The AM Forum

I found this excerpt online for the 1940's Radio Handbook, by Orr. See the attachment. Aside from the older tube types, is it a good, workable design?

Philip
KA4KOE

More pages

And Page 6.

Philip

Seriously consider the price for the tubes (and spares) you will need.  That is an interesting design but a lot of the tubes used are going to be quite expensive now.  There is nothing wrong with plug-in coils but they are very inconvenient if you plan to change bands often and you need to consider the safety factor of either designing in safety interlocks or being very careful to always power down and short HV to ground when reaching inside to change coils.

I would seriously consider looking towards the mid to late 50s designs with more recent finals (like 813 or 4- series), a Pi type output network and more recent modulators (811A, triode connected 813, etc) unless your focus is more upon building a period specific rig than a big signal plate modulated transmitter.

The modulation transformer (and even more so a modulation reactor choke if you want to use one) are going to be some of the most difficult components to find.  You can take a dual path of choosing a design and looking for a suitable transformer while also keeping an eye out for mod transformers and if you see one that looks interesting consider looking for a circuit to match.

There are a lot of user friendly designs in the 1950s handbooks (both Orr and ARRL), QST and CQ of the era, and other magazines like Radio and TV News.

With a little diligent sleuthing, all the tubes in that design are available.

The 203Z's are the most difficult, easily subbed with 838's.

You will find it more difficult to source the rest of it, than the tubes.  It would be a great transmitter if done properly with period parts for looks and newer iron for longevity.

73DG

The 35T's are pricey lately and HK254's very scarce but the various 25W WW2 triodes such as the 25T/25TG/3C24 are plentiful and provide sufficient drive.

NOS USA 2A3's are out of sight and Id suggest triode connected 6V6/6F6/6L6 (all are 30's tubes) And 203's are mostly gone to air. The 812A is a direct zero bias sub.

That circuit is easily capable of 500-600W input with a single PS of 1500V for the 3C24/100TH/812A

Carl

Carl sure likes to rain on other peoples' parades... :P

203Z= 838 available <$75

2A3= 6B4G available  <$100 pr.

I'll GIVE you a pair of 35T's if you promise to finish the project... :D

Here is a power supply, only one week old...everything there was sourced within the last year.

73DG

Interesting!

Keep in mind I've still got to refurb/recap my Valiant, which is my first old style transmitter project. We're talking in a manner similar to one of the "5 year" plans. Just playing around with ideas. Note that WB4GWA has a nice design in ER mag as well. BTW, I've renewed my subscription.

Philip

You should be able to build a xmtr from scratch without looking at anyone else's design.

Fred.

More itch than scratch with me, Fred, but I know what you mean.

73DG

That's a nice power supply Dennis.  First time seeing 866s conducting.

Phil, I do not want to discourage you but you're probably better off at first getting an existing rig and going through it and breaking/fixing/modifying it.  Something like a Globe King 500 or better, a BC610.  However, if you want to build from scratch FB.  I'd start going to every hamfest especially dayton with a lot of cash and start looking for and collecting parts.  Be on the lookout for chassis/rack panel assemblies.  They are getting hard to find.  I don't even know if new ones are being made.  How is your machine shop?  If you are like me, you are going to have to start learning about metal working and getting tools and learning how to use them.  I advise building the rig into two short 5 foot racks.  Easier to move and get into places like a basement.  If you are patient building a big rig is wonderful and a great retirement plan.  You will really get an appreciation of homebrew gear.  Most of the time what you don't see on a nicely laid out chassis is more important.  You don't see all the different ways parts could have been laid out wrong, botched holes, and other boo boos avoided by years of practice.

Whether to build the project as documented, or do something more modern (and it wouldn't take much in this case  ;D ) depends on whether you want a higher power transmitter that will work and sound better, or a nostalgic higher power transmitter that's true to a time period, but won't perform as well, and probably be more complex and expensive to build.

You could accomplish a better result, and still stay in the tube domain by using more modern tubes like 813s.  A pair modulated by either another pair or perhaps by 572Bs or 810s is a reasonably sane choice.  PI section output networks are easier to configure and require few special parts.

A much better modulator design could be had by eliminating all driver transformers and using some type of direct coupled driver, either tube based or solid-state (easier than a tube driver and works better).

