Demetre SV1UY wrote: > Hi Jose, > > Happy New Year to you and your family.
Happy New Year to you and yours, too (also, to the readers of this list). > As for the early KAMs you are right, but after a while they brought > out new firmware and they fixed the problem. I have an early KAM with > a special addon PCB so that it can take PACTOR 1 modeand I followed > all the firmware upgrades up to 8.1 I think. It is now in the basement > somewhere so it is not handy for me to check. But as I said before it > was always a lousy PACTOR controller (probably it had a bad modem > design because even in HF packet it performed badly. I did never own a KAM, but a friend of mine owned one of the early ones. He always used it on RTTY, but I began experimenting with that "packet thing" when I visited him then... Packet on HF...required a LOT of patience. Seems I achieved it (I know quite a few that did not...), as I spent some 5 years of HF packet sysop, and then, some other six or seven in pactor. In packet, a single erroneous bit trashes the frame. Fading, sparks, collissions, all of that made it too easy to generate a retry. For some time, I ran my homebrew linear (about 400 watts out) to keep the link to the US. Really, not affordable, it cooked a final tank that was quite OK for SSB or CW, but not for packet. I had to rebuild that pi-network. I could do the same in pactor2 with only 25 watts, not only to the neighborhood, but also to Africa. So, actually, the pactor 2 and 3 modulation schemes are good for low powers. Where is the key to it? The protocol. Using ARQ plus FEC (convolutional code), data interleaving and block codes allows to recover frames that packet layer one would lose. It is similar to what CD's and digital broadcasting uses nowadays. In retrospective, packet radio layer one belongs to the dark ages. Could it be changed? Yes, AX.25 specification only deals with layers 2 and 3, and Q15X25 did it with some success. But in general, manufacturers did not innovate on this. I was not really aware of that back then, either. What is missing on this scheme: bandwidth/speed negotiation, like pactor does to survive bad links. SCAMP failure is associated with its unability to negotiate the link. > So in the end I had to buy an SCS Controller because as you know it is > superior in PACTOR and in PACKET RADIO. I have never got any addons to my SCS PTC-II. And the newer robust packet adittion also requires a RAM addition to 2 MB. I just have loaded the tiny38.pt2 firmware upgrade and it still works quite OK. That is another example I did not mention: robust packet, using PSK instead of FSK. I don't know in detail the tricks they added to robust packet, but it would be interesting to dig and see (if that could be possible) what they did. But certainly, data modes require some coding tricks to survive the HF hostile environment (Olivia success is based on the Walsh code layer it uses), as has become usual nowadays for data transfers (keyboarding is something with a different twist, the simpler the better). It was a mixture of sheer good luck and naiveness to get a raw Bell 103 modem to work on the lower HF bands. Maybe Kantronics Golay COULD have been better, but 300 baud is generally too much. And it never really became popular, with each manufacturer having its own "pet project", that did not achieve the numbers required to have an impact on the community. PA0R comments about PSK speeds in PSKMail seem to agree with what is well known: PSK63 works, PSK125 somehow, but nowadays PSK250 has only a 60% success. On 10 meters and using a single propagated ray (as usually happens close to the MUF), I would not be surprised to see that PSK1200 (or QPSK1200) would work as well. > 73 de Demetre SV1UY 73, Jose, CO2JA __________________________________________ Participe en Universidad 2008. 11 al 15 de febrero del 2008. Palacio de las Convenciones, Ciudad de la Habana, Cuba http://www.universidad2008.cu
