Driving two 1625s

[N4LTA]

I am testing my dual 1625 AM transmitter and it appears that I don't have quite enough drive.

The buffer driver tube is a 5763 which I though would have been plenty. The 5763 has about 280 volts on the plate and the output is a LC circuit tuned to resonance coupled with a 1000pf on each side of the resonant circuit. I get a nice peak at each of the three switch positions - 40 -80-160. Nothing is unstable that I can see.

The grids are tied together and the resonant circuit feed the grids directly with the 1000 pF capacitor.

I am feeding regulated -22 volts of safety bias through a resistor and 2.5 mH choke in series with the choke tied to the grids. The resistor is an 8.2 K 1 watt - I picked it to allow for about -90 volts at 8mA grid current which is what the data sheet asks for for Class C phone at approx 600 volts.

(-22) +  8200 x .008  = -87.6  - grid voltage at 8 ma of grid current.

I realize that I can stick a 12 K resistor in there and do away with the safety bias but I build this for a test bed.

The screen has regulated 250 volts.

I can't get any more than 6 ma or so grid current

Any ideas on getting more drive or anything to look at?  Last night I got about 80 watts carrier out on 40 but less on 80 which is a puzzle.  That works out to abou 66% efficient - I expected a little better. I'll get into it more tonight.



Pat
N4LTA

[N2DTS]

It should work, I think they used a 5763 to drive a pair of 6146's in lots of rigs, with drive to spare.

Measure the plate current on the 5763, see what its doing.
Screen voltage on the 5763 might have an impact on the grid drive.

You might try changing the grid circuit Q, maybe try a bit more or less coil there.
If the Q is too high, losses in the coil go up and tuning is touchy.

The more grid leak you use, the more drive you need to get the same current.
Since I drive all my rigs using the flex, or the 32V3, I have lots of extra drive power to waste going deep into class C.

I dont have books with me, but what is the output power of a 5763, and what are the drive needs of a pair of 1625's?

Gito would tell you to put in an 813 as a driver, as you need 3x the power to drive the 1625's.

Brett
 

[WA1GFZ]

Check your fixed bias voltage when drive is applied to see if it is going more negative pushing you further into class C away from grid current. 5763 should be plenty of power. Look at the RF voltage swing on the grid with a scope.

[N4LTA]

Thanks Brett and WA1GFZ.

The 5763 can do an easy 2 watts in doubler service an dthe drive requirements are .2 watts per tube - so I thought I had Gito covered.

I need to give it a good look over - but I would have thought that I woul dhave enough drive to overdrive it pretty hard.

It should be good for AM though as it ran fairly cool doing 80 watts out on 40 meters for 1/2 hour - Fine tuning is what I need.

One funny thing happened last night. I was bending some wires in the tank circuit and I heard a funny kind of whoosh sound - not loud at all. Everything looked OK and I powered it up and only could get about 40 watts out - nothing looked out of place until I noticed a 1625 was not glowing. It had a tiny pinhole at the plate seal and all the power got out of the tube.

Pat
N4LTA


Pat
N4LTA

[Gito]

Hi

Bret, two 1625( 807 with 12 v filament) needs only 0.4 w drive, and it needs only 1.2 watt as the books says 3 to 1 drive power (not my personal opinion)

And I believed like Pat wrote 5763 can easily give an output of 2 watt.
So the conclusion do I need a 813 to drive two 1625?(as You wrote)
of course not since 5763 can give more the drive needed for two 1625.( it can give 10 watt output in class C)

So Bret before You wrote please check the tube data sheet handbook.


Gito

[KE6DF]

To go along with your 813 driver for your pair of 1625s, consider using a pair of 833s as modulators -- that way  you will be sure of having enough RF drive and audio. 

[Gito]

Hi

What I wrote is an answer to Bret,when He wrote .... Gito  would tell you to put a 813.....as a driver....
So don't miss understand what I wrote before.

Gito

[N2DTS]

Gito,
Just a poor attempt at humor....

Yes, a pair of 833's, blow those 1625's into orbit!

Brett

[WU2D]

Actually a 1626 is the proper driver for a 1625 - Ha

[Gito]

Hi

Sorry Brett,If I have taken your reply seriously ,any way  this AMfone net is a good way to change knowledge  from one to the other.as I think I still have much to learn.
The first transmitter I build is 40 years ago using only one 807 and a 6AG7 as the oscillator/driver with two EL 34 in push pull as the Modulator.

And Not even knowing what an "good" antenna is ,at that time I used a random wire as the antennas.
Later from friends an the Old ARRL handbook I begin to know how actually a transmitter works,what is a "good" Antenna ,what a feed line is

Thanks Brett

Gito

[N4LTA]

Found the problem - Duhhhh

I had the crystal oscillator hooked to the regulated power supply. It (and the screen) drew more current that the supply could handle - so I removed it and tied it to the doubler/driver supply. That supply was generated with a dropping resistor and filter cap. I didn't look at how much the voltage dropped.

I doubled the current through the dropping resistor - duh - and guess what - my driver plate voltage dropped to 150 volts.

I now have plenty of drive and the bias is about exactly -90 volts with 8 mA of screen current.

Now - On to the Pi-net and why 40 works great and 80 doesn't.

I expect another Duh.


Pat
N4LTA

[N4LTA]

Learned one thing about toroid inductors last night - The #6 material is excellent as a pi net inductor   - I have used it before and have used the #2 material with good results.

I tried the #3 Blue core for a 160 meter section of the coil and that doesn't work too well. Got hot and the efficiency of the network dropped way down. The #3 material is rated up to 5 mhz but didn't work well on 3.7 and 1.8 in the pi net.

Back to rewinding with #6 material. I am using a stacked pair of T130s - I was using a stacked pair of three T130-3s.

Pat
N4LTA

[N2DTS]

Is there any downside to using toroids instead of air coils?

Besides the size factor, it does not seem like you could do much better than an air wound copper tubing coil.

Brett
 

[N4LTA]

The toroid are much smaller and they tend to keep the magnetic field enclosed so that they can be mounted closer to the chassis and other coils. The downside is that the cores can be lossy. There is always a downside - everything is a compromise.

At this power level and frequency - A coil wound with #12 wire would work fine but be larger and require more space.

I'll post a photo when I get things like I want them.

Pat
N4LTA

[KM1H]

Toroids have been use in 1-2kw amps on 160 & 80 for decades Pat and with equal efficiency as air coils. Its often the only way to get those low bands in a confined space and small cabinet.

I prefer the 2 mix at HF and have run 6 mix on 6M at up to 150W OK in amplifier input pi networks.

Carl
KM1H

[N4LTA]

Yes - I had a 160 meter amp using a GG   4-1000A and the tank was a T400-2  about 15 years ago.

I don't have any suitable  type 2 toroid material so I am using the 6 material. I just got finished testing 40 and 80 and I have that working good.  Now to add the 160 meter toroid and get the proper padding caps switched in.

I hope to get the modulator working before the weekend is over.

Pat
n4LTA