagging as you up the drive.

If you increase the drive and the outpoot falls off, check your grid leak resistor.
It could be possible that you are rectifying the RF inpoot and creating excessive biass. You will need to adjust your leaky grid resistor accordingly.
With leaky grid biass some of the inpoot RF is rectified into DC and creates the biass.

Or you've just hit the point where the tubes just won't take any more grid drive, and your low plate voltage is holding you back from making any more outpoot. 360v on the plates of class C 807s isn't very much. Those things really come to life around 700v.

[N4LTA]

Thanks,

The screen is taken from the plate voltage with a resistor. I'll check and see what is going on. That could be a problem.

I am wondering if increase drive is making to much bias voltage after your comments. I did not expect to much performance at 350-400 volts and I am getting over 70% efficiency and its nice and stable so maybe I should be happy as is. I'd like to find out why I can't drive the tubes to 100 mA though.

I put a 10K  2 watt resistor across the grid tank and now it is tunable but still sharp - sharp is not a bad thing I guess.

Pat
N4LTA

[WQ9E]

I expect the low plate voltage is the major reason for low current draw.
 
My Johnson Navigator is similarly handicapped by low plate voltage running 340 volts loaded and the screen is regulated at 150 volts.  I had to select from several new 6146 tubes to find one that would draw the rated 115 mils when tuned and loaded.

For the 807, the tube tables recommend 250 volts on the screen with 400 plate volts in class C.

[KM1H]

I need some help.   I can only drive it to 130 MA (both tubes and I would suspect that 100 mA per tube would be about the correct drive. I am running lower voltage than some  - 363 volts at 130 mA at about 1 watt drive. When I increase the drive - the plate current and power out actually drops and that has me confused.

That sounds perfectly normal to me Pat and according to tube manuls you are only 30ma max off from expected.
If you cant reach 6-7ma total grid current then either drive losses are high, the 47 Ohm resistors are excessive (Ive found 10-15 Ohms more than sufficient unless there is drive to burn), tubes are a bit soft, filament voltage is low under full load, or screen voltage is below 250V at full load.

Also dont trust anything digital when measuring DC or 60Hz AC voltages in a RF circuit.  My trusty Tripplet 630 is the go to device here.

[N4LTA]

Yes Carl I have had problems with measuring grid bias voltage with a Fluke

The 47 ohm resistors seem a little high - I got that value from an old Ham Notes article. I have a Tripplet 630 that I'll have to look for.

The only thing that puzzled me was that I could double the drive to 2 watts and the plate mA dropped.

The screen may be sagging - I'll check tonight - Right now this thing has no meters except a plate meter on wires and it takes 5 arms to measure everything all at once and not get zapped reaching over the tubes.

You have given me some stuff to look for and I sure appreiate your and everyones help. Thanks again

Pat
N4LTA

[WQ9E]

Pat,

With low plate voltage and high excitation you can easily get into excessive screen current so run some calcs on your screen dropping resistor before you change it.