As often happens, I may have jumped the gun in declaring the exorcism of my 17 meter transmitter to be a success. As readers of this blog will recall, my problem was that when trying to "net" my separate 17 MHz receiver and transmitter, at around 18.116 MHz I could hear more than one tone as I tried to get to zero beat. The 8th harmonic of my 5.176 MHz carrier oscillator was mixing with the 23 MHz VXO signal and producing a spur. I could probably knock the level of this spur down below FCC limits, but -- and here is the problem -- I probably could never knock it down to the point that it would not be audible in the sensitive receiver that sits right next to the transmitter. So this is really a netting problem, not really a spur problem.
I don't want to try another filter frequency -- I have VXO crystals that really work only with a filter at 5.176 MHz.
So here is my current idea: Build a receiver board and turn this thing into a transceiver. Switch with relays the input and output of the 5.176 MHz filter, and use relays to switch to the receiver board the VXO and carrier oscillator signals.
Making this thing a transceiver would eliminate the need for netting. This should solve my problem.
What do you folks think?
73 Bill
Posts
...why not just reduce rx sensitivity during netting?
ReplyDeleteHenk/ PA0EME
Good idea Henk. Assuming spur is 40 db down from 30 dbm main signal,and that the RX has an MDS of around -110 dbm, if I reduce RX gain by 110 db, this should put the spur below the MDS while leaving the main signal audible. I will give this a try as soon as I can. 73 and thanks, Bill
DeleteAs much as we love to see the open guts of our rigs, would this problem be minimized by fully-enclosing the TX and RX? Maybe they usually are when you're using them, but your pics show them open. I'm a bit of a prude when it comes to exposing too much to the outside world.
ReplyDeleteHow about this half-a**ed idea: avoid relying on an *audible* zero-beat by installing a Sanjian frequency readout on the TX and RX. If you use the 8-digit version, you could them together to within a Hz or two. It's a higher-tech solution that would solve the lower-tech spur issue. That is, as long as you get the spur >43dB below the fundamental.
Oh man, you guys out in Portland just need to LIGHTEN UP! Al fresco is the way to go. But I hear you, and I did put the carrier osc in a box. It didn't help.
DeleteI had the same idea about the San Jian counters. Back on Wednesday of last week I ordered 3 of the 8 digit models. But now I doubt this will work. These counters often just take the IF freq and subtract it from what is being produced by the VXO. But they often can't handle negative numbers. So in my case the IF would be plugged in as 20 MHz. The VXO runs from 1.89 MHz to 1.832 MHz. 1.89-20 = -18.110 and 1.832-20=-18.168 If the counter can't display an absolute value, this won't work. I had this problem with San Jian's before: https://soldersmoke.blogspot.com/2021/02/a-problem-with-san-jian-plj6-led-counter.html
It would be cool to net visually instead of by ear. Maybe these counters will be able to handle negative numbers. Fingers crossed. 73 Bill
Let me just go on record as saying that *no one* in Portland needs to "lighten up." Quite the opposite: they need to get their act together. Also, I'm a Francophile, so I prefer *en plein-air* to *al fresco*.
ReplyDeleteTwo more thoughts: you've shielded the BFO, but have you shielded every place its signal goes? For instance, is the product detector shielded (including the wiring from the BFO)? I keep thinking about shielding because I've never recovered from my encounter with a TV-36 AM VHF transmitter built for the FAA by Collins (hybrid tube & solid state). Every amplifier stage, and every mesh of every filter, was isolated in its own enclosure or RF-tight segment of a larger enclosure. After trying to get it working, I parted it out: it had more than 200 feed-through capacitors feeding DC to its various amplifier stages and even relays. Its power supply included several 1200V oil-filled capacitors and three 10-Henry chokes the size of milk cartons. Nothing, but nothing, in the way of RF was getting anywhere it didn't belong. I'll be haunted by the specter of that transmitter until my key goes silent.
My other thought is that you could still use the Sanjian readouts even if the IF offsets were off. You'd still be able to visually match them for zero beat. --Todd K7TFC
I think people often go over the top on shielding. I was influenced by free and open hardware like the BITX20, the BITX40 module and the uBITX. No shielding in those rigs.
DeleteI may be missing something but I don't think I can use the San Jian's to net. TX is single conversion with the VXO sig for 18.110 = 23.286 18.168 = 23.344
But RX is Double Conversion with VFO
18.110 = 1.89 18.168 = 1.832
So how could I make the San Jian spit out numbers useful for netting?
Just kidding about Portland and lightening up!
73 Bill
"Just kidding about Portland and lightening up!" I know, I was kidding too.
ReplyDeleteI hear the ice starting to crack under my feet, but it might not be a Sanjian show stopper that the TX has an up-converted IF while the RX IF is down-converted. At the single IF port of the TX, or either IF ports of the RX you have access to both the sum or the difference frequencies. Suppose on the RX you picked off the sum and fed it to the readout (though a filter just good enough to cut the difference signal (6-10dB) so the Sanjian wouldn't be confused. Now you have a positive-going readout signal for both the TX and RX and no negative offsets. If that works, it might be less trouble than turning the TX into a transceiver.
Several years ago, there was a posting on one of the reflectors from someone looking for a visual indication of zero beat. I don't remember if he found one or not. I'm going to try to find it, and if I do I'll send it along to you. I guess it would have to be one that could distinguish between the carrier from the TX and offending spurs--by relative signal strength, perhaps? --Todd K7TFC
Todd: I too hear the ice cracking... but I may be missing something. I've tried several combinations of IF setting for the San Jian and can't come up with one that provides a useful readout for the receiver. Please let me know if you can figure out how to do this. Even if it doesn't display "18", the only think that really counts (!) here is the figures after the decimal point. Readout on the transmitter is easy. I have another idea: I could do with the receiver San Jian what I did with the counter for the Hodge Podge rig: https://soldersmoke.blogspot.com/2021/03/a-frequency-counter-for-hodgepodge.html
DeleteThat would work. 11 MHz crystal for the NE602. Then select the difference frequency and set the San Jian IF to ADD 9 MHz. Boom. That would be a lot easier than building a receiver board. And I could experience the joy of visual netting! What do you think? 73 Bill
Well, I might mess with it a bit. I want to have a few on-hand (and not in rigs), so I've ordered two more Sanjian readouts. I'll simulate your TX and RX using two channels of an Si5351. You'll see! Tennessee Tuxedo will *not* fail!
DeleteHi Bill,
ReplyDeleteHave you thought of netting with the aid of a magic eye tube?
73 de Peter VK2EMU
Peter: Good to hear from you. Hope all is well. I think a magic eye would get me into more trouble with the waterfall watchers who complain a lot if I am as much as 14 Hz "off frequency" (I actually had a guy tell me my rig was 14 Hz off. How would he know?) But thanks for the suggestion OM. 73 and SITS. Bill
Delete