I have built a basic transmitter on an ugly dead-bug board to investigate the properties of part 15 AM and otherwise learn about it before starting a station (with an SSTran most likely).
I have built a basic transmitter on an ugly dead-bug board to investigate the properties of part 15 AM and otherwise learn about it before starting a station (with an SSTran most likely).
In my transmitter there is an LC Colpitts oscillator generating the carrier frequency (~1600 kHz). This passes through a buffer amplifier stage and then on to the final amp/modulator stage. I am using the series modulation technique (two transistors, one is fed the audio into the base and the other fed the carrier into the base). Plain old 2n4401 transistors are used througout.
Everything works, but my issue is with antenna matching. I have no idea of the output impedance of the final stage to build any kind of matching network, or filter to cut out the fairly loud harmonics. I've read about it being based upon resistors (of which I have none in the final except to bias the base of the first transistor), characteristics of the transistors, etc. but have not yet found a clear answer.
Here is a schematic of pretty much what I have. I would like to know how to figure out the impedance of the "To Ant." connection.
Also I understand that loading coils are used to cancel out the capacitance of the antenna, so that it shows up as a pure resistance to the transmitter. Given that my 9ft copper pole radiation resistance calculation showed about 0.4 ohms, it still seems that it would be necessary to match the the antenna with the transmitter - driving the whole antenna system, which may be 5-10 ohms or so total with a higher-impedance final stage doesn't seem like it would work well. Kind of like hooking an 8-ohm speaker directly to the output of a tube amp with no transformer...
Do I need a matching network beyond that of the loading coil (I think I do, but am not sure)?
Thanks.
Hello Rankinecycle
From your abbreviated circuit drawing I have no guess as to the impedance at this juncture, but output filters and matching that I've known are usually comprised of choke-coils and capacitors and not resistors.
There will be other comments in a few minutes.
Try Googling "Pi-Network filter". There's lot's of info regarding designing the popular filter out there.
It serves as a harmonic filter and impedance matching network.
Here's a link to an application note which I used to design several 2 meter amplifiers. At VHF the calculations were close enough to reality that very little tweaking was needed. The advantage I had was that I knew the output Z of the power transistor from the data sheet.
In Section 3 there is an equation to estimate the load resistance from the transistor operating parameters. This may get you into the ballpark.
I have had good results using a pi network with two C's and one L. This gives good harmonic suppression and can match a wide range of Z's.
From the schematic I cannot tell if this is running class C. If so you might consider replacing L5 with a parallel LC resonant circuit and take the output through a link loop.
Keep us up to date on your progress.
Neil
Good Morning, The first thing I see is the collector directly tied to the antenna. I would recommend a capacitor in series there. Also the 2n4401 likes to see a high impedance load if I am correct. Here is a link that may help:
http://www.aoc.nrao.edu/~pharden/hobby/HG_DS1.pdf