The AM Forum

I dug this out today after several months of debating over what tubes to use. I had initailly wanted a pair of 4-400's in the PA, but I had to get back to reality. I have quite a few 813's also, and this would definately give me a pretty good head start. I remembered that I had bought this several years ago to save it from "CB amplifier" conversion; someone obviously put a lot of effort in building it. It appears that it was set up for 20M when I got it since the the input coil is labeled as such. I don't know what happened to the coils mounted to the output jackplate, but I do have the remnants of the swinging link.
Any ideas? Strip it and use the basis on a Pi-net, or rebuild it as link coupled output? Does the latter actually need to have a swinging link, or can it be fixed; especially if I was going to use it as a monoband TX (75M)?   
Phil

If you found yourself a bunch of BC-610 coils with the variable link inside, you could use it pretty much as it was designed.  The  coils with fixed coil will also work but you would need to put a variable cap in series with the ground side of the link to ground, for it to avoid over coupling the final on some bands. The coils with the fixed link that look like BC-610 coils are in reality the input coils for the BC-610 amplifier know as the AM-141.

What is that big resistor in there? It looks to be a very low value?   

It looks to be in series with the filament transformer primary along with the front panel mounted rheostat. The transformer is a Stancor P-2030. I haven't researched it yet, but maybe it isn't a 10v 10A unit; maybe a 12v and that is why the use of the monster resistor. It would have 10A flowing through it? Rheostat doesn't look that tough though.
Phil 

The resistor thing could work in your favor by eliminating inrush currents to your filaments. Lots of old buzzard rigs used them. Perhaps they were cheaper than variacs or were around before variacs?

It a good basis for a nice HF amplifier  ;D ;D 

Having the input network right next to the tubes is real sloppy design. At least have a shield between them or better still put it underneath.

Also lower the 813 sockets down so that the visible grid leads are below the chassis.

A pi network output and parallel tubes might be a better choice depending what you want to use it for.

Carl
KM1H

I think the nameplate on the transformer reads P-3020 not P-2030.

stancor P-3020 is 11.0 ct @10 amps.

I wish I could find a place that has those old buzzard style nameplates concerning radio that you could buy new.

You're right Derb, I'm going to have to get my eyes checked again! It did say P-3020 -- maybe a little dyslexia too!
I'm just going to settle on 813's and start sketching out a layout tomorrow. I'll worry about the modulator after I get the Tx. up and running. Probably going to drive it with an ARC-5.
I just ran across an article in an old ARC issue where a guy made them like etching PC boards. I can scan it and send you a copy if you'd like.
Phil

Last part of that referring to the nameplates.

that would be kool, tnx.  :D Pair of 813's will strap big time, no need to apologize there.

Forgot to mention earlier that if you want the best stability with 813's use sockets that have the push and lock metal bases and ground them good. The GE tubes with the ground ring built in the envelope aid stability also.

Carl
KM1H

means a rf ground - use some good copper STRAP to ground it, no little tiny wire. use the screw that mates the metal shell to the ceramic on both sides.  :D

Tim, I sent the nameplate article. Pretty cool idea.
I'm going to start taking that deck apart today to get it to a workable level. I'll post my progress on it.
Phil

Tim who? Tiny Tim ;D ;D

Thanks for no help guys ;D ;D

I finally figured out that the Derb is a Tim. I always thought that was your first name Derb.

My next project is designing an outhouse for the moon. Any suggestions?

Sorry O.M. I've been wearing blisters on my digits working a flat file on this front panel refurb. Figured out his name when I got his Email, but I guess we can call him Mr. Derb.

Very nice rig! I'm jealous! ;)

derb came about from N3DRB.

Yeah, I kinda figured. Actually I'm pretty familiar with it another way. I worked for Chrysler dealerships for about 10 years as a line tech. The Chrysler "scan tool" is called a DRB. With only 2 DRB's in a shop of about 20 techs, you would hear throughout the day "Hey! Who's got the Derb!" About 80% of the work required the use of the DRB III. You could own one of those setups complete, if you had about $10K to part with. Not many mechanics I knew who were that fortunate.

