Thursday, July 30, 2009

Finally, the station has a VHF/UHF antenna

Freshly back from a bit of a Summer jaunt around BC with my sister and family who were over from Switzerland, I finally got to tackle the installation of a main station VHF/UHF antenna. The procurement of a suitable antenna has been a bit fraught. Originally, I had intended using a old Comet dual band antenna that was last in use back in the UK in the 90's. That plan was kiboshed when I couldn't locate the three screw-in radials for the base. I'm sure these will eventually turn up, but I was surprised to find that they were not where I expected to find them - in a bag full of antenna paraphernalia that I had rather nicely organised when we moved to Vancouver almost 10 years ago.

While wondering whether to contact Comet to see if some new radials could be acquired, I developed an interest in D-STAR and indirectly in 23cm (the only frequency that supports the D-STAR 'high speed' DD data mode). I had installed a Comet 23cm Yagi to point at my local D-STAR repeater, but the experience got my thinking about adding 23cm in general - i.e. obtaining a VHF/UHF antenna that included this band along with 2m and 70cm. Taking the plunge, I initially ordered an MFJ antenna from Radioworld. This turned out to be a 'special order' and was backordered for weeks. When it finally arrived it was missing mounting hardware that had clearly popped out of the little plastic sleeve that most VHF+ antennae seem to be delivered in these days. Radioworld promised to obtain the missing hardware from MFJ (and I'm still waiting), but meanwhile I decided to opt for the Comet GP-98 anyway. This antenna has a slightly higher gain across all bands, and I was rather pleased with my Comet CYA-1216E yagi for 23cm, so decided to give Comet another shot.

The GP-98 duly arrived, with all its ancillary hardware, and as a two piece I assembled it without too much trouble. The only thing that caused a bit of a sweat was whether it was OK to pull a conductor/connector down from the upper section in order to mate this with the lower selection as shown in the sketchy instuctions. As it happens the whole inner section (the actual radiator assembly) does slide, but it took quite a bit of force to get it moving, and I wasn't convinced that I was doing the right thing! This antenna has a little compression fitting to physically connect the two fibreglass sections together once the radiating elements have been joined inside. There's a little rubber grommet that gets compressed as you screw a bezel from the bottom section into the upper section. All seemed to go well until a few days later, when the temperature in the room reached the high 20's, even 30's, thanks to a record-breaking heat wave here in BC. I was doing some email, with the assembled antenna lying on the floor a few meters away, when... *CRACK*. It didn't take more than a second to associate that noise with some plastic explosively failing, and sure enough, when I turned over the antenna, the compression fitting had failed, with a nasty big vertical crack up the joint. Presumably a combination of design/manufacturing defect, over(?) tightening and the high temperature combined to destroy the plastic under stress. So, this episode created the need for a new work item - repairing the connection. Luckily the strength of the antenna is mostly achieved by having the upper fibreglass slide into the lower section, so the joint is really only there to resist mild forces pulling the sections apart, and (mostly) to provide water-tightness. I affected a repair by cementing the crack with a sealant/adhesive before wrapping the whole centre connector in three layers: vinyl tape, self-amalgamating tape, vinyl tape - exactly like waterproofing an feed line connector. Then a number of tight nylon zip-ties around the connector were added to compress the connection and also to secure the tape. The outcome was quite acceptable, though it's always annoying to have something fail like this in the first place.

The next challenge was how to rig a side mount for the antenna. The standard arrangement is to use two lengths of tubing/pole horizontally strapped to the tower with two U-bolts apiece. A standoff mast is then secured between these using two more U-bolts, and the antenna is then installed on this. Taking some advice from John White (see previous posts) on the subject provided some recommendations for type of tubing (1-1/4" OD aluminium pipe for strength, with a nominal 1/8" wall thickness). I opted to get a mast made of the same pipe, and decided (somewhat arbitrarily) to make the distance between the horizontal members quite a large fraction of the almost 3m long antenna, for stability. I therefore ordered a 2m length of 'mast', with the intention that this distance be around 1.6m or so.

