Home AM Transmitter (again!)
 

The saga continues. I picked this up again after two years on the shelf.

https://lh3.googleusercontent.com/-q...07%2520020.JPG

Going back to Phil's original schematic and instructions, I still couldn't make it work. Then I tried to build an Antique Electronic Supply kit that I bought some years ago, but I busted the graphite slug in the oscillator coil. I bought a new coil and I busted that one, too.

So I started to build a "Li'l 7" again from scratch. This time I put it on a big board so I could lay it out just like the schematic with all new components.

https://lh3.googleusercontent.com/-O...07%2520101.JPG

When it was finished I powered it up, turned on a couple of radios, and adjusted the RF coil until I busted that one, too.

https://lh4.googleusercontent.com/-G...07%2520102.JPG

I'm down to my last RF coil, and I'm convinced I'm missing something very obvious, so I'm going to ask you radio-smart people to check what I'm doing one step at a time. Here's step one: Putting in the Audio Signal. I have a stereo phono plug in a CD player. The lines from the left and right channels (tip and ring) are joined to carry the audio in, and the line from the sleeve goes to ground. This is how I have it set up with a 100K pot, and I've followed the signal with headphones all the way to the ll7L7.

https://lh4.googleusercontent.com/-D...520diagram.JPG

So, does this look right so far?

The potentiometer shows a linear 0-100 K ohms until I plug in the stereo plug. After the stereo plug is in, I can't figure out what the potentiometer is doing--sometimes it looks like a bell curve peaking at 25K. And with the headphones, the audio cuts out within 10 percent of the range of sweep.

That looks correct, but I think that I would put a capacitor in series with the input, so that the low resistance of the headphone output of the CD player (or whatever) does not affect the resistance measured at the pot. Also it appears that you have the pot wired backwards, with ground going to what should be the high end. It will work ok as wired but the pot will need to be turned CCW for max input, rather than CW.

But back to more basic stuff... does the tube light up? does the power supply make around 125 (or so) DC? ... why do you keep breaking the coils? do you not have a plastic (nylon ?) hex tool for adjustment?

jr

Edit add:
Also check the wiring of pin 2 of the oscillator coil... schematic shows it connected to ground, not to pin 4 of the tube. Connection of C-6 would also not be correct if this is so.

Yes, I realize the pot is backwards--and I am using it backwards--and I'll fix that.

The leads from the tube pins are:
1- dark blue
2- black
3- yellow
4- orange
5- light blue
6- red
7- black
8- green

Oscillator coil pin 2 is green wire. All greens go to ground. Tube pin 4 is an orange wire.

I started with the audio input because it's so basic that Phil's instructions do not describe it in exquisite detail. Would it be better to use the output for a 4-ohm speaker than the headphone jack?

Breaking the grapite slugs: I never had this problem before, but now the adjustment is very stiff (I am using the appropriate plastic hex). I finally figured out that I've been moving past the useful range, to the end of the threads. Then it binds, and when I try to back it up, it breaks. But it only took me 3 coils to figure that out!

Yes, the tube lights up. And the electrolytic caps hold a charge.

I'll fix the pot and see how it works with the speaker output. Then I'll do the voltage readings.

Thanks very much.

"Oscillator coil pin 2 is green wire. All greens go to ground. Tube pin 4 is an orange wire."

If that is the case, do you then have *both* ends of the 47k resistor as well as the grid of the tube (orange wire) connected to ground? Take a close look at the wiring between the two lugs that the resistor is connected to.
jr

Hey that looks like a cool project...... Can you post a schematic? with voltages you are getting ?

jr_tech has a point about matching the source and load.... look at this:

http://en.wikipedia.org/wiki/Impedance_matching

What you want is for the output device to "see" a speaker or earphone load, And for the next stage to see its natural input resistance, or "impedance"

The circuit you have, the output device will be very unhappy whenever the volume control gets turned down under 1000 ohms. (just because I have seen headsets in the range of 1K to 8 ohms) The input of the tube circuit would like to see around 50Kohms or more I would imagine....

I found the circuit, you should try to get it working with the volume control at mid point, so it don't cut the tube off....

Remember maximum power transfer will only be 50% not that you will need that......

Another look.
 

