hi all, Here is some info a friend of mine made to the ICOM DStar repeaters to make it perform better.
73, ron, n9ee/r The loss isn't the problem. It's that Icom is using a single shield cable, that's only 60% shielded, so the problem is desense -in- the box itself. What's the loss difference between the two? Almost none, they both have slow velocity factors, so the losses are almost identical. We monitored the discriminator and dummy loaded the TX, sig generated the RX and dropped almost 1dB of desense by just putting RG142 dual shielded cable in the chassis to the back panel. Now I did RG142, as that's what I also use as my interconnects, so I also add the 14" of cable we put inside to my cable lengths to my filtering so I can get my interconnects at 1/2 wavelength so I don't have any current nodes in the wrong place. I could have put grommets on the back panel, but the extra loss a pair of N-Male/N-Female Delta brand connectors wasn't worth it to "not" have disconnect on the back panel. The other problem is you -have- to use a preamp. The receiver is only .5uV sensitive! Also has horrible side channel rejection, so I have 4 cavities in series as a preselector with a 17dB PHEMT preamp (.4dB NF). The combination of that preamp, that cable, has gotten both NI4CE and KJ4BYI's repeater the best performance. Bill and I are going to try and build our own RF deck, looking at the simple circuit that they units use to convert the FSK into data, we think we can. If we do, we can use a -real- RF receivers and exciters. Regards, Dave