If you're willing to move into the 21st century  ;)  and are primarily looking at 160-40 meters, Wayne WA1SSJ brought Al W1VTP's class E rig over to my place the other day, and we hooked it up and ran a nice cool 400 watts output with DC coupled audio and 175% positive peaks.  I can lift the entire transmitter by myself.

The design depends on your ultimate goal  8)

Regards,

Steve

"You should be able to build a xmtr from scratch without looking at anyone else's design.

Fred"

Fred, likely the answer is yes. However, designing a high power transmitter from scratch with zero experience and only a textbook with formulae is a risky proposition. Why reinvent the wheel?

My idea of a successful project entails the following criteria:

1. The operator is not distilled into a smoking, stinking, and steaming pile of carbonized fat.
2. The transmitter outputs a clean signal at rated output and modulation.
3. The various components do not self-destruct from any variety of failure modes.

Given the cost of the components involved in a high power, plate modulated transmitter, I'd rather not see a ton of invested money become "one with the snows of yesteryear".

I think the biggest lesson here is not to work in a vacuum and obtain help and advice from those more learned in these things than I. This rationale is the primary reason we have this jam up forum with which to exchange ideas.

I have already started on my "5 year plan". I recently obtained a rather nice 6' rack on casters. Just needs a little wire brushing and a coat of paint to look new.

PHILIP
KA4KOE

"FEELEEP"

Not even close. Maybe a typo? 811?

http://www.r-type.org/pdfs/812a.pdf

WOW Carl
You are taking a beating in this thread.

But I agree about vintage special period transmitter designs. Limitations to what was available, and accepting performance for that time period. Not today's 125% pos peak TX and hi-fi audio.

Thats a neat transmitter. The parts are out there if you look hard enough.  I like the old design and old tubes. Its different and different is cool.

More Modern designs will be easier and might work better in some respects, but are common.  "813s here"  Blah... Heard it a hundred times on the air.  I always wanted a 450 TL or 450 TH for final.  Maybe 304TLs for modulators. 

I would never expect you to build something like this with no plans or help.  Maybe try to find an old home brew transmitter somewhere on a classified add.  Then, Fix it, Mod it, update it and get it working.

C

When you have never built a big rig before that is a tall order. 

Phil, I agree with you on your approach.  Look for the parts first and then you will have a feel for what your going to build.  With every rig you built, you will definitely learn something.  At all cost, don't let those with attacking personalities bother you.  If some of these people are as good as they think they are, they would have far less to beat their chest about! The true innovators will help you and you may want to do that on a one on one basis since here you will get some good advice and some bad advice.  Pick someone you respect and work the design issues through them.  I would certainly be willing to donate parts if I have anything you can use once you determine what you want to build! 

Joe, W3GMS   

Well, I've got power fever and want the lights in the house to dim when I key and hear that ominous KERCHUNK!! Most amps I've seen aren't really designed for this duty cycle. I could find a T368 a lot cheaper than a legal limit rated amp.

PHILIP

Joe

Definitely going to do the paperwork first using Orr's calculation format for a high power Class C transmitter. Then I need to do some learnin' on modulation transformers. Orr is a good read. I take it to the "reading room". However, my edition is the 22nd and doesn't delve into plate modulation.

I will post my calcs here where they can be critiqued by the unwashed masses and propeller-heads alike.

There are some very talented folks here and their work in steel, glass, copper, and ceramic is inspiring.

PHILIP

Phil, PM me and I'll send you a much more pertinent issue of the 'Jones' handbook.

73DG

Joe,

You have to lighten up a bit.  I don't always add smiley faces at the end of my comments.

I was just trying to inspire Philip to try building his dream xmtr.  You're correct, he will learn a lot with whatever project he decides to build.

Philip,  everyone here, including myself, welcome all the questions you have, keep them coming.

And, before Joe beats me up again ;D, I don't have all the answers.

Fred, KA2DZT

Started on lunch break. Here is a screen capture using a spreadcheek program.

P

Well if we are postings links here's a really good one:
http://www.somis.org/

Ya but Feeleep
You can not see the pubes in a T368. Big grey box with noisy fans and it will dim the lights hehehehe

The 5 year plan sounds just about right. The gaining of experience and expertise on working with and building HB tube transmitters doesn't happen overnight, let alone the accumulating of all the parts, tubes, transformers and pieces needed to undertake a HB transmitter project. It takes time.

