Tuning a small DIY am transmitter

[Alex_Baker]

[ Specified attachment is not available ]Hello Everybody

I have been on the forum a bit, just usually in the repair section. This topic is about a home built broadcast AM transmitter that needs some tuning  , and I figure that the beginners section would be better than the repair section. I know little about how AM transmitters work, so part of the purpose of this project was to learn a few basic things about how they work, but my other goal is to have a device that can transmit music to my old tube radios.

I found the original design on this page; http://techlib.com/electronics/amxmit.htm . I liked this one because it did not use any special components or transformers, and the designer gave lots of details and tuning tips. With that being said there are still a few unknowns that I was hoping to get some input on, I will mention them later.

Here are the details about the transmitter, My schematic is attached below.

VCC is regulated 15V
The crystal is 1MHz(rather hard to find)
I have a 36 inch extendable metal antenna attached to it
The antenna loading coil (not shown on schematic) is a tunable inductor that I can tune between 1700uH and about 5000uH (I think).

As it sits right now, the transmitter is working and it sounds pretty good, however it is incredibly weak. The signal all but disappears more than 6 feet away from the transmitter, and that's with nothing in between the TX and the radio! I have tried some tuning steps, recommended by the designer on the webpage. First I tuned the tank by adjusting L for the highest output with no antenna attached, but they also talk about antenna loading a bit, which I understand little about. I understand that antenna length and frequency need to be considered for choosing the inductance of the load coil (and  location), but beyond that I don't really how it works. The guide says that when attaching the antenna and coil the meter reading should drop significantly, I can't reproduce this effect.

Here are my main questions: 1) is there a good way to verify that the crystal is working as it should?
                                           2) what peak to peak voltage should I expect on the antenna output with nothing attached? currently it is about 10V pk-pk.
                                           3) How does the antenna loading coil work, and what is the best type of coil to use for it?

I hope this was not too much to look at, I appreciate any input 

[iMo]

1MHz TX frequency would require an antenna of around 150m length (lambda half) or a half of it (a quarter of lambda). Also you would need a different amplifier with an antenna tuner to match the wire properly. Btw, 1MHz is off the amateur licensed bands (provided you are a licensed ham).. 

[Transconductance]

Tuning should be simple.  First, tune the tank inductor "L" to maximize the meter reading.  Then tune the antenna loading coil to minimize the reading.  Think of it as maximizing the amount of power going into the antenna so minimizing the waste power going into the meter.  What you're really doing is minimizing the energy reflected from the mismatch.

You can also use your receiver for antenna tuning.  If your receiver doesn't have an S meter, then move it farther away or otherwise attenuate the receiver input until you have some static hiss in the audio.  Then tune the loading coil on the transmitter to minimize the hiss.

The author gives you a few values you can verify.  For example, the current in the differential pair is supposed to be 20 mA.  Therefore, you should see about 2 volts across the 100 ohm resistor in the modulator.  Go through the article and find these hints.  Verify them in your build.

You should post a photo of your circuit.  At radio frequencies, even this low, construction techniques matter.  Note the construction of the prototype in the photo on the web page.

A proper antenna will make a big difference.  If you have room for a 75 meter long-wire, you could get past the need for the loading coil.

Do you have an oscilloscope?

The way to debug the transmitter is to look at the signal at the output of each stage.  In this case, there are two stages.  The output of the oscillator is on the 0.1uF capacitor closest to the left side of the schematic.  The output of the PA is at the collector of the right-most transistor.
(No reference designators - WTF?)

[Transconductance]

OK, I did a better job of reading your post.  10V P-P on the TX output makes me think your problem is output mismatch.  So, focus on the loading coil tuning.  Even if you don't have room for a proper antenna, a longer wire will help.

[Alex_Baker]

Thanks everybody for the tips so far.

First I will say that transmitting on the AM broadcast band in the US is legal without a license provided that the TX is less than 100mW and the antenna is not longer than 3m or about 9.8ft. Since I do not have a license these are the limits I am working within.

OK, I did a better job of reading your post.

 
Sorry, It was an intimidating mess of a first post.

I have been able to tune the tank inductor no problem, but regardless or what I put on the antenna output, the meter reading wont drop, short of grounding it of course. I have also verified that the differential pair is passing 20mA, and its spot on. I need to do some more testing with antennas and loading coils, but like I mentioned earlier I (legally) cant be over 3 meters antenna length. I suppose I will  just start at the maximum allowed.

Pictures are attached below, along with the board layout. They are large files so I may need to attach them in multiple posts.

[radiolistener]

Your coil near antenna looks like bad. You're needs winding in one direction with the same gap between the turns. Avoid winding overlaps.

Also metal bolt inserted into the coil affects coil properties, are you sure that this coil needs metal bolt inside? material properties of the metal bolt also matters.

[Alex_Baker]

Ah, that makes sense. 

The coil is something that I had laying around salvaged from an old radio, so I presume that the metal slug in the middle was chosen specifically for its properties.

I will see what I can come up with to get 1000uH or so...

[MathWizard]

Why are R22,23 so small, in the audio amp, plus considering it's running from 15V too, isn't that an oversight, and you get a plenty stable reference with way bigger divider resistor's

That crystal osc. also works with an inductor and a blocking cap, instead of the crystal, without any fiddling in LTSpice.

I think the diff-pair + current sink, with RF and AF mixed this way, is called a balanced modulator, something like that.