Obtaining the bolting hardware was the next challenge. I wanted stainless steel, as recommended, but the local purveyor of such things only had the U-bolts, lock washers and nuts - they were 'right out' of the backing plates required to cross-join two tubes with a diagonal U-bolt. I took what they had, but my hunt for the backing plates resulting in finding ready-made 'U-bolt kits' at a large big-name hardware store. These weren't stainless steel, but they had a backing plate. I intended using this backing plate with the other stainless hardware, but then realised that the holes in the plate were a tad too big. This meant also getting some flat washers. In any case, long story short, eventually I had a full set of fixing hardware, albeit a rather ragbag pick and mix.

John had also suggested installing the two horizontal members first, loosely and in such a way that the mast end is close to the tower. Then he suggested adding the mast, then the antenna itself, and finally sliding the whole assembly out to its desired standoff distance from the tower. When I sat down to think about this, it occurred to me that there would be an awful lot of dandling about on the outside of the tower with this approach. The fully assembled antenna, with feed line attached, with also have to be lifted to position, and the whole structure would also be rather flimsy until many of the U-bolts were properly tightened, as these diagonal connections all rely on each other and the maintenance of proper right angles. Instead, I opted to assemble the whole thing on the ground and then lift the entire subassembly up the tower, securing the horizontal tubes to the tower at the appropriate place. The main challenge with this approach is maintaining the structure, because as noted earlier, all the pre-tightened joints must be kept at right angles. To achieve this, the 'missing' side of the box, opposite the mast, was provided for with a wooden stick from the garden that was nylon-tied to the horizontal members (hoorah for nylon zip-ties!). The antenna was fully installed on the mast, and the feed line was nylon-tied down the mast, with the spare coax coiled and tied securely to the bottom horizontal. Not only does tying the coax keep it under control and out the way, but having it at the bottom also handily acts as a counterweight to the antenna. This meant that lifting could be done from the top horizontal and the antenna would stay upright.

Lifting of this whole subassembly was achieved by attaching pulleys higher up on the tower than the intended final height of the top horizontal. The pulleys were shacked to a tower cross member, and they were temporarily nylon-tied to the tower verticals in order to stop them migrating inward. Ropes were threaded through the pulleys, and the ends tied with appropriate position/spacing to the top member. The raising of the whole subassembly was relatively straightforward. The two ropes provided good side to side control, and another rope attached to the outer-bottom corner provided extra control when necessary. The only fly in the ointment was a realisation that my garden stick was protruding a little too high above the assembly and would hit the eaves soffit boards before the assembly was at the desired height. This was easily remedied and on the second try the whole assembly was easily lifted to the desired location and the ropes were tied off at the bottom of the tower.

With the whole assembly fairly accurately positioned, the next task was to go up the house side of the tower (affording the possibility of leaning back against the wall and using both hands) and the two horizontal tubes were U-bolted at the desired height. This was relatively easy, though it was at this point that I realised the backing plates I had were rather 'bendy'. I'm hoping that this doesn't translate to failure later - they are more 'strap' than 'plate'. I opted to secure the bottom horizontal first, then inspect the assembly from the ground to determine the final extension of the upper horizontal - in order to get the best 'vertical' on the antenna. Indeed on inspection the antenna was not quite vertical, but pushing out the upper horizontal a little was the right medicine. The upper horizontal was then properly secured in place. At this point, the "garden stick" is no longer required to keep the horizontal members parallel, and I could have cut the nylon ties holding it and let it fall. However, I opted to keep it in place for now (in case there some reason to need to disassemble the thing in the near future).

Finally, the feed line was run down the tower and into the house in the usual way. My initial attempt to fit the inside connector was challenged by failing light and I managed to miss a small fleck of braid material that must have shorted against the centre conductor when the plug was fully tightened. On my third try I had managed to find the culprit and had myself a non-shorted plug! Early tests (receive) show the antenna appears to be receiving the club repeater well (as expected), so hopefully all is OK.