Al VE3AJM

" However, my edition is the 22nd and doesn't delve into plate modulation.  "


Try this....

http://www.tubebooks.org/technical_books_online.htm

Scroll down for the 15 ed.

klc

Downloading it now. Taking forever.

Philip

A homebrew transmitter is available and I wasn't looking for it and its not that far of a drive. It has a 4-1000 modulated by a pair of 4-400s. However, I must be really cheap, as I priced replacement 4-1000s and swallowed really hard. lf the price is not crazy I may take the plunge.

Opinions? How rugged is a 4-1000? This one has the glass chimney and is ventilated via a blower.

FEELEEP

4-1000's are fairly reliable, just don't overdrive them, the grid is a bit fragile.

Personally, I prefer big triodes. RCA transmitting tube manual 4th edition has a great push pull 833a's R.F. amp my HB rig is based on. 833a's are common and easily replaced, and triodes make stable and simple amps. The biggest drawback is they need ALOT of drive. A pair of 833a's could be driven by your Valiant when the time comes, and a big PA system audio amp could drive a pair of 833a's in the modulator, no need for mic and speech amps.

Everyone has their opinions, and some people will tell you that tetrodes are better because they require less drive; which is true. It's just a matter of personal preference and situation, but give 833a's a look if your looking for QRO.

Mike

Oh, c'mon Mike - you're just saying that because you have oodles of NOS RCA 833s that you give away randomly to anyone who doesn't live north of Albany.  ;)

I prefer triodes myself, the biggest drawback these days being desirability of more and more 'radio' tubes by the audio crowd, driving up prices. The drive issue pales in comparison. I guess the upside is, tubes like 845s are now being reproduced to meet the demand. So availability shouldn't be an issue.

I like 810s!

Hi Mike,

For modulators, what is your opinion of running a pair of 4X1's, tetrode connected, vs: a pair of 833A's?  I'm on the fence right now and could always go wid the 833A's.  I have a solid state driver for either.

The specs show about the same power output at 3KV, but the 833A's are drawing grid current and the 4X1's are not.   I wonder mostly about which is cleaner.

T

Philip,

That HB xmtr may be a very good way for you to get some hands-on experience with big rigs.  If it comes with the power supplies even better.  You should check it out, maybe takes some pics and post them here.

Not sure what the price is, but, it may prove to be cheaper then trying to obtain every part needed to complete a xmtr starting from nothing.  I'm assuming that you don't have much in spare parts.

Even if the rig doesn't work perfect, most of the folks on the forum can trouble-shoot the xmtr just from pictures.

Fred

"I wonder mostly about which is cleaner"

Tom,  it depends on which dishwasher rack you place them on, top or bottom.


Another solid helpful comment from,

KA2DZT

I dunno, 6 to 1 , half dozen to the other. There are going to be non linearities using a mod xfmr. The 4-1000 tetrode connected are fine if you don't have a lot audio drive, but you will either need a choke or dropping resistor on the unmodulated B puss for modding the screen. the 833a's are a lower Mu, and are class B so each tube conducts on opposite cycles, but needs a lot of voltage for audio drive, hence a large audio amp and line to voice coil xfmr. I think you already know where I stand with tetrodes, but I have always kicked around the idea of a 4-1000 modded by 833a's.

I would be interested to see how the numbers crunch (THD) 1 vs. the other. The SS driver, is that GFzed's design? I never got around to building it. How does work?

Well, the 4X1 modulators, tetrode-connected,  will need a very well regulated screen supply. I use a shunt electronic regulator type that drops maybe 1 volt under load.The data suggests 1KV screen voltage when run in AB1. That's how I did it with Fabio I and it tested very FB.   I think you were refering to the RF final with the choke, etc.  Yes, I have a choke in there for the screens.


The GFZ solid state MOSFET driver works very well. I was able to get negative feedback from the sec of the mod xfmr back to the low level 1 volt audio input. I had maybe 10dB, so it reallly helped clean it up even further in AB1.  This can be done with ONLY one transformer in line, the mod xfmr. If a driver xfmr is used, forget tapping the sec of the mod xfmr.