Jr,
Regarding the wiring of the 47K resistor (R1), I don't see what you see. One end is grounded; the other end goes to tube pin 4 and C5 (68pF). Here's an updated photo of my build and the original schematic:

https://lh6.googleusercontent.com/-v...0/P1110216.JPG

https://lh4.googleusercontent.com/-p...07%2520200.jpg

(Note to all: This is a tried and true design. See the whole story at Phil's Old Radios. This will take you to the AM Transmitter article: http://antiqueradio.org/transmitter.htm
If I knew what I was doing I'd be tempted to improve on this design, but I need to stick with the original which others have built successfully.)

And here are the voltages I get:

https://lh5.googleusercontent.com/-x...07%2520103.jpg

This is as far as I got with the project two years ago. I looked at the old notes, and the voltages were basically the same as they are now. So, do these values or the wiring seem true to the schematic and the expected function of the device?

To Username1: Thanks for the link to the impedance-matching article. I have not yet read it, but I see its relevance and will get to it as soon as I can.

Thanks for posting the improved picture... In the old picture, it looked like the green wire under the resistor (R-1) was connected to the lugs at each end of R-1, this is a much better pix, and at first glance the wiring appears to be correct, but I will print it and study more closely.
Have you tried swapping the wires to #1 & #2 of the coil (or # 3 & #4, but not both windings) to get oscillation? Phase of the coil matters, and perhaps these coils are reversed?

jr

In my previous (two years ago) efforts, I tried switching the coils--someone had suggested the primary & secondary might be reversed on the schematic. I tried that without success, and errors in this long-used schematic are unlikely.

At this point, I haven't even tried the transmitter. I had been asked previously, "Does is oscillate?" I assume that means do I receive any kind of audible signal from the transmitter, and the answer is, "No, I have no audible signal."

Here's what I'm going to try, and I'm explaining this so any ridiculous errors will come out. I'm going to turn on my 1931 Airline and my Atwater Kent model 37, tune both to a quiet spot in the expected frequency range, and turn up the volume. I can receive local stations on those sets without an external antenna.

Then I'll plug the transmitter into a CD player with the volume turned up high, and I'll set the transmitter's potentiometer at mid-range. The receivers are 15 feet away from the transmitter. Phil's instructions say to DISCONNECT the transmitter antenna for this trial.

Here's a question that may seem silly, but it's important. When tuning the coil, where should the graphite core be in relation to the windings? Fully within the windings?

I try that the way it's set up now. If it doesn't work, I'll try reversing the coils.
Thanks for the help. I'll let you know what happens by tomorrow afternoon.

I guess that I would start with the core nearly falling out of the top of the coil and adjust in from there until it nearly bottoms out... cover the entire range.
With the transmitter antenna disconnected, trying to get a signal to a couple of vintage sets 15 ft away might be pushing the distance limit somewhat... do you have a nice sensitive little portable radio that you can put *very* close to the transmitter?

jr

Pin 5 on the tube is the screen grid. It shouldn't be such a low voltage; I think 100V or so would be better. I am not looking at the tube data so perhaps I am in error.

In order to get plate current in a normal pentode, the screen voltage needs to be positive, not as high as the plate. If you increase the screen voltage by a series resistor to B+, be sure to decouple the audio from the pot with a coupling capacitor. And bypass the screen for RF with perhaps a 220 pF capacitor.

Indeed the screen grid connection and voltage is suspect, but in the construction article, Phil explains that it is a "feature" not a mistake.
I am scratching my head over that one, however.
jr

I don't think this is a good way to operate the tube. Raise the screen voltage and then modulate it with a capacitively coupled audio signal.

Maybe I will take a look at the tube characteristics to see if this is feasible.

Okay, typical operation shows 105 V both plate and screen, resulting in 43 mA plate current, 4 mA screen current and about -5V control grid bias.

The tube simply will not work with such a low screen voltage, I firmly believe.

Estimating 4 mA screen current, add about 10k or so from screen to B+, then capacitively couple the audio to the screen. This will require a substantial audio signal, perhaps 20V, not easily obtained from every audio generator and certainly not from a microphone or phono pickup. So a 'speech amplifier' is needed.

A comment from the earlier thread about the very unusual operating conditions of this circuit:
http://www.videokarma.org/showthread.php?t=251917
Still scratching my head...
jr

I don't know enough to scratch my head.
 

