On Thu, 21 May 2009 08:06:38 -0400, "Jeff DePolo" <[email protected]> said: > > > This leads me to a question that I have had on my mind. How > > are people > > doing desense testing with D-Star systems? (Remember, it's digital.) > > > > 73, Joe, K1ike > > How about this - record a clean D-Star transmission (not decoded, just > the > "raw" output from an FM receiver) on a PC with a good sound card, then > use > it to modulate the sig gen in your service monitor. Do desense test > using a > lossy tee like you normally would, except instead of comparing 12 dB > SINAD > points, you'd have to rely on listening to the repeated/decoded audio on > another radio to gauge performance.
Jeff's techniques should work. Another method is putting an IC-91AD or IC-92AD on extra-low power in a WELL SHIELDED box and passing it through a variable attenuator/iso-T with the best quality interconnect jumpers you can possibly muster, if you're cheap/frugal. We did this for "baselining" our system off the mountain, but haven't done it yet on the hill. The shielded box came from the cellular phone test lab industry, and didn't leak AT ALL. > This isn't an ideal way to test, but it's better than nothing lacking > real > D-Star test equipment. IFR/Aeroflex just announced that they're releasing D-STAR capability into one of their new model Service Monitors in June also. List price is ... astronomical, of course. It also does TDMA 2-way systems and NXDN and all the other newer-ish commercial stuff. > Is there any way to get pre-FEC and post-FEC BER metrics out of a D-Star > repeater or user radio? Not that anyone's been able to find. There's a mystery pin on the "serial" connection at the repeater modules themselves that's labeled "RSSI" in the service documentation (usually found only in Japanese, but some english versions have wandered out), but most folks who've played with it think it's just a standard analog voltage from the receiver, and not any indication of how the digital side of the repeater is coping with thing. To add insult to injury, VOICE in the D-STAR data stream is HEAVILY forward error corrected, so to REALLY test it ... you need to feed the IC-91AD/IC-92AD some serial data from a PC and then COPY it on another receiver and PC to see when things really start to fall apart. You can KINDA hear when the voice becomes error-corrected, but feeding something like a 1000 Hz tone through it is useless... like in the cellular industry you need to feed real words through it and determine the "copyability" of that voice for yourself... or just use the "completely garbled" falling out point as your known test point. No one's found any documentation on what BER rate is needed to be reached (in the failing/downward direction) before the rigs go from "copyable voice" to "garbled", but there's definitely a stage there where that happens in a CONTINUOUS transmission... the system can "pick up" if it re-syncs in that mode (mobile flutter/multi-path) but it often will NOT REPEAT if a signal STARTS OUT that way... digital hysteresis of some value... unknown. The UTAH VHF group has done the most accurate and useful "engineering" data on D-STAR I've seen yet... google for their website. Very civil debates have also "raged" on some lists about whether or not pre-amps help or hurt with the "broad" nature of the receiver's front-ends, and over time... as someone else pointed out, since these are "mobiles in a box" that's not shielded well, folks have figured out that the quality of the internal interconnect jumpers from the rear-panel N-connectors to the rigs themselves are pretty piss-poor in SOME bands radios. You pretty much just have to open yours up to find out, and of course, you're playing with fire for your warranty at that point, I suppose. Would have been nice if Icom had just spent a few extra bucks on a $2000 "repeater" and put some semi-rigid or at least good quality double-shielded stuff in there. Although most who've opened the 1.2 GHz modules have found good quality jumpers in those. The "cheapness" seems to be in the VHF/UHF modules. External amplifiers are also a bone of contention... the rigs and repeaters send a bit of a "preamble" prior to the start of real needed header data, and some have external amps on their D-STAR repeater modules with no particular problems... but "ramp-up" time/switching time is very criticial... the "routing" information (callsigns) is only really sent ONCE at the beginning of the transmission, and if it's lost, it's not like P25 where the Unit ID information is continuously interlaced in the data stream. Later, some folks found that the single transmitting callsign *is* interlaced but it's non-standard (not in the D-STAR specification) and something that obviously Icom decided to do, but doesn't contain the full four-callsign "routing header"... just the transmitting station's callsign. So yeah... there's some "issues" with it... but generally it's been fun to mess with it here... I've rattled on other lists about the Gateway and Internet connectivity being the "key" to it really being useful to hams... if you don't have IP connectivity that's low-latency, and even though the marketing material says it's not needed... a STATIC public IP address available AT THE REPEATER SITE, you really can't enjoy the Gateway as much as you can with those things... the Gateway really makes the system. Otherwise, it's just a digital repeater which is novel, but not novel enough to spend a LOT less money on an analog repeater at the site and look for another site for D-STAR. In all, a fun new toy -- we're having a great time playing with the low-speed data that's interlaced with the voice... simultaneous PC "chat sessions" going on while we're having voice nets (it can get annoying/confusing... you have to explain to people why they're seeing other callsigns go by in-between voice transmissions at first) and sending files, e-mail gateways, blah blah... mostly with Dan's D-RATS software, but Brian's D*Chat is fine for simple chatting too... and I haven't yet tried out the D-STAR TV thing, where someone recodes what's basically SSTV into a serial data stream and sends pictures around but I should... just not enough time in the day to play with it all. In all, a good "first" designed-for-hams system. There's some things I would have done differently if I were Icom, like build a little better quality repeaters, and be more open about API's to the proprietary Gateway software, etc... but it works... It's definitely a PITA to measure performance though. One application I've been meaning to write has been a "loopback" type test for the low-speed data system. If one had end-to-end BER by measuring how the low-speed data is doing... you could do a lot with that to predict coverage, etc... mix that with the DPRS/GPS location data, and you could have an application that you drive around spitting out data for a few days, and you end up with a coverage map... then if you made incremental CAREFUL changes like adding a pre-amp and appropriate attenuation, you could take your baseline data and see if you made things better or worse... Nate WY0X -- Nate Duehr [email protected]