I think Steve/ QIX has a direct-coupled ss MOSFET version that I may look at too. I don't know if he has a board made for it. I sent him an email axing.

The only problem I've had wid 833A's is when I make a mistake, I usually burn a hole in the plate. I then change over to 4X1 modulators and they can take anything.  But the 833A is simpler and kinda cool - and LESS air noise, if any.

I can see why you like the 833A's by 833A's rig.  Have you had it on 40M and above?  I plan for 160 - 40M.   I can always use my linears above that.

T

Yes, I was referring to the RF section. Modulator screens need to be well regulated like you said. Just more parts to worry about was my point.

Any tube off resonance will crap out after a while. The carbon plate chinese jobs seem to be better at taking a beating. They work fairly well and are cheap, more watts per dollar...

Like I said, personal preference. A 4-1000 may be more forgiving, but when she lets go, its a costly lesson learned these days; except for Tom Vu, leal estate mogul and pletty ladies extradonaire ;)

I think I have the schiz-matic for the SS driver around here somewhere. I might just have to re-visit things and get my sorry ass back on the air. I really like the idea of eliminating as much iron as possible, especially the non linearities of driving push pull grids.

I think one of the biggest advantages of the ss driver is to permit heavy NFB without more phase shift introduced by a driver transformer(s).   I am still amazed to be able to include the secondary of the mod transformer into the loop and still get 10 dB of NFB.  Ever try it  with a tube driver?

I've never blown a 4X1 after years of running many, many homebrew rigs with them. I've used them in linears up to four tubes, modulators, RF finals, etc.  They are truely amazing tubes.  The only problem I've made for myself is once running the air too slowly and seeing the red paint turn brown from excessive heat. Still, the tube runs FB and I use it today.   Now I have a digital temp gauge sampling air coming out of all my amps.. with alarms.

Well, guess I'll stick with the tetrode-connected 4X1's for modulators. I was hoping you'd sway me but it really is a toss up, as you said.    I have run them triode-connected, which is easier, but found they run a little cleaner as designed in tetrode service.   

Hope you get back on OM.  My goal is to have the new 4X1 AM rig running before Deerfield. That's about three weeks away.

T

I have to say, I prefer triodes as modulators.  The biggest reason is more plate swing without distortion.

Using a self-modulating screen is OK if the choke is sufficiently large to allow full modulation at very low frequencies, and also not to introduce a phase shift as the modulating frequency gets lower and lower.

My last big tube rig was a pair of 450TLs modulated by a single 4-1000 pulse width modulator.  I like triodes in the RF amp as well  ;D

Yes, I just remembered that the plate voltage cannot swing below the screen voltage, thus there is 1000 V (of screen voltage) that would be subtracted from the swing, right? 

Maybe I shud try triode-connected 4X1's again.   

Something that I don't understand....  in modulator service, if the screen voltage in AB1 is so critical and must be regulated precisely, how is it that we can actually connect the screen to the grid and let it swing all over the place with the signal?   How can both cases be linear and do justice to the transfer curves?

I know 813's work very well triode-connected and so do 4X1's. But I didn't have the testing setup like I do now to measure THD and IMD.  The last time I went from triode-connected to tetrode-connected 4X1's I based my results on hearing and reports from guys on the air who thought the tetrode-connected sounded better.

I wud go with the good old 833A triodes, but don't trust them to hold up under full strap.

T

Tom,
You could always use (4) 833A's in Push Pull Parallel!  I did that with 810's and it worked excellent.  That shoudl give you plenty of strap.
Joe, GMS 

I would think 4 833As would be just as good as a pair of 4X1s, and the 833s would very likely produce better audio, and be easier to drive.

That's the way I would go for sure, if I had the tubes.  They are are also, I believe, easier to cool.

No blower noize....

Come to light tom, come to the light.....

I'm thinking, I'm thinking.    Four 833A's would require about 80 watts more filament power than a pair of 4X1's.

But maybe I can make it up by saving money on HV wire -  like this guy did:   ;D


(Note the Romex on the secondary of the pole pig)

Some folks should be forced to turn in their license.

Seems that the BCB boys used 4 833's in QRP rigs and 4 4X1's in a mans rig and both had low enough distortion to fit inside their channel and sound fine.

Triode vs Tetrode is strictly a personal choice and not a clean signal one if built properly.