Sorry I can't add much to this discussion, but it's way over my head. I spent some time checking the pin 5 voltage. With no connection to the audio source, we get 0 to -.8V.
When I plug in the earphone jack with the audio power off, we get -5V. With the power on, we get +5V. If I leave the meter on pin 5, the + or - 5V bleeds down, losing about 1V per minute. I put in a spare 117L7 and got the same results.

Tried tuning the coil with a 4-ft antenna set up 15 ft from the receiver, but received nothing at 900 KHz. I have a couple of table radios with loop antennas, but they're too noisy between stations. So, If the radio won't come to the transmitter, then the transmitter will go to the radio.

https://lh5.googleusercontent.com/-v...7l%2520105.JPG

Alas, this didn't work either. We're being visited by our grandsons for the next couple of weeks, so I won't have much time to play with this. On the other hand, at 7 and 10 years old, they're both a lot smarter than me, so maybe they can figure it out.

Try my idea of adding a resistor from the screen to B+ and I think you have a good chance to make it work. It's obviously not oscillating because the tube has no gain without screen voltage. Of course, with a big enough audio signal there may be parts of the cycle where the screen voltage climbs enough to start oscillation but in any case you will have terrible distortion.

I have a circuit for you to try. I built this in 1968,and had marvelous results. Circuit #13. I'd still have it, but a friend lent me one of those battery eliminators that looked like a 9V battery with a power cord attached. It was hot to the power line and non-polarized at the AC plug. I had the circuit grounded, so there was a big bang when I plugged it in with hot on the wrong side!

http://cdn.ka6wke.net/AmateurRadio/S...e_Hobbyist.pdf

My idea for such a circuit would be to put the screen at B+ voltage
and use the grid and cathode for the oscillator (screen as plate
of oscillator, grounded-plate type circuit) Then modulate from
a low impedance source (i.e. speaker impedance) using a
plate transformer in reverse to plate modulate the tube. Gain
would be adjusted using a say 25 ohm wirewound pot before
the primary of the modulation transformer.

the basic idea is at

http://nmwilliam.tripod.com/el84.html

though this uses a crystal rather than a coil. However ...
that circuit generates several watts and needs watts of audio.

Is it time to throw out the bathwater and the baby?

I don't think so... before giving up and embarking on a re-design or totally different circuit, a couple of questions perhaps should be answered.

1.As designed can the transmitter work?

a. There are several arguments here that indicate that the tube possibly could operate (perhaps poorly) outside of its normal operating parameters.

b. Calculating from the voltage drop across the 10k resistor to the plate of the 117L7 (pin 5) indicates that the tube is drawing about 2 ma current which is about 300 mW of power. This would seem to be consistent with a typical power output of 100mW that is generally considered
permissible for un-licensed operation on the AM band if a very short antenna is used. More power might incur a visit from the FCC. A good reason, perhaps for this low-power configuration.

c. Phil says it works... there is no doubt that what he built indeed does work fairly well!

2.Does it work?

a. Perhaps, but not on the frequency tested... you might want to test at the high end and low end of the broadcast band just to make sure that it is not transmitting.

b. I still favour testing with a sensitive little portable held near the transmitter. Sony for example, sells a sensitive little analog portable for around 12 bucks that would be ideal for this testing. http://www.amazon.com/Sony-ICF-S10MK...y+pocket+radio

3.If not, can it be made to work?

a. The circuit is pretty simple, and appears to be wired correctly... a big question involves the coil phase and/identification of windings... perhaps the primary and secondary are swapped, or one winding is simply out of phase with the other.

b. I suggest disconnecting all 4 leads to the coil and measuring the resistance between all the pins... perhaps we can figure out what is happening here.

OBTW, your test radios are very cool!
Not affil with Sony,
jr

Hi All;
I have a copy of what I think is the origional article for this Transmitter, it is copied in jpeg, but I don't know how to make it show up here..
The Schematic is slightly different, but maybe it may help you find what the problem is with your unit.. It was origionally for a phonograph with a crystal pickup..


http://www.videokarma.org/attachment...1&d=1376186080

THANK YOU Marty

Hi All;
Winky, I have this file in PDF format, but the file is too large for me to upload here..
If you want I can email it to where ever you want If you cannot see the details or fine print..
THANK YOU marty

Yes R2 is the critical component missing from your unit. It's from the screen to B+.

speaking of AM transmitter circuits, I have some at my web page http://www.wa2ise.com/radios/amxmit.html, http://www.wa2ise.com/amtxmit.jpg and http://www.wa2ise.com/amxmit.gif

Sorta works !
 