O.K. guys, I have this chassis stripped and I'm trying to figure out which way to go. This being my first big HB rig, should I go for the gusto with the 2 x 813 PA, or start with somthing simpler like triodes; 572B's or 811A's? Like I stated before, this will be a 75M single bander for now. No bandswitching to fuss with. I'm set with enough parts to go either way, just needing a little advice from others who have already been there.
Phil

I have built both.
I had a push pull 812A rig modulated by 811A's, with seperate power supplies.
Smooth working rig!
I used KW plug in coils and 30kv miniture vacuum variables for neutralizaton caps (out of an old 4-400 paging transmitter).
Very nice 300 watt rig. I sold that rig to somebody who paid to have it crated up and shipped to Washington state I think!

I also built the 2x813 rig. Another smooth working rig, and very robust, 600 watts carrier all day long.
You cant go wrong with 813's.

I guess it depends on what you have for modulators.
I always want as much audio power as carrier output power, that is 600/700 watts for the 813's.
If you cant do that, then do the 811/812 rig, as 811A's will modulate that fine.

Brett

Depending on the driving power available the choice would be 572's or 813's.  A good layout of the 572B's in parallel along with around 47-100 Ohm 2W carbons in the grid leads should eliminate the need for neutralization as a 80M monobander.

Also when feeding the bias use a  roughly 3000 Ohm 1W resistor across the feed RFC. This eliminates any chance of a TPTG oscillator.

Another tube choice is the usually ignored 803 which are at the low end of the price range on Epay and hamfests. They can take a beating.

Carl
KM1H

I was figuring with the 572B's, I wouldn't have to worry with the screen supply either. I could also modulate them with a pair of 811A's using the ART-13 mod tranny I have on hand. I'd also prefer to use pi-net on the output. Anyone have some schematics I could smoke over?
Phil

I'd flip the tube complement - 811s in the final, and 572s in the modulator. But I like a lot of audio.  ;D

Ive heard the ART-13 iron cant take more than 200-250W on a good day? Also dont hang any high level clipping or other junk on the secondary that can reflect back to the primary. Don has commented on xfmr failures recently on here or QSO.

Carl
KM1H

If that's the case, I think I have one more mod tranny that might work better. I'll be honest, I can't afford to ruin anything; literally. Work has been pretty slim for most of this year and we've been working an average of 6 days a month at the plant. I'm really fortunate that I have collected a lot over the years, so at least I do have things to play with in my newly acquired spare time.

Here is the start on the RF deck...

All the parts look too small to me.
What are you going to do with the rig?

The RF parts look good for 200 watts, and you can get 700 watts carrier out of a pair of 813's....

Brett

I don't plan to exceed 2KV Ep and that is what I figured it for, but your right. After I actually checked the C/tune air gap, I found out it was too small for that; it's only 3/16" and needs to be at least 1/4" for that voltage. I could reduce the plate voltage to 1.5KV, but then I don't have enough C. The tank coil looks like #14 and I could bump that up to #12-#10, but it would have to be hand wound; that's not a problem though. The C load will probably be too small of an air gap also. Trying to juggle all the figures and finding the necessary parts is a lot harder than I realized.
I'm not looking to build a powerhouse on my first time out. I just want to be sure I'm doing all the figuring correctly and have good construction practices.
I'm basing all my calculations on info from the Bill Orr handbook. So, to figure plate load resistance, I got my starting figures from the RCA handbook:

For 813

ICAS Ep=2000v
Ip= 200mA   x2 tubes=400mA
For RF plate load resistance
ohms=Ep/2Ip

=2000v/.800mA
=2500 ohms

Am I on track here, or have I goofed up?

Phil

Your calculations are ok, but with small parts you get arc over and melted coils.
I run my 813's at 2000 volts and 400 ma, they do that all day long, but with mod peaks you can get 5/6 KV across parts with the extreme audio peaks.
For loading caps, you cant have enough, 2000 pf is a place to start, and I always disliked fixed caps switched in.

For the plate tuning cap, I think Bill Orr lists 4x the plate voltage, which is a place to start, but more is better.

I suppose you could run 1500 volts and 300ma, or any combination, but must resist the urge to crank it up..

When I build, I always overbuild, things last that way.

