As I thought about having grown up around vacuum tubes I realized how much I miss them...the glow, the smell, the distorted audio, burns, and so on. Therefore I decided to scrounge around my junk box and build a tube something. The something is a Part 15 AM transmitter.
I wanted the transmitter to be transformer isolated from the power line, be frequency agile, be crystal controlled, and have some of the controls of the "big" transmitters. The junk box yielded all parts except for two tubes and a panel meter for a total cost of $17. The third tube (small one in back) came from the junk box and it is marked "Amperex" with a date code of 1951 yet it tests good and works fine.
The design is a plate modulated class C unit with the only non-tube concession being using a PLL to set the frequency by means of DIP switches. Otherwise, all is old school tube assembly techniques.
The finished (except for labels) unit:
The two front tubes were selected for the relatively highly visible filaments. The meter indicates the final plate current which is tuned for a dip with the knob next to the meter. The two switches control the filament power and the B+ power just like with the big units. The hole next to the switches will someday be used for a meter selector switch to select current or modulation but this hasn't been designed yet.
Here's a low light view showing the glow from the tubes:
The top of the cabinet is smooth with nothing but the tubes protruding. This was done by using a sub-panel as the chassis for the tube circuitry show here with the cover removed:
The power transformers can be seen in the lower left. One powers the filaments at 12.6 VAC and the other steps the 12.6 VAC from the first up to 120 VAC for the B+.
To make servicing and changing things easier the sub-panel rotates up and can be secured in place exposing the circuitry. In this view the circuitry can be seen as well as the tuning capacitor and the tank coil and the PLL is in the lower middle of the pic:
One off label use for tube equipment such as my first good stereo amplifier has been to warm cashews. So, to commission this unit here
Now for some boring details. The final input DC power is 74 mW when tuned and connected to a 3 meter wire antenna. The maximum modulation is 84%. The audio is not as clean as the solid state transmitters I use but it is acceptable. I found that the tubes are spec'd at 7% THD typical but using two negative feedback paths seems to have improved this.
Range tests haven't been done yet but a quick check with a cheap portable the signal from the antenna in my basement could be heard at the curb 100 feet away. More serious checks to come.
There is no matching to the antenna at all. The signal comes directly off the plate through a capacitor (for safety...stops the B+ from getting on the antenna). With this simple arrangement the harmonics are all down by 34 dBc or more. There is no hum seen on the scoped RF and audio and none is heard on the signal. The tank tunes from 1100 kHz to 1700 kHz.
During prototyping I tried coupling through a link coil to a 33 ohm load resistor and achieved good results. The transmitter loaded nicely up to 100 mW input and gave an efficiency of 68% which is pretty good for such a simple circuit. This has no meaning though when used with the simple wire antenna.
This was a lot of work doing the prototyping, the chassis and panel work, wiring, testing, and adjusting but the final product was worth it.
Neil
Beautiful, *drools* I love to see some actual work here on the site, and with tubes, that's almost a lost art!
12CU5s, big cathodes make them seem like 50C5s, they look like them. With 100 mw in, it seems you'd get lots of modulation capability with a good circuit, in something direct coupled like Heising mod. .5 ma, 200 plate volts?
Larry Lythall PLL design? Looks like his older one with the bigger chip. I like putting the new with the old world, frequency agility and stability where needed in an oscillator with a tough tube circuit.
Lookin' some more.
Thanks, Nate, for your comments. The 12CU5 tube is the 12 volt filament equivalent of the 50C5. They are rated at a maximum average voltage of 150 volts. This circuit operates at 121 Volts through a full wave bridge and three stage RC filter. Raising the voltage by lowering the filter resistor would increase ripple so I didn't push to 100 mW for this reason. The grid drive was adjusted to give the maximum RF output. Going higher just increases the DC input power with no gain in RF output.
The modulator is Heising and there is a big problem with this. The modulator plate load is an inductor which changes the load with frequency resulting in a roll up in frequency response to about 700 Hz. This is one reason the audio is not quite right but good enough. The modulation percent is limited by the cathode bias of the tube which is at 10 volts. This means the output to the RF amp cannot go lower than about 14 volts but I applied a trick which somewhat overcomes this limit.
The PLL is my design but is very similar to other designs I have seen. I built this one a few years ago but never used it until now.
Neil
It seems that 12CU5 tube would be pretty rare, but in this low power use it will probably never go bad. Interesting controlling the output with grid drive, as some builders do with MOSFETs.
There's a peak at 700 hz, or is it that audio below 700 hz is rolling off? Maybe you could EQ the audio going in or put a resistor across the inductor. The inductor is too small do you thinK?
Another idea to get closer to 100% mod could be the 'modified Heising', with a paralleled resistor and capacitor, like a 100k with 1 uf, connected between the modulation inductor and RF stage. That would lower the DC voltage into the RF final so you could drive it with RF more, and the cap would bypass the audio frequencies at full strength, which should allow more mod, to 100% and more.
That might be for another rig if you make one, I wouln't see cutting up the nice circuit you have now!
I think the Lythall synthesizer uses a 4059 chip I think.