Carl

Pair o' 4X1s?  Sho 'nuff.

.

OK John,

Now we need for you to channel the late Paul Harvey because we must hear "The rest of the story"  :)

With the "death" naming of the JPGs one might assume you are building a suitably creepy rig in order to get  Stephen King to join us on AM.

I think it is more like JN covertly photo-ed a defunct Bayou CB blaster... 8)

73DG

Another tube that the audiophools use as a sub for 2A3s is a 1619 triode connected.

It's a directly heated tube, which the audio types like, and it uses a 2.5v filament like the 2A3.

There is even a guy out there that makes an adaptor socket for it so it can plug into the 4 pin 2A3 socket.

1619s are cheap and plentiful.

The grid is directly varying with the input signal (well, almost). The screen is not.

I stole those pictures from an ebay auction some years ago.

Just looking at the pictures in that auction scared the hell out of me.

OK, thinking more, here's my take:

When the screen is fixed, there is higher gain and still linear. (tetrode config)

When the screen is tied to the grid, the tube is still linear but has lower gain. This is because it approximates a triode.

So, we are really only changing the "type" of tube - from tetrode to triode and both modes operate linearly.  


In contrast, in tetrode config, if we had a badly sagging screen supply and the screen voltage dropped with signal, this would cause severe non-linearity.  The same holds true if we had a sagging grid supply in triode config.

T

No it didn't. I'm willing to bet you have a lot more hell left in you.   ;)

Looks more like a ham 20M amp by the looks of the coil and the B&W butterfly cap.
A Cber wouldnt have the iron in a galvanized bucket ::)

A real ham uses chimneys, a CBer uses a pair of table fans. ;D

The bucket is great to immerse the tranny in oil when you run it 100% over its rated current.

The key to using 833s successfully is proper ventilation and plenty of it, as demonstrated here.

Makes you wanna just reeeach over and tweak that big bread slicer, dunnit?


This should clean out the rest. ;)

Some folks should not be allowed to have electricity.

With 833 pics like that, someone's going to dig out the one of that South American 833 station.

Looks like a good subject for studying the cumulative effects of RF exposure...Assuming the subject isn't electrocuted before the data can be collected :)

That photo reminds me of a story told to me by some well drilling friends about a customer who was using various pieces of garden hose and garden hose connectors to bring natural gas into his house through his attached garage from his backyard well.  He and his neighbors were both lucky he didn't blow his house into the next township.  Some people DO need a keeper.

YIKES!

I wonder if that plywood is fire rated?

Now if he had added a few more studs plus horizontal reinforcements he could have nailed up a sheet of galvanized metal salvaged from the dump or recycler.

Then screwed/soldered on some more tin to support the tubes and other stuff.

Some people just dont think these things out beforehand  ;D ::)

Carl

I wonder if foil-backed drywall would work?

Is it my imagination or is that little green RF choke across the 50 ohm output?

This may be documented scientific proof that the builder puts his safety first.


T

 ;D good one T!

How about 304TL's Tom.

You can run them in B, ab1 or ab2. They will have much less distortion in AB1 as the Grid is 0.  You are allowed to run up to an amp per tube in plate current where the 833 is 500ma each. A pair makes 750watts ab1 and 1800 in AB2.  AB2, Just needs more drive... Even at 1000 watts carrier and 125% positive its not even in Ab2. They are pyrex and just need a simple fan across the raised socket pins.  No blower motors. 

Robert built a transmitter with 304tls with one 4-1000. It had a strong voice.

Both the 833 and the 304 Tl sure look cool and I would be happy with either but the 304 will be cleaner.


C

Yes, a lot to be said about using 833A or 304TL triodes. I almost went with the 833A's but opted for the tetrode-connected 4-1000A's.  Just finished wiring them. Now adding an electronically regulated screen supply.

To compare power, according to Eimac's datasheet,  a pair of 4X1 modulators in AB1 push-pull will put out between  2300 to 3800 watts of audio depending on voltage.   In AB2, they will put out even more.

Since I like to run my modulators and linears at less than 1/2 to 1/3 of rated power in class AB1 with heavy NFB, the 4X1's in AB1 will give plenty of conservative headroom for cleanliness. That's why I went that route. We'll see what the THD tests look like in a coupla of weeks.

