TheRadioBoard

I have made progress building my 40-meter tube transmitter. Since my last update to the RadioBoard,

http://theradioboard.com/rb/viewtopic.p ... ight=ab8if

I have added a VFO and a driver amplifier. The driver delivers a sine-wave output that is amplitude adjustable up to 100 volts peak into a 10kohm load. That should be adequate drive for an 807 final.

Here is a current picture of the transmitter…



At present there are three chassis boxes. The power-supply sits to the left and is clamped to a pair of angle-aluminum rails that make up the sub-chassis. The VFO box sits on top of the driver which in turn is clamped to the sub-chassis. The empty slot to the right is reserved for the final amplifier.

The pictures that follow show the VFO and driver in various states of development. One of the best pieces of advice on the RadioBoard is the need to breadboard circuits if you are building an untested design. I speak from personal experience when I say breadboarding can save you a lot of time, money and frustration!

So this is how the VFO started…



It looks more like a rat’s nest than a breadboard, but it proved the feasibility of using a Pyrex beaker for the coil former. Phenolic formers are pretty much unavailable these days while those made of ceramic cost a small fortune. Once working, the circuit was transferred to a chassis box.



The VFO uses a pair of 6J5 triodes. One functions as a grounded-cathode Colpitts oscillator. The other is a cathode-follower buffer. At this stage I found that mechanical flexing of the aluminum chassis caused severe frequency shifts. One culprit was varying parasitic capacitance between the coil and the bottom cover plate. To improve frequency stability the VFO box was stiffened internally with corner braces, and the VFO was mounted on top of the driver chassis using some 20 spacers.



This made the VFO chassis much more rigid, while reducing the coil’s parasitic capacitance to ground. The VFO was finished with the addition of a front panel and a Model N National Radio Velvet Vernier tuning dial. I’m curious as to the date of manufacture of this dial. Nineteen forties??



This picture shows the VFO hooked up to the driver breadboard. The driver uses a 6AG7 pentode to drive a pi-coupled output. Here the breadboard worked perfectly at first turn on. Before building a breadboard I like to run a computer simulation. It is always encouraging when a breadboard works the same as its simulation. That was the case with the driver. Sometimes they do not work alike. That leaves me figuring whether the circuit design is flawed or the simulator is in over its head!

Finally (for now) here is the bottom view of the transmitter.



I make generous use of barrier strips and lock terminal lugs. They essentially serve the same purpose as terminal strips (now made from nonobtainium) with the additional advantage that I can add or remove chassis boxes without soldering or unsoldering.

If there are any ideas here that will assist or encourage a future builder, they should suffice as my excuse for bragging!

To be continued…

73, Roger

Roger,

That rig is beautiful! Can hardly wait to see the PA.

I've been meaning to ask: What's the glider in your avatar photo?

73,

_________________

http://kr1s.kearman.com/
http://qrp.kearman.com/

The National N was available at least as far back as 1928, but certainly still available in the 1940's. However by then it seems the ACN was the most popular for this use since it allowed direct calibration.

I assume you're doing telegraphy only, since that is probably sufficient for telegraphy, but not for telephony.


Phenolic formers also also pretty much useless at HF. Polystyrene makes a good, easy to use, low cost coil form.

Hi Jim,

Thank you for the compliments! The glider is an American classic, a Schweizer 1-26. It provides my altitude adjustments!

http://www.retrothing.com/2007/04/schweizer_126_b.html

73, Roger

Hi Jon,

Thanks for the information on the National dial. I considered going with a direct calibration version but the plastic cursor seemed less than robust. The classic vernier readout was also a factor in going with the earlier model.

Yes, this rig is for CW only. The extra cost of a larger power supply and modulator did not appear to translate into extra operating fun.

I ruled out using polystyrene only because of its limited temperature rating. My concern was that a hot tube nearby could soften or deform the material.

Roger

That's beautiful Roger. And the way you're documenting the build is an education for us novices. Thank you.

73, Dan

Thank you Dan, it is my pleasure to help.

73, Roger

Very nicely done!

-Bill

Very nice. I like your use of terminal stips and lugs. Hadnt thought of that.
I build my VFO on wooden chassis. (pic in the German section) A grid hartley on wood is so stable I think there is something wrong. I imagine a colpitts would be no different but I dont know for sure.
Looks like you are having a lot of fun.
When is the test TX?
Don VE3LYX

Thank you Bill!

Hi Don,

Thank you. I'm not surprised that your wood chassis is very stable. The temperature coefficient of expansion for pine (for example), is only 50% more than aluminum. Add to that the higher thermal resistance of wood, and its physical dimensions could actually be more stable than aluminum! I have seen users cover wood enclosures with aluminum foil to eliminate "hand capacity" effects as well as radiated emissions (such as TVI). And you can't beat the price of wood and the ease of machining!

Yeah, I gotta admit I'm having fun at this! As the old saying goes, it's the journey rather than the destination that is the most fun!

The test TX is probably a few months off. I'm working on the circuit simulation for the final amp right now. Here is a screen shot where I have used a pair of 6F6s to approximate an 807 (the simulator did not have an 807 in its library).



The results come pretty close to replicating an 807 Class C amp running off 400 volts. I think I'm ready to draw up an actual schematic and order some parts now. After that there's the challenge of making a base shield for the 807 and an RF-tight enclosure for the final amp. Plenty to keep me busy... and that is good!

73, Roger

Very interesting project. I didnt know about Electronics simulators I use engine dyno simulators in my engine business and drag track similators and found the results very accurate (if you dont lie to the program)Several time we were able to predict with in a 1/10 of a second actual eventual results. Once we were bang on. (Must have been an accident)
Ill be watching for the test TX. And yes for sure the journey is the best part of the trip.
Don VE3LYX

very nice !

is the vfo like the one shown in the Dec 1953 RCA Ham Tips ?

Peter

Calumet baking powder used to come in metal cans with a twist-off lid. The can was a bit wider than the diameter of an 807. I trimmed down a can so the top extended just a bit higher than the bottom of the anode, and cut a hole for the 807 socket in the lid. Painted it battleship grey, and it looked great. Of course, this was in 1963. They're probably using plastic containers now, but you might something useful at the supermarket!



An option is to mount the socket beneath the chassis with spacers, so the bottom of the anode is just below the chassis.

73,

_________________

http://kr1s.kearman.com/
http://qrp.kearman.com/

I used a coil shield I got for a nickel at a hamfest...reaches up to the same height as the unobtainium Millen shiled: