GPS/Computers

[qsiv]

Re: Given that most NMEA is slow

I'm reassured someone is thinking of the same solution - I've come very close to ordering an embedded development kit on more than one occasion, but as it is a back burner project I havent quite committed to it yet. I keep hoping one of the NMEA MUX suppliers might do what I need!

<hr width=100% size=1>

 

[Aeolus_IV]

Re: What about a variation on ....

No, not a printer buffer, but a cache of the latest version of each NMEA sentence "seen" on the input. After all, you do not want to replay your position over the last 30 minutes at high speed, just where you are now. It always strikes me that this should not be too difficult to put together - but unfortunatley I'm not hardwere engineer, on the other hand the software/firmware ought to be fairly simply. Same thing with the "NMEA combiner", still can't understand why they cost £100+ each...

Jeff.

<hr width=100% size=1>

 

[tome]

Re: What about a variation on ....

Have you tried using a Kalman filter for this? It sounds like an ideal application. Incidentally, we take care of latency by timestamping each datagram into each port to 10 msec resolution.

<hr width=100% size=1>

 

[qsiv]

Re: What about a variation on ....

I hadnt thought of that - but KF's are perhaps better for linear systems, subject to noise. I'm not sure whether the issue of a 'rapid change of course (tack or gybe) really fits the model, as it is in effect a discontinuity. I could though apply a Kalman filter to the true wind, as that is effectively a 'trend', and look simply for inconsistencies in the AWA, which approach I hadnt considered - so thanks!!!

<hr width=100% size=1>

 

[tome]

Re: What about a variation on ....

I presume you have a reasonably fast fluxgate (5Hz)? If so you can resolve AWA relative to TWA without any steps caused by course changes.

<hr width=100% size=1>

 

[qsiv]

Re: What about a variation on ....

I think the issue is to do with the damping rate being applied to the wind instruments (in order to minimise roll and yaw influencing the AWA) - as a result there can be inconsistencies in the TWA which is a calculated output. There are other issues to do with the masthead describing an arc as the boat tacks, which generates an apparent wind input which further messes up the AWA that is measured.

The current band of IACC boats use 3 axis solid state gyros (at least one uses a Willow AHRS (automatic heading Refernce system) - I havent tracked the rest). These were originally developed for 'a missile program' - I'd guess Tomohawk, but are now increasingly used in standard avionics. The hassle is that the AHRS are very pricey (for recreational market) at about £10,000 (even if they do have the dreaded Kalman filters built in<g>), but they do run at 100Hz, so it is possible to completely filter out the effect of masthead acceleration from the AWA. B&G kit already corrects for upwash from the main, so that doesnt have to be done. It is surprising how difficult it is to get accurate input data!

<hr width=100% size=1>

 

[Searching]

Blimey!!!!!!!

I thought it was a siimple thing to sort out using a USB for GPSs

Now I know what is simple and its ME

Thanks for the info I have found it VERY enlightening but I wont hold my breath for a solution. The main problem for me is the demise of NMEA on lap tops.

Thanks

<hr width=100% size=1>

 

[tome]

Re: What about a variation on ....

Qsiv

I've PM'd you on this as the thread is getting a bit specialised for some? It's a very interesting subject all the same!

Regards
Tom



<hr width=100% size=1>

 

[tome]

Searching

I don't think it's you that is simple, it's the notebook manufacturer's not appreciating a specialist market need for serial ports to deal with NMEA data. Sorry if it became too technical, it was a perfectly reasonable post you put up in the first place but with such a diverse user base on the forum it can be interesting how these threads develop.

Regards
Tom

<hr width=100% size=1>

 

[AlanPound]

Nobody has mentioned it yet (and I don't know if it is 'old hat') but, when you *do* get your USB->RS232 adaptor wire up and working, there is sometimes an operational issue to deal with.

If you connect your GPS to your USB device, and *then* plug it into your laptop, your Windoze often sees the stream of incoming data, and installs it as a MOUSE! (with an immediate effect on your arrow pointer...)

To avoid this, plug the serial adaptor into the PC, let Windoze recognise it as a serial adaptor, THEN plug in your GPS... All should be fine.

There is a little registry hack to avoid all this nonsense, but I don't have it handy (and there are health warnings)...

Also, as far as I have noticed with my own setup, *real* serial ports occupy COM1..COM3, so at least some USB serial adaptors install themselves as COM4 and up. Problem is that some software won't recognise COM4 (as it doesn't believe it exists) MaxSea is like that (or at least mine is...)

OziExplorer works fine with COM4 though...

<hr width=100% size=1>