logging Setup for Garmin GPS 18x5hz

[jeanmi30]

Hi,
Is some one have yet used a serial GPS Garmin 18x5hz connected to a Gems logger ? I would be very interesting for a base logging setup. Also, witch frame have you select in the sensor.

Best regards

[CMW]

I am logging that GPS receiver via CAN from my Motec M800 ECU. Motec have started suggesting that anything higher than 5 Hz be connected directly to their loggers or dash boards, and I think the reasoning is the GPS stuff uses a lot of processor power which they don't want diluting too much doing none engine control stuff. I have 3 LA99-L2 loggers and would maybe like to try reading the GPS direct into one via the serial input, so am watching this with great interest. I have not, as yet, managed to get a perfect track map from the GEMS logger, viewed in either GDA Pro or via Google Earth, whilst Motec's i2 standard shows the same data fine. The problem is odd spurious lines appearing unless I only read the middle of the logged data. I will post a screenshot of what I mean later. What logger are you using? People claimed the coordinates could simply be added together from the CAN data and a map drawn, I have found fairly complex maths is needed to show positive and negative coordinates correctly. This may not be an issue via serial data.

[jeanmi30]

Hi,
Thanks for your reply.
I use, for this aplication, a DA1 logger that collect from engine and other ECU data by CAN, so, the serial line is availlable for my GPS. I don' t try yet to connect it but I'll do soon.
Do you know how is configure your GPS device, witch sentence is transmit ? ( PGRMC, etc )
Your car have a major difference on mine : the GPS is connected to the MOTEC ECU, and then transmit by CAN to the Gems logger. May be some informations should be usefull

[CMW]

I left the Garmin receiver set to the default settings, and it's been fine. If I can help at all, I will.

[RickS]

Good work guys, that's what a forum's all about!

Just to let you know that we're currently collating information regarding this aspect of data logging and will post an entry on the Q&A forum. Keep an eye open for updates.

Regards,
RickS

[CMW]

Maybe you could comment on why, when logging to my DA99-L2 logger the track map shows these weird straight lines, which appear to be created before the car moves? If I select data from later in the log (say a second lap) the lines are not shown. Logging the same session and the same GPS data inside my Motec M800 and viewing it in i2 does not show these lines. I would also like some tips on filtering, as the track lines have the odd small zig zag in them, which again doesn't show in i2? Looking forward to the info on GPS logging, thanks.

http://www.gatesgarth.com/GDA/track_map.png

[RickS]

Will get into it. Do these lines disappear as soon as the vehicle starts motion or only if you start review the log some way into the playback? Initially, they look like tracking lines but why they appear is, at the moment, a mystery to me.
It's my first week at GEMS and am in the process of getting set-up, please bear with me.

Regards,
RickS

[RickS]

Usual question.... What version of GDA are you using. Apparently, a new release was implemented recently with enhanced GPS features. I am looking into 'filtering' options also.
Regards,
RickS

[CMW]

Sorry Rick, the version is 4.0.2.0 Build 567 It does it when you review the log in its entirety, if you only select a mid section of laps the lines don't show. To clarify, I have run in on board data logging in my M800 alongside the external log from this ecu, via CAN to the GEMS logger. So the GEMS is reading the same data as the ecu logging itself. I'll dig out the log file and if you want it I can send it over when you are settled in. Any info on the new loggers mentioned on the main GEMS site, yet?

[RickS]

Thanks for the update. I don't seem to be able to recreate the 'weird' lines of the track map, so I would be grateful for a copy of your log. Best to email it to: tech.support (at) gems.co.uk when you get a mo. I suspect that 'filtering/conditional error removal' is the area we should be looking into although, through looking into the version history on GDA, there is a later version for your perusal.

Had a quick chat with 'sales' and release of details for the logging developments is imminent.

Regards
RickS

[CMW]

I have zipped the log file up,and posted it at:

http://www.gatesgarth.com/rock4.zip

There's no rush, the car is put away for the winter now, but I would like to understand what is happening here, thanks Rick.

[RickS]

OK. I think that I'm onto a winner re. these spurious Aztec lines appearing over Rockingham!

I've had a detailed look at the data that you posted and the GPS Latº and GPS Lonº that come out of the maths processor, have intermittent spikes. Now depending on the time that you set up in the Fixed Range fields in the track editor will govern whether you get the spikes or not and so show on the track map . See second file attached.....

Now by adjusting the time to a section of the data that has no spikes (I used 12:30.00 - 15:00.00) the track renders without our spurious lines. See first file......

All I will do now is to start experimenting with 'DECLICK' and try to get rid of them. I'm sure what is causing these spikes but tactfully wonder if it's 'dirty' data coming from the GPS device. If this is not showing with the i2, is there some suppression code already in place?

Regards,
Rick

[CMW]

I believe there *IS* some filtering in i2 as the same date renders both without the spikes and with a smooth contour to the track map, sans the tiny zig zags and occasional more major one. I will have a close look at the i2 stuff, and also mail Motec to see if they do, or will, share what sort of filtering is applied as standard, without any end user input. Do you agree that the rather complex maths is needed, as initially people told me that it should render without any maths other than simple addition of the high and low values? I tried this and because some values are negative I had to enlist the kind help of a maths savvy friend to sort out some algorithms that became ever more complex to do this. I hope we didn't miss anything obvious?

Thanks for your timely reply.

As an aside will your next generation loggers simplify this? And will the software be backwards compatible with my and other earlier loggers?

Cheers RickS

[CMW]

I will ask on the Motec forums what internal filtering is done, but meanwhile please see http://www.gatesgarth.com/motec_filters.png Thanks.

[RickS]

Thanks for the new info. Will analyse when I get a moment.

Regarding the maths. I tried not to be influenced by previous entries in this thread and just followed my nose through the 'help & tutorials' and managed to get the calculated traces (that I forwarded yesterday) with the following:-

combine16('GPS LAT HW','GPS LAT LW') / 10000000
combine16('GPS LON HW','GPS LON LW') / 10000000

I haven't had time, as yet, to experiment with the peak removal but will keep you up to speed on my process.

Regards,
RickS