My only concern remains with the quality of the backing 'plate' I have used with the U-bolts. I have no idea whether these bendy straps will end up providing the appropriate continued force, though they are done up nice and tight, so hopefully things will pan out fine. Also, some of the hardware is not stainless steel as I had wanted - time will tell what this will mean with respect to necessary maintenance due to corrosion. For now though, it's doing the job.

Tuesday, July 14, 2009

DV Dongle update

Well, I have the DV Dongle working after a fashion, with some interesting discoveries along the way:

  1. You really need to have your application (optionally your applet) JRE set up to be Java SE 1.6. This preference can be achieved in the Mac by using the included "Java Preferences" utility and moving this JRE to the top of the list(s). Note that AFAICT 1.6 is a 64 bit only option (at least it doesn't appear on my early Intel MacBook Pro).
  2. You must start the DV Tool application from the command line, rather than double clicking the JAR file, as is normally possible on the Mac. Apparently the application is quite sensitive to the relative location of some auxiliary binaries. I've played with the JAR packager tool included with the Mac dev tools to try to wrap this conveniently as a Mac OS X application bundle, but so far the correct setting of whatever search path is required to find these binaries has eluded me. I'm sure this is possible though, and presumably an app bundle could be created that just launches the shell script that can usually be used to fire up the app, rather than having the Java launcher load and run the JAR.
  3. On my set up, using a USB headset (at least my Logitech headset) for both input and output audio creates latencies that destroy the audio. You get about a half-second period clicking/drop-out in the audio in this configuration. The simple expedient of turning the output audio to another Mac audio device clears this up so the audio is clean. However...
  4. There is an approximate 2 second delay in audio pumped through the dongle. I have not been able to make this go away. I am running on a super-fast 8-way Mac Pro with loads of spare capacity (idle CPU and low memory pressure).

So, at present, I can connect to gateways and presumably (though I have not tried this yet) transmit audio to a repeater port. The approx 2 second latency though is still a question: is this nominal or degraded performance of the DV system? If the latter, can it be fixed and how?

I'm sure a 2 second latency would not unduly prevent reasonable use of the dongle on the D-STAR network (even if this behaviour isn't quite as designed). Except for timing a break-in, such a delay will only manifest, of course, at the ends of overs and you are supposed to leave plenty of time between overs on the D-STAR system anyway. The lag does sound quite noticeable however when you use a radio to monitor the same port that you have the dongle connected to.

I think my biggest use of the DV Dongle will be to monitor remote repeaters, which I don't think is possible on current radio D-STAR implementations - though I'll be the first to admit that I still have much D-STAR-fu to learn!

Wednesday, July 1, 2009

Fldigi for the Mac

I just bumped into Fldigi, a digital modes client that supports the Mac (...well as a cross-platform app, not quite a full Mac app, but that's OK).

I was idly looking for a client that supports the Olivia mode, having heard a few hams waxing about this on the air, and good ol' Google surfaced it as a result. There may yet be other clients to be fished out of the great internet ocean, but I'm going to give this one a try. It seems to have a fully fledged macro system, and an auto-logger that exports adif - for import to other logging software, as well as things like eQSL. It supports a fairly wide gamut of modes; certainly most I've ever heard of, though more flora and fauna are popping up regularly. It also integrates with, so my subscription for web queries that I paid for in order to feed VisualQTH can keep on giving!

Anyway, I now apparently have a choice of two clients (CocoaModem and Fldigi), though I've yet to check that it will interface properly to my SignalLink and CI/V rig control. I think there's a good chance as these are both pretty common interfaces/devices. I've noticed that the IC-7800 radio profile for the rig control is labelled as "untested", but hopefully that's just the authors being conservative... we'll see.