120 watt AM Rig

[W9ZSL]

I'm still fishing.  In regard to the 807 amp in the '57 handbook, I can put 600 volts under load on the plates of the parallel 807s (in my case 5933s) at 200 MA just fine.  I gather from previous answers that a slightly-higher voltage on the modulator tubes (also 5933s) would be desirable.  My existing supply will give me at least 650 VDC under load.  Back to the drawing board...sort of.

The ultimate rig will be the 813/811A combo built around a UTC S-22 with two supplies: 2KV RF and 1350 AF under load.  Yeah, I know, I've been wrangling about this for years.  Before I do that "Masterpiece" rig, I have to get up to speed.  Enter the 5933 rig.

Previously I proposed a single supply (775 VDC no load) to do the job but after your feedback, building an RF amp with it's own supply (600 VDC loaded) for AM makes sense and it can give me some experience before I tackle 2 KV.

It also makes sense to use the existing supply for the modulator.  Under load, I'll have around 650 volts.  Now the big question.  I want to use an ART-13 mod iron.  It was designed to use a single 1250 volt supply.  In the past inquiries here, I was told that impedance matching didn't matter much as long as one supply was used regardless of voltage. 

I can't imagine that 600 volts to the RF and 650 to the modulators would give me a lot of mismatch with that iron or would it?  The RF screen voltage would be picked off the unmodulated B+ supply and applied to the ART-13 screen winding.  The modulator supply already has variable screen and bias voltages built in.  The driver will be separate and can be used with both modulators.

I've also learned that under these circumstances, a higher voltage on the modulator is preferred but on the other hand the ART-13 iron is not multi-impedance.  I think the plus or minus 50-100 volts difference primary to secondary might be a slight but not an unacceptable mismatch...or am I wrong?

[KA2DZT]

Just use the same supply for both the RF and modulators.  600 volts or 650 volts on the RF PA will be OK.  Use the two separate filter circuits in your PS.  Is your plate xfmr able to handle about 300 ma load??

[W2JBL]

The 807's will be fine up to 750 volts ( that's maximum ICAS rating). I'm using an ART-13 mod transformer with 807's modulating the 1625's (12 volt filament 807) in an ARC-5 (WW-II aircraft transmitter). The setup works great. I used the screen winding on the ART-13 transformer (with a 22k 2W load resistor across the winding) to feed inverse feedback to the speech amp. The rig sounds excellent- very clean and good frequency response. At 620 volts there's just enough audio power available to modulate 100% negative and 110% positive. Higher voltage would help but in my case using 70 year old transmitters I never went to the high tap on plate transformer for 700V for fear of crapping them out. You only need one power supply for the modulator and PA- using separate supplies would be a waste of energy, money, and parts. Good luck with your project.

[W9ZSL]

Looks like I'm in the ballpark.  I'm not too concerned about audio feedback because the separate speech amp / driver already has it built in.  That will have a pair of 6V6's running about 10 watts which clearly is overkill for a pair of 807 / 5933s.  That's why there is a volume control!  I can switch it to drive the 811As  The driver iron is a Stancor NOS Poly-Pedance universal A-4762 rated at 15 watts.  It will have its own supply.  I'll pull the choke under the chassis and put a Thordarson 200 MA swinger on top alongside the two octal sockets.  Those were originally intended for voltage regulator tubes.  That HV will go to the modulator.  Note the bias and a tap on the 100K bleeder for the Modulator screen.  I can split a second high voltage branch off the bridge and tie it to one of the terminals on the strip.  I'll mount another Thordarson choke (smoothing) on the RF chassis along with the filter cap.  Screen voltage can come off of that and run through the ART-13.  I can build a separate bias supply for the RF.  I may go with just 75 and 40 meters. I have what I think is an Air-Dux 2006T; 2 1/2" D, 6 TPI of #12.

The bridge was originally used in a HB supply for a Hot Water 100 as was the choke. Once the choke is gone, I probably will replace the bridge with something more stout.  

The bias is a full-wave bridge that gives me 5 to 135 volts varied by the pot salvaged from a BC-1T.

[w7fox]

In my earlier note I mentioned my rig that actually has two 1625s in the final and two 807s in the modulator.  I had the tubes and sockets in the junk box so I used what I had.  The art13 transformer is made for 813/811 tubes running the same plate voltage.  I am not a vacuum tube expert, so I struggle with the constant current curves and such, but when I was designing the rig, it was pretty clear that I wouldn't have enough audio power if I ran the modulator tubes in AB1 at 600, or even at 700 volts.  To get enough power in AB1, I would have to run the modulator at 750 volts, and the art13 trnsformer has the wrong ratio for that.  The ARRL handbook talks about power and impedance matching.  However, the west coast handbook talks about plate voltage and audio swing.  Both approaches are valid, but the voltage swing approach seems easier to understand to me.  If I have a final with 600 volts on the plate, I want a modulator that will swing that voltage down to zero, and up to 1200 volts.  So I looked at turns ratios of the transformer, and plate swing of the modulator tubes.  It was clear that AB1 wasn't going to do it, so I designed for AB2, which creates driver issues.  I used transistors as emitter followers to drive the 807 AB2 modulator tubes.  The resulting rig worked well, and I know the purists are disappointed there are transistors in the rig, but I've been using silicon diodes for years and it was just fun to build.  It could also have been done with cathode followers.  Good luck!

[W9ZSL]

That's why I was wondering about how far I can go with using different voltages.  On the other hand, I could build the modulator using 6146s which can deliver about 95 watts in AB1 with 600 on the plates.  I could use those 2 x octals for a pair of OC3 regulators to get 210 volts for the screens. I'd have to dig up a driver transformer but the circuit is pretty basic.  The pre-amp is no big deal  I have a recording studio.  I know modulating parallel 5933s with 6146s is weird.  There is a similar modulator in the '64 handbook with the preamp.

I could always just go for the 813/811A combo using the ART-13 iron and a single 1250 supply as an option but considering I have a UTC S-22 and a second plate transformer that can give me 2KV for the 813, why settle for a rig that will run about 200 watts max when I can shoot for 350?

I have all the parts I need for a 807/5933 simple amp and a power supply,so why not build it?  Good practice!

[KA2DZT]

I would just build the 813/811A rig.  It will be easier than the lower powered rig.  Use the 1200 volt PS for a rig that you'll get about 120 watts output.  Later you can use the higher plate supply for the same rig, but maybe use the lower plate supply for the 811s and the higher supply for the 813.  With those two voltages you won't make 100% modulation.  Use the S-22 mod iron and not the ART-13 mod iron.  With the S-22 you can get a step-up ratio which will work better with higher 813 plate voltage.

Fred

[W9ZSL]

That's the eventual plan  I just though I could do something with the ART-13 iron in the meantime.

[W9ZSL]

I just put the ART-13 mod iron up for sale.  Best I use the single 1250 supply with 813 and 811As  with the S-22 for starters.  If I add a second supply for the 813, the plate iron is a dual-voltage so I can start low and go high to bring everything up to speed.