For grins, yesterday afternoon I dug most of the parts out of my junk box to build a simple version of the Li'l 7. I do not have the proper oscillator coil, but found small antenna coil (likely out of a junked AM radio) that had 2 separate windings. The windings both had fairly low resistance (under 10 ohms), but I decided to give it a try. I did not use a power transformer for isolation, but made sure that ground on the board was always cold. I could not find a 1 meg pot, so I substituted a 1.2 meg resistor to ground at the input (pin 5).

Since the coil was not identified in any way, I connected the lower resistance winding (untuned) to the plate (pin 3) and B+ and used the higher resistance winding (tuned with a 270pf cap) in the feedback circuit (pin 4 and ground). A 6.3 VAC 60Hz "signal" from a bench power supply was connected to the input (pin 5) and ground.

I had to reverse the two leads in the feedback circuit to get oscillation, but when I did, I was able to find a loud buzz with a portable radio very near the transmitter. Since my "oscillator coil" does not have an adjustable slug for tuning, I merely scanned the band with the portable radio... found the buzz "signal" at about 700kHz. The range was only about 5 feet with no antenna connected to the transmitter.

Now the "sorta" part... when the signal was removed, it appeared that the transmitter stopped transmitting! In fact, there seemed to be a threshold... if the signal was reduced below a certain point oscillation stopped. I suspect that a real audio signal would be transmitted, but with much distortion, as bob91343 suggested. :thumbsdn:

Perhaps with the correct oscillator coil, the circuit *might* work ok, I don't know... today if I get a chance, I will try to "bias up" the G2 connection to a positive source to see if I can get undistorted modulation. As can be seen in the 'scope photo, alternate parts of the 60Hz test signal (likely the negative portions) are missing from the modulated output of the transmitter.

jr

Hi All;
Jr tech, If you put a tuning cap in parallel with your coil, you could tune it to where ever you want..
That is in the schematic I posted, He had a tuning cap in the circuit.. He states in the article that an antenna coil or an oscillator coil will work, with a tuning cap..
THANK YOU Marty

For sure... but I used all the 365pf tuning caps that I had when I built a number of tuned loop antennas for myself and several friends... still digging in "junk boxes", might find something suitable. :)
"Fixed tuned" worked well enough for my purpose of testing the operation of the circuit.

jr

Update: I "biased up" the G2 somewhat by connecting it through a 470k resistor to the B+ supply. G2 now sets at about + 25 volts. The modulated output is no longer missing the negative information from the modulating signal and the transmitter continues to oscillate when the modulating signal is removed, but it is not very symmetrical (sorry 'bout the fuzzy handheld pix). The tube is now drawing more current and the B+ supply has dropped to about 110 VDC. Obviously, the 10k resistor in the power supply will have to be reduced, and likely more filtering will be needed to handle the increased current load. More work is needed, but out of time for today.

Referring to the posted original article above, I think R2, 3.5 meg, from B+ to the screen grid must be large enough to allow modulation
with a very high impedance source. With a lower impedance source, [headphone output from a CD player, etc.], it could be much smaller.
Hope this helps.

The Wayback Machine
 

(for some reason I can't embed my photos, so I have to use old-fashioned links)

After several frustrating attempts to get a Li'l 7 AM transmitter to work, I decided to try something else. After reading Jay Allen's (radiojayallen.com) description of the product, I bought the SStran AMT3000. I'd never worked with a circuit board, and it was rough going at first...

https://picasaweb.google.com/lh/phot...eat=directlink

...but I managed to finish the build in two days...

https://picasaweb.google.com/lh/phot...eat=directlink

...and, son-of-a-gun, it works beautifully. Good reception throughout the house, and now I know there's always something good to listen to on 1530 AM radio. And now, when the grandchildren visit this summer, we can enjoy an authentic experience listening to old-time radio shows.

The name "Wayback Machine" refers to Sherman and Mr. Peabody on Rocky and Bullwinkle, a favorite TV show from my youth.

https://picasaweb.google.com/lh/phot...eat=directlink

With the experience of building this transmitter and with all the suggestions above in this thread, I'm going back to the Li'l 7 once again.

I'm glad you're happy with it. I've had my 3000 running 24/7 for months with no issues :)