Electronics > RF, Microwave, Ham Radio
Homebrew transmitter questions
cncjerry:
So having all this great test equipment, I decided to put it to use for once.
About 10yrs or so ago, I friend sent me a KVG XF-9B 9Mhz filter and two crystals (LSB/USB) and said, "have fun." So I built a transmitter using mostly the schematic in "Solid State Design for the Radio Amateur" beginning on page 202 incorporating some other more robust circuitry and used an FET for the output. I followed their modulator and IF design but changed the BFO, mic amp, bandpass filter (post IF) and for a VFO I will be using an AD9851 or AD9854. Right now for the VFO I am driving it with my HP 8648C synthesized signal generator.
When I built it 10yrs ago, I had trouble with BFO stability but mostly I didn't have a spectrum analyzer and frequency counter so getting the BFO set correctly became such a hassle I threw it all in a box until last week.
With this background, I am wondering what level of performance I should be seeing. The carrier is down 73db and the opposite sideband down 59.4db. That's the first question, is this decent for the filter? good enough? I don't have everything shielded so that will improve things a little.
The second issue I see is that the harmonic distortion is higher than expected. The 2nd harmonic within the sideband is only down 45.7db and you can't see it on this plot, but the 3rd is only down 33.1db. So looking at the plot below, the larger signal is the 1k tone being fed from my signal generator. To the left is the 2nd harmonic of the tone and then to the right, the carrier (in the center can't be seen) followed by the opposite sideband. I can tweak things using the carrier balance, IF gain and audio drive but the harmonic distortion is about as good as it gets. I assume some is coming from my audio generator, an FG5010 in a TM5006 frame, not known to be the best.
I'm going to hook up the power amp chain and see if I can make some contacts with it.
BTW, The second plot is somewhat interesting as I took the time to set the RBW/VBW on my 8568B to 10hz/1hz respectively. I was trying to find the carrier. Note the intermodulation around the 1Khz signal that you couldn't see in the prior plot.
uncle_bob:
Hi
That all looks believable. With some care you can do better. There are many examples of people doing worse. The normal approach to improving things is to look at it stage by stage and get each one working the best you can. One example: You can sweep the filter if you analyzer has a tracking generator. If you don't have a tracking generator it can be done manually with a signal generator. That sort of test will let you vary the input and output match on the filter and optimize it.
Bob
German_EE:
Some things that can help here:
1) The transmitter audio stages and the RF modulator need to be as good as possible. Any mistakes here will show as poor transmit IMD before you hit the filter but it's also a problem with the crowd who insist on turning microphone gain up all the way.
2) For transmit tests drive your audio with a clean sine wave audio signal, distortion in this waveform will cause poor IMD again.
3) I don't know what mixer and modulator you are using but if they're the little MiniCircuits mixer modules then these MUST be terminated at 50 ohms on all three ports, anything else and performance will suffer.
4) Look up the specification of the crystal filter you're using and make sure that it's correctly terminated at both the input and the output. Ground the filter case.
cncjerry:
The balanced modulator is a 1496. I'm going to take another look at the incoming 1Khz tone I am using and rework the mic amp today to see if I can clean it up.
The filter case is grounded and terminated correctly, at least consistent with other designs. I have the carrier oscillator/BFO dialed in now pretty closely. The IF gain on another 1496 seems to change the harmonic distortion somewhat balanced against carrier level and opposite sideband so I'll take another look in there as well.
This filter and mixer setup was used in a lot of transmitters years ago, I think ten-tec used them in some models. I've not been able to find a plot as to how well they performed.
Thanks for the pointers.
uncle_bob:
--- Quote from: cncjerry on June 30, 2016, 04:36:02 pm ---The balanced modulator is a 1496. I'm going to take another look at the incoming 1Khz tone I am using and rework the mic amp today to see if I can clean it up.
The filter case is grounded and terminated correctly, at least consistent with other designs. I have the carrier oscillator/BFO dialed in now pretty closely. The IF gain on another 1496 seems to change the harmonic distortion somewhat balanced against carrier level and opposite sideband so I'll take another look in there as well.
This filter and mixer setup was used in a lot of transmitters years ago, I think ten-tec used them in some models. I've not been able to find a plot as to how well they performed.
Thanks for the pointers.
--- End quote ---
Hi
1) Even with "pro gear" a tweak of the actual terminations on the real pcb layout is a standard part of an IF strip design. Things like coil Q are often the issue.
2) The 1496 is a very old part. There have been a *lot* of chips introduced since it came out. That's not to say it's junk, only that there are other chips that migh do a bit better.
Bob
Navigation
[0] Message Index
[#] Next page
Go to full version