T

Yes. That is a much larger tube. I have a new one here in the box if you need it.

I thought you had this transmitter running before?  I seem to remember you putting 30 DB over the top of Steve out here in AZ.  HEHEHE

C

Thanks for the tube offer, OM.  Though, I already have six 4X1's and five chimneys.  This makes two spares, so all set for now.  They never seem to wear out. Been using a few of these same used/pulls  4X1's for 20-30 years now.


You are probably thinking of Fabio I, the 4X1 Plexiglass rig. When I built the class A linear project last year, I ran out of room, so tore Fabio I down.  I'm using some of the same parts to build Fabio II and doing a much better job. Better layout, bigger parts and higher voltage ratings this time.  And, all aluminum construction, no Plexi... ;)

I have my workbench, office desk and shack all crammed into one room, so if a rig isn't actively used, I get rid of it.

T

Yeah. Thats it.  Same config though right?  4-1000 by a pair triode connected?

C

I am simply amazed at the length of this thread after posting my 5 year plan!

Philip

It's easier for some then starting a new thread. Six months from now no one will remember under what title all this information was post. And then, regurgitation will start somewhere else. It's not uncommon.

way to much qso'ing in the technical area. come on people get with the program.  ;D

We have to conserve the internet.
No QSOs unless the content is important, never repeats, and is in the right category.

Peek Freens.

The same topic started six months later can produce different participants, different opinions and sometimes evolved techniques that were not available in the first thread.

I look back at some of my old threads even two years ago and realize how differently I would have done things.

Nothing stays static - especially in the technical whirl.

The alternative is to just read and reread old threads and never start a new one. There is so much info on the web, we could go through several lifetimes doing it. But there's nothing like a new thread, even a repeat, to get the blood flowing of old and new participants.   [caw mawn]

T

.

Ditto, I usually go no further than my property lines for what I need as what isnt in the basement or attic can usually be found out in a storage trailer, old van or truck body or an outbuilding.
The best thing about old age is Im always finding something new ::)

The odd needed part can always be found somewhere on the Internet or a visit to ESS.

A slightly newer edition of the Orr Handbook is available on line as a PDF.

http://www.tubebooks.org/Books/intro_orr_radio.pdf

I think it is the 15th edition.

I don't think Steve QIX would approve of those clipleads

IF it is it will have a little red stamp on it that says FIRE RATED. Fire marshals don't like it when you paint over that stamp.

Carl sure likes to rain on other peoples' parades...

203Z= 838 available <$75

2A3= 6B4G available  <$100 pr.

---------------------------------------------

And 203's are mostly gone to air. The 812A is a direct zero bias sub.

Carl

----------------------------------------

Not even close. Maybe a typo? 811?

http://www.r-type.org/pdfs/812a.pdf

-----------------------------------------

WOW Carl
You are taking a beating in this thread.

But I agree about vintage special period transmitter designs. Limitations to what was available, and accepting performance for that time period. Not today's 125% pos peak TX and hi-fi audio.
-----------------------------------------


Nah, I just consider the sources and completely forgot about this early Page 1 part of the thread.

1. I dont consider $100-150 for a pair of 838's reasonable and a 203Z is not a 203.

2. Nor do I consider a pair of 6B4G's at around $100/pair reasonable either.

2. A 812A is a low mu tube as is a 203A. I dont see how anyone could call a 811A a sub.

There are many other candidates that the audiophools havent ruined yet such as the 814/VT-154 if you want pre WW2. It is also very conservatively rated at 60-65W Pd. The 803 is going up in price but also capable of a lot more power than listed due to a specific transmitter design being used to set the specs. It is really a 813 on steroids with about 20M the upper limit for a KW input.

Some Euro tubes are cheap in the US as the BC/SW TX's are being junked and NOS spares still around. Check out the TB5/1750.

Dont overlook the 4X150A/4CX250B family for audio. They deliver a lot of power even at 1000V. There are cheap Svetlana and Chinese ceramic sockets that fit the standard 1 1/8" hole and add some air holes for a quiet under chassis muffin fan.

For the late 40's look there is always the 4-65A, 4-125A, and a pair of 4D32's will get you a 250W carrier at 700V.

The Russian GU-81M is impressive looking and delivers a lot of power without air. Tubes and sockets are quite cheap at Eu hamfests.

Carl