This has happened three times now on three different nights and three different contacts. Here's how one instance went: I was calling CQ on 40m PSK31 and Fred, K5QBX was trying to reply. His signal looked sort of okay but just wasn't enough to read. It took two attempts before I even got his callsign. It was enough for a contact but not enough for a decent conversation. It was late and the band was empty so he switched in the amp and dialed it up to 500 W PEP. His signal was still perfectly clean on the waterfall and IMD was -21 dB, and now I had him at about 98 percent. So then I suggested, just for fun, switch to PSK63 and drop back down to barefoot. He did, and guess what? I had him at 95 to 98 percent. He had never tried PSK63 before and was very pleased. We had a long chat.
I've observed this same phenomenon two other times where on a weak contact on a noisy and weak 40m band at night, PSK63 provides better readability than PSK31. Of course, this makes no sense. PSK63 should be 3 dB poorer than PSK31 and yet I observe the opposite and then some.
I am at a loss to explain this effect. Perhaps the noise or path characteristics on 40m are such that PSK63 modulation is harmed less than PSK31? I'd be interested to hear any theories about this and maybe someone can duplicate my results.
73 de KW2P
3 comments:
Hello Phil,
Just for funnies today I put my call into Google to see what would come
up...much to my surprise I found your blog about our PSK31/63 QSO. For
your further information, I was running 50 watts when we first tried to
hook up that evening, so for good print, 10db gain was applied (500
watts). I was pleasantly surprised that PSK63 worked better barefoot
and was much faster and like you, am wondering if that was an anomaly.
If you'd like to try to duplicate that experience, we can try for a
PSK31 sked on Thursday evening (Jan. 11) at 0200 UTC (2100 EST or 2000
CST of Jan. 10 local time) using 7.070.150 indicated on the radio and
700 hz on the waterfall which is actually 7.070.850 Mhz (I think). At
any rate, I'll have 7.070.15 indicated on the rig and have the
"spectrum" view going so that I'll see across about 2500 hz of
spectrum. If you have a better way of explaining that, please let me
know. I'll also have the amp warmed up...
If this time isn't good for you, please offer a different time. I
picked that time as I have obligations both Tuesday and Wednesday
evenings. Looking forward to chatting with you again...73.
Fred
K5QBX
Hello Fred,
It's very nice how the Internet works and that you found this topic in a search. I'm a little surprised that this post got indexed so quickly so you were able to find it.
Yes I would like to have another go at it and the time you chose is fine. I'll set an alarm to remind me five minutes before.
PSK63 is very noticeably different on the waterfall so we won't have trouble finding each other. The method you used for quoting frequencies seems pretty standard for digital modes: dial frequency plus audio frequency. 7070.0 KHz dial plus 1500 Hz audio would put us at 7071.5 KHz actual center frequency.
My radio's audio passband starts rolling off below 1200 Hz. The sweet spot is between 1500 and 2300 Hz so I just compensate by lowering the dial frequency. For digital operation, I set the VFO to go in 1 kHz steps. It makes the calculation easy. Usually when operating PSK, I set the radio to 7069.0 to bring most of the action into the "sweet spot" of my radio's passband. The ideal spot on my radio, and where I try to put a weak signal, is at 1900 Hz because it's in the middle of the passband and I can also use the IF shift to eliminate all interfering signals above or below 1900 Hz. Now if only I had two IF shift controls--one that came in from the top of the passband and one that came in from the bottom, life would be perfect!
I look forward to a QSO on Thursday.
73 de KW2P
Fred,
Well we were foiled by absolutely terrible propagation on 40m last night. Once in a while I could see a faint signal on the waterfall but no copy. I doubt that you heard my calls either. We'll just reschedule and try again.
73 de Phil / KW2P
Post a Comment