|
|
|
Forum Newbie
Group: Forum Members
Last Login: Monday, March 15, 2010 11:58 AM
Posts: 2,
Visits: 13
|
|
Hi,
Newbie here with a nice used SDR-1000 with 100 watt PA, no ATU, that was recently bench tested before shipping to me. I've put together a system with a reasonably fast computer, Edirol FA-66, PowerSDR 1.18.3 as recommended by the seller, Astron 35 power supply. The unit came with the USB to Parallel cable as well as a short parallel port cable - both of which have been used successfully with it in the past. The computer is running winXP, SP3, was recently updated and then isolated from the net and dedicated to flex use. It's a dual opeteron 254 system with a Thunder K8W (S2885) motherboard with on board lpt and firewire (TI chip) support. CPU usage runs 7-20%
I set it up initially and tuned around on 75 meters and was able to receive well and love the filters and ability to see what's going on. Transmit power seemed low and when I tried the PA calibration, it failed. I sorted out some cable issue with the FA-66 (balanced output on tip and ring, not tip to ground mono as I had assumed) and now the PA calibrates, but only on 80 meters. All other bands it doesn't show any output beyond a watt or two. And it's deaf on all of the other bands. To me it seems as if the band filters are not switching properly except for 80 meters. I can hear a little bit of activity on the other bands, and the frequencies match up to other transceivers but where a signal might be s9 on another rig, it's barely audible on the sdr-1000, except on 80 meters where it's comparable. On the broadcast band, where normally I see dozens of strong BCB stations, I can only see a few of the strongest on the flex.
I've tried running it with both the short supplied LPT cable off the computers parallel in both normal and ECP mode - same results. I've tried installing USBIO and running it with the USB to parallel cable and get the same results. I have listened to the SDR unit and as I select from band to band, I can hear relays clicking - with the sole exception of going from 12 to 10 meters, which I suspect share the same filter, so *something* is switching or trying to.
At this point, I'm pretty stumped. The fact that it works ok on 80 meters tells me the basic hardware, the sound card, and my computer are all more or less ok. There's obviously some parallel port communication going on else I wouldn't even be able to turn it on, and the relays moving argues that something is changing on the port when I select the different bands, although maybe not the right stuff. One oddity I did notice as I selected through the bands while connected to my dummy load, was that the noise floor on 30 and 17 meters was much higher than on all the other bands. On most bands it's in the high -130's or low -140's, but on those bands it's in the -108 region and the panafall display shows wideband noise.
As the unit was just checked out before it was packed up and shipped, and my computer is newly dedicated to the task, it seems likely that the issue is on the computer end - but as the system is so synergistic, how do I sort out which domain the problems lie in? Is there a troubleshooting procedure or software to sort out issues on the parallel port or USB to Parallel adapter? Do the lines on the parallel port stay static when the bands are selected stay static and measurable if the connector is pulled off the flex unit to measure? Is there a listing of what pins should be in what state depending on band that I could check against? Is it possible to emulate the computer output onto the parallel port (I'm envisioning something like a 5 volt supply, a parallel port plug, and some clip leads to take each line high or low as needed) to check the flex hardware band by band independent of the computer?
I don't know if there are any flex folks in central or western NY (I'm in Ithaca) but I could sure use a Flex Elmer and maybe a known working computer to plug on to. I'm looking for any ideas on a next step for troubleshooting this issue.
Thanks in advance.
73 de Kevin, WB2EMS
|
|
|
|
|
Supreme Being
Group: Forum Members
Last Login: Tuesday, January 17, 2012 4:25 PM
Posts: 215,
Visits: 503
|
|
Kevin,
A couple of bits of info that may help... There are two sets of relay switched filters switched per band... The the PA board there is a set of Lowpass filters. And on the RFE board there is a set of half-octive filters that are used on both TX and RX...
( one set of relays could be audibly switching bu the not )... The parallel control port on the SDR-1000 uses mulitplexed data lines so the data is latched and the signals on the cable connector is not static.
You can download a set of schematics for the SDR-1000 from the flex downloads sections.. do a search for schematics
The other thing you could try ... is to reset you database.. Assuming your using 1.18.13 PSDR, go into your install directory and rename database.xml and then restart PSDR
AL, K0VM
SDR-1000, Flex5000, Flex user since 2006
|
|
|
|
|
Forum Newbie
Group: Forum Members
Last Login: Monday, March 15, 2010 11:58 AM
Posts: 2,
Visits: 13
|
|
Al,
Thanks for the reply. I figured the lines on the parallel port would be chunking several bytes into latches, but figured it didn't hurt to ask. 
Since I posted, I've opened up the radio and started poking around inside a bit. By pressing on the top board, near the connector stack that goes through the various boards, I'm able to pretty repeatably get some relays to click when I apply pressure, and then un-click as I release it. After a bit of cycling like that and other general wiggling, it seems to be working correctly (at least it will pass the transmit calibration and seems to receive normally) on all bands. So at this point I'm pretty sure the issue is a connection issue in the hardware, and possibly in that top board.
Other than the connector itself, which was recently replaced to solve just such issues, are there any other well known problem areas that exhibit this kind of behavior? To me it sort of has the feel of an smt device with a bad solder connection to the pad - makes contact when flexed one way, lifts just enough not to when flexed the other.
I think the relays that are changing with pressure applied are in that top board, and that's consistent with not properly selecting the right filter which was giving the symptoms I reported.
How hard is that top board to get in and out? I'd like to look at the SMT devices on the bottom of it and check their soldering.
Kevin
|
|
|
|