In any case, I love the old stuff, be proud of that sitting on your table.
Neil, another possibility is to modulate in the cathod as Cunningham did with a transistor.
Rob Veld
Neil,
This is truly a beautiful effort. On all counts from design to execution. You've got my vote for the "Gold Soldering Iron" award. If there was one. LOL. Very nice indeed!
TIB
tubes have a special nostalgia about them. nothing sounds or looks better, at least to me, than an old tube AM transmitter.
Tube AM transmitters was the beginning of the AM Broadcast Band, and all these years later, in many cases the tube transmitters dating as far back as the 50's, are very much still in service and chugging along.
i don't think we will ever be able to say that for these solid state boxes some 60+ years later.
i think as long as AM remains analog and parts remain available there will be those stations who will keep the tube boxes on the air.
Two minutes ago I was dreaming in a dark room, now I am viewing a dream transmitter on the brightly colored computer monitor. What a wonderful way to wake up!
Neil, you already know I am nostalgic about tubes and have been expressing the wish for a part 15 tube transmitter for several years, and here it is!
I noticed you have been silent here on the board for awhile, now we know why.
I second Tim on the ""Golden Solder Award"!
The World's Newest Old Transmitter.
If ever the FCC allowed more power for part 15 you can make a linear amp out of a few tubes to boost the power to 1 Watt or even more if allowed. Grandpa had a tube linear on his Realistic Navaho CB and that linear was very clean. It was black and I think only had two tubes and a pot on the right. It had a power switch and a switch that put it in standby as the tubes warmed up. Very simple linear amp as the top looked like a cage and you could really see those tubes glow. I'm not even sure how many watts it was but when in a fringe area Grampa could go from an S2 to an S10 with the flip of that standby switch. No hum and didn't have all the harmonics like the solid state linears too. I remember Grandpa telling me that tube linears and transmitters were far better. He was gonna buy me a tube AM transmitter to play with as a neighbor kid had one and he could get 3 miles with a very nice signal too. We could hear it in the car on the way to the Cash and Carry in New Port Richie, FL. I'm sure too he was pushing past part 15 cuz he'd try and get the longest piece of copper wire he could set up and let it run up a tree. But boy the TX sounded good from what I remember of it. I was a kid back then so you know its a bet of a blurr.
Nate, I tried to equalize the audio in the preamplifier stage using an RC network which gives a roll off up to 700 Hz and then is flat but this made things worse distortion wise so I took it out. There is probably some problem with the current balance between the modulator and oscillator and I will pursue this later. I hesitate to try too much experimenting because it will make the unit look hacked. Been there before.
The audio sounds pretty good on newer radios but sounds tinny on older radios, perhaps due to the modern pre-emphasis used today. Since this is for my personal use with old radios I would like to get a more flat response so I need to do some homework and calculations. Otherwise, I am happy with the outcome of the project.
Finding the 12CU5 tubes was easy. There were several offered on the popular web auction site and I bought a pack of five NOS in original boxes for $15 shipped. If anyone needs older tubes it does pay to shop around since the prices vary quite a bit as does the tube condition.
Neil
Very well built unit you have there.
Looking forward to more reports as you tinker with this vacuum tube marvel.
TheLegacy, I never liked the idea of linears for AM because of the need to keep them linear which implies using something other than a class C final making them run very hot. It makes more sense to me to build the power into a class C transmitter to begin with.
Many years ago I was the "chief engineer" at a campus carrier current station and the transmitters were pretty bad and sounded awful. A hi-fi enthusiast and I decided to build our own transmitter. He designed the audio and I the RF section. The unit operated at 25 watta DC input and was matched, as best could be done with variable caps and B&W coils, to the power line in our dorm. It produced clean audio and radio signals. It never went on the air other than for carrier current on my watch.
After I disengaged with the station, the story goes that some genius decided to string a long wire between two poles and put the transmitter on the air. This couldn't have worked right since the output network was not designed for this but the FCC heard the signal and shut them down, even their carrier current operations, for a year. This is third hand info but it is plausible.
Neil
The bass rolloff problem is fixed. I had neglected to install the cathode bypass capacitor on the modulator (arrggghhhh). Putting a few microfarads in there brought the bass up to par and it now sounds really good. A slight side benefit is that the bass has the "tube sound" with just enough second harmonic distortion to sound like the old time jukeboxes and guitar amps. It is really a nice sound.
Nate, I did some research and found the modified Heising circuit you suggested. I am going to try this to see if it improves things more. Right now the audio sounds crisp and clean with nice bass but the modulation is still a bit low. This might be the fix and thanks for the heads up.
Neil
During all this discussion about modulation, how difficult would it be to build several modulation choices via a selector switch?
You could change your mind every once in awhile and try a different way.
Neil deliberately used an analog meter because they were the real-deal in the tube era.
Now there are digital meters, but the analog meter gives a flavor that provides authenticity.
I'd go overboard and have meters for AC power (110v); filament power (12V); B+ (DC power), the millAmp meter as used by Neil, and as he intends adding, an audio level meter.
If anything got left out from my meter wish list let us know.