Its easy to wind a coil with thick copper wire, or tubing, solder on some terminals and mount it on insulators, or install bannana jacks and plugs and switch coils for bands. The grid is easy to make multi band, but the pie net handles a lot of power and voltage, and its nice to not have big switches and shorted coil turns to cause problems and lower Q.

Yes, parts are getting hard to find, and sometimes you have to go with what you can get, or make better parts yourself. You can make all kinds of coils, wind enameled wire on PVC, I did that with my 813 rig, seperate coils and links 160 to 10, you can make big coils also, big caps would be tough.....

What is the modulator going to be?

Brett

 

Also, I would not have made a hole in the front panel, with 813's there should be nothing to see, no plates glowing, and a lot of RF will come out and infect your microphone.....

Brett

The tubes look a bit close fer my tastes....

klc

O.K., nothing is written in stone at this point. I still have more parts stashed that I'll have to go through. I could use the front panel for the mod deck, use a different tube or combination of tubes for RF, or totally bag both and start over. Let me give a run down on the tubes I have on hand, that's probably a good starting point to decide from.


4-400 QTY-3
4-125 QTY-6
4-65   QTY-2
813    QTY-10
304TH QTY-1
572B   QTY-2
811A   QTY-10

I'm going to include the triodes. Since this will never be more than a 75/40M Tx, I wouldn't imagine that neutralization would be much of an issue.
The largest plate iron I have is the Stancor from a BC-610. Remember, this Tx doesn't have to make max smoke; I'm sure I'll build another pretty soon. I'm not going to really worry about the modulator for now: I'm just going to concentrate on the RF deck and the power supply.
This is basically just a learning experience, and a chance to hone my layout and construction skills. so that when it is time to build "The Ultimate", I'll be 100% sure I'm doing it right.
Phil

Phil,
This is the fun part!
Makes you think, and its fun.

I might use 813's modulated by 813's, 813's are very rugged and problem free.
A pair of 811a's at 1500 volts would be a bit light for a pair of 813's but would be simple, run the entire rig at 1500 volts, you would be around 12.4k ohms and 340 watts out into 3750 ohms for 600 watts input.

813's as modulators at 2000 volts would be 16k and 335 watts out, but in AB1 (no driving power).

Part of the design must reflect what parts you have or can make.

If you had killowatt plug in coils for push pull, 811a's (or 812a's)  modulated by 811a's works real well and is very simple for 300 watts out. If using 811a's, you need no bias supply for the modulator or rf deck, no screen supplies, just a grid leak resistor....

Mod transformers you can get might also restrict your choices.

I have a stanley mod trans 8000 ohm to 2800 ohm 300ma trans (with screen winding),
a thordarson multi match 300 watt (T11M77), a freed 20,300 ohm to 9100 ohm 250 watt job, and an old bc610 mod trans 16k to 8325...
I could trade  any of those for something...

Brett




   

Sink those tube sockets so the internal grid lead is below deck otherwise it most likely will need neutralizing. Its also a good idea to use sockets with the bayonet base and then ground the shell. The sockets you have are OK for GG or audio.

A pair of 572B's or 4-125A's will modulate it well.

Id stick with that Tune cap for now, see how it performs. Ive had no arcing in a TX with a single 2 KV 813 at the same spacing. A lot depends on how well you do the pi-net impedance transformation. The ART-13's dont arc either when the voltage is cranked up.

Carl
KM1H

I dont think you need to do anything with the sockets at 80 and 40 meters....

Here is another neat window idea....dont make holes in the front panel, but mount the tubes (modulator) on a small seperate chassis, mount the small chassis with the tube sockets on it on the top of the modulator deck front panel, then install a piece of glass in the rack/cabinet in front of the tubes.

That way you dont have to cut holes in panels, a piece of glass is about $4.00, and it looks sharp.
I use screw terminal strips to connect the tubes to the mod deck.

For the RF deck, I would rather try and shield it, or at least not have big holes to let the RF out, although in many cases it does not seem to cause problems in the shack.
It MIGHT allow harmonics out (pre pie net) and into your TV's and so on....

Brett

It never hurts to do a quality build instead of a schlock one. Now it will be ready for other bands if this builder or another wants. Besides 40M can get squirrely also with a nice high gain pair of tubes.

Carl
KM1H