Building a PC Plotter – Part Two

Keen readers will have noticed there’s not been an update for six weeks- in fact, not since arriving in Swansea. Well, this is because the weather- particularly the weekend weather- has been either too wet, or too windy, or both, and since sailing over to Mumbles in the last blog post, Karisma hasn’t left her berth on Pontoon M. In fact, all that’s happened is I’ve added an extra spring line, in order to ensure she can’t surge forward on the finger and hit the main pontoon, as the berth is a bit small for her and so much wind has been whipping through that I’ve had to re-attach the starboard dodger three times as it keeps getting torn off in gale after gale.

We have, however, spent quite a few weekends on the boat, enjoying the company of our friends the Bridgewaters (and their dogs!) who also have their boat berthed in Swansea. If the weather holds next weekend, they plan to head around Lands’ End to the South Coast and we will go to Pembrokeshire, so at least we have made the most of the time in twin-town.

I thought I’d return to the theme of the boat’s onboard PC system, which was installed about a year ago and the description of which has been one of the more popular posts on this blog. The reason for this is that I have had to do it all again- because the core components all failed within a year of installation, and I thought I’d now document the replacements.

Firstly, the AOC LM20 monitor, which was such a bargain on Ebay at £26.99, gave up the ghost over the winter- I switched it on in February to find a green line of failed pixels running from top to bottom of the screen. Fair enough; it was probably around fifteen years old, had probably completed thousands of running hours in its lifetime, and had been a bit of a punt anyway. After quite a bit of searching I sourced a suitable brand-new replacement: a generic 17-inch CCTV monitor, imported from China as a brand called SAC systems. It cost £161.19 including VAT and a little P&P, and has performed perfectly well for the six months it’s been installed, although how easy it is to get hold of seems uncertain, as it’s already no longer available on Amazon. It requires 12v DC delivered via a standard 2.5 x 5.5 mm jack (exactly the same as the AOC) and has VGA and HDMI connections. It uses the same 75mm VESA mount as the AOC, but- unlike that monitor- also has internal speakers which connect to the PC via a standard 3.5mm stereo jack cable, so it can be used to play back video.

This ran happily with the Sumvision mini-PC until a couple of weeks ago, when suddenly the VGA output from the mini-PC failed, and it was impossible to get a picture on the screen. After a lot of troubleshooting, all of which failed, I asked a question about it on the YBW.com forums, to discover that another chap with a Sumvision onboard had just suffered an identical failure. Sumvision have now released a ‘Mark 2’ version of the mini-PC, and after a lot of thought I simply replaced the units like for like, as the VGA component of the original PC is just part of an all-in-one motherboard and they aren’t really user serviceable, as well as there being nothing really competing with the Sumvision as a VESA-mountable mini-PC in the hundred notes sort of bracket. At least on Amazon, there seems to be no history trail of VGA failure, so I figured perhaps I was unlucky, and perhaps the Mark 2 unit would have such issues ironed out.

So when the replacement unit arrived I was confident I would simply have to plug it into the power, connect to the monitor via VGA, and these difficulties would be behind me. I had spent some time setting the unit up at home- laboriously re-installing all of the plotting and NMEA handling software, and talking to the paid-for software vendors to ensure all the licences were released from the previous machine (a thumbs up to Eltima, PC Plotter and Meridian who all understood the situation and did so without any issues). However, on plugging in to the boat, the PC suddenly went beserk- it began freezing and crashing, and I could not get it to see the multiplexer which collects the NMEA data from the boat’s systems. Fortunately, my friend Simon was on hand and after a lot of trouble shooting we tracked the issue down to- the power supply, which was the only remaining bit of ‘original’ hardware from last year’s install. When run on the boat’s power supply, the PC would crash; when run on its own 240v AC to 12v DC converter from shore power, the PC worked perfectly.

The power supply is shared by both the screen (which was still working perfectly) and the mini-PC and is a ‘proper’ stabiliser/regulator, that is, it is supposed to supply a continuous 12v DC to the computer whatever the input voltage (coming from the boat’s batteries) is, precisely to prevent freezing, crashing and damage. Shorepower connections and charging batteries via the engine alternator can result in the actual voltage of a boat’s nominal 12v system being anything up to 14.5v, which will damage sensitive electronics, and the regulator is supposed to prevent this. The unit I had installed was a comparatively expensive regulator too, made by a company called Amperor. In fairness to the manufacturers, they state its nominal output as 12.6v; when I checked with a multimeter it was chucking out 12.89v. The Sumvision’s 240v PSU was supplying 12.28v, and the approx 0.6v difference was wreaking havoc. Of course, the $64,000 question quickly became: what burnt out the VGA adapter of the original Sumvision PC? Had it simply failed (in line with the experience of at least one other user on the YBW forums), or had it been burnt out by the Amperor slowly supplying a higher and higher voltage? We shall never know I guess. I replaced it with a (cheaper!!) stabiliser from a company called Hunterfield, which puts out a nominal 12.0v and finally had the whole system working again.

Autopilot issues

When we delivered the boat to Swansea, I mentioned that the Autopilot simply refused to steer to a waypoint and seemed unhappy steering to a compass bearing either. It was suggested to me that the power wiring might have been responsible for the difficulties experienced in getting the unit to steer to a compass bearing, and I have still got to get around to examining the full loom to assess that. However, the lack of steering to a waypoint turned out to be another computer issue: the AP was taking its NMEA from the output of the multiplexer, and the multiplexer was talking gibberish: its NMEA output was completely corrupted. I hooked my spare laptop up to the NMEA output and the sentences were corrupted n a way reminiscent of a stop bit error. I have seen this before because the multiplexer needs to be run with two stop bits, and most applications automatically configure for a single stop bit. However, changing stop-bit or other definitions had no effect on the output, which remained garbled. I have not got to the bottom of it but seems that the output of the multiplexer might have become damaged in some way. Of course I am now wondering if it might be in some way related to the excessive voltage supply being delivered to the original mini-PC, which in turn powers the multiplexer via USB. In the end I simply re-wired the NMEA loom such that the AP gets its NMEA directly from the Simrad AT10 (the conversion box which translates the N2K output from the Simrad Chartplotter and instruments into NMEA0183). Although this means the AP can no longer be driven by the PC or by the old Garmin 128 at the chart table, the AP is driven perfectly by the chartplotter in the cockpit, meaning that waypoint navigation is a simple as touching the screen, setting that point as a waypoint, and pressing the ‘navigate’ key on the autopilot itself.

Sometimes, especially if you have installed a PC on board a sailing yacht, simple is also best. Fingers crossed for fine weather this weekend!