any ideas?

[Old_Bill]

The attached .pdf schematic is of a simple? low power AM transmitter that I am trying to get to work. 

It powers up all right, and will produce a very weak carrier on frequency.  The oscillator and first tripler tune as expected to a definite peak, as does the second tripler.  The problem is in the PA, with its rather unlikely tube, a 12BH7.  When the PA is in the circuit it provides a very small amount of amplification and does function as intended, as a doubler.  However, the tank and load do not tune- there is no dip of current or peak of power and the RF monitor barely moves off the pin and makes no change as it is being tuned.  The coils are wound exactly to specification, and the tank is resonant at the proper frequency- but almost no RF is being generated.  B+ is in the mid 300's and the tube is brand new.

I found an error in the original schematic and am wondering if there is another obvious design typographical error I have copied.

Thanks for any ideas.

[KM1H]

I wouldnt even attempt a guess without having a GDO available to both check the coils and sniff for RF in diode mode.

It could be not enough signal to overide the bias, soft tubes or any number of things.

Carl
KM1H

[Old_Bill]

Thanks for the response.  Let me restate the question a bit- if this thing were built according to the schematic, and the tuned circuits were within operating range, would it likely work.  Are there any obvious missing or wrong connections.  The print I am working with already had some mistakes I had to find the hard way.   The plans are over 50 years old and there is no one to go back and check with.

Also, could anyone say, if the first stage were converted to a doubler instead of a tripler, and the PA were made a push-pull tripler instead of a push-push doubler, retuning all the coils of course, would that be any better design.

Thank you all for reading and for any ideas or advice you can provide. 

[AB2EZ]

The schematic looks ok up to the point where you have modulated B+:

The "Second Tripler" consists of two triodes (each half of a 12AT7) in parallel, and fed out of phase by the prior stage. The pair of third harmonics that each of these parallel triodes produces will also be out of phase (that's good).

The outputs from the Second tripler feed the two triodes in the PA/doubler. The inputs to the PA doubler will be out of phase, but the outputs will be in phase (that's good)... and the outputs are added together.

A problem is that the grid drive signal needed to put the 12BH7 far enough toward cutoff to act as a doubler is pretty large... particularly when the modulated B+ is at its highest value (modulation peak)

So... you need to check to see if your actual physical implementation of the schematic is working properly. That can be tricky because the impedance of a typical scope probe (15 pF) will disturb measurements you try to make at the grids of the PA/Doubler tubes.

As stated in the earier post... check to see if the pair of signals that drive the grids of the pair of output/doubler stage triodes are large enough to self-bias those triodes into cutoff (to obtain the doubling effect). Assuming that the peak of the modulated B+ is something like 400 volts, you need a peak grid voltage of around -25 volts at the 72.846 kHz frequency. Those two signals should also be approximately 180 degrees out of phase, and approximately equal in amplitude.

If the signals on the grids of the PA/doubler triodes are ok... I would check to see if the rf load impedance on the triodes is large enough. Since you are not seeing a "dip" in the plate current, and the output (at 144MHz) is very low... that would suggest that your output link-coupled tank circuit is not placing a high enough RF load (at resonance) on the PA/doubler tubes.

In my opinion, using the 2nd stage as a doubler and the third stage as a tripler would make things worse. Furthermore, you would have to change the schematic to get the phases right. I.e., if the second stage were used as a doubler, then you would need to place the outputs of the second stage tubes in parallel, instead of push pull.

Good luck

Stu  

 

[KM1H]

Scope loading at those frequencies is why I didnt suggest it. This is where old tech works better unless you have a real modern lab full of gear.

When I get myself into a corner the old Measurements Corp Model 59 is often the first and only thing I need to grab.

Carl
KM1H

[Old_Bill]

All.

Thanks for your excellent and detailed analysis- this is exactly what I was looking for.  This has turned into quite a learning experience.  So if I understand correctly, the schematic is likely correct as is, and the most likely explanation for the lackluster PA response would be insufficient drive to the PA final.  With that in mind, I will go back and re-check the coils and stages for resonance at their assigned frequency.  Maybe I can get a better peak.  By the way, the alternative construction I was thinking of was to make the first stage an oscillator-doubler, the second remains a tripler, and the final becomes a push-pull tripler.  However, excellent point about having to feed those stages either in and out of phase to make it work.  I'll keep tinkering and see if I can get it to do anything.

[Old_Bill]

Having made little progress with the the 2M transmitter, and giving the matter much thought, it seems to make a lot more sense to stay down on 40/80 rather than 2.  So now, the project has been reduced to parts, and I am looking for any good schematics of tube-type AM transceivers for 40/80, and of course, any suggestions as to good AM crystal frequencies on either/both bands to start with.  This will be about 5 watts or so and portable, in the SF Bay area.  I had pretty good luck building a WWII Paraset (CW-40/80) and that regen receiver worked out great for AM.  So it's really the transmitter that we need the design for.

Thanks for all your ideas and suggestions.