Garmin GLO vs QStarz 818XT Part 2

RC-GLOIn my previous post on the Garmin GLO vs QStarz 818XT I looked at the two units from a physical perspective without having done a real world test. In this article I’m going to look at and analyse the data produced by RaceChrono using both devices riding on Silverstone’s GP circuit.

HeadstockProtectorFor the test, I installed both GPS receivers on the bike. The Garmin Glo was strapped to the rear of the bike on the tail piece while the QStarz 818XT was taped between the top yoke and the instrument console on the head stop lock bracket (circled in yellow).

The test was conducted for one session with a detailed analysis of the fastest lap from that session. These are the raw numbers produced by RaceChrono for lap 4 of the test session:

  QStarz 818XT Garmin Glo
Lap Time 3:02.12 3:02.02
Distance (m) 5846.835 5854.346
Data Points 911 1346
Sample Rate 5Hz 7.5Hz
Trace Colour Red Blue

Taking the raw numbers from RaceChrono and plotting them against a schematic for the track is quite illuminating. The diagram below shows the two traces obtained from the GPS devices. The red trace is the QStarz 818XT while the blue trace is from the Garmin Glo.

Vale-Club

The first thing to notice is how both traces cut the right handed corner between Vale and Club. Neither trace is correct as I’m pretty sure that I was close to the right hand side kerb but definitely not over it. This isn’t totally unexpected because the reported accuracy of these devices is down to around 0.5 – 2m depending on the quality of the satellite signal and fix.

The second thing to notice is that the Garmin Glo (blue plot) records my line through those two corners more accurately than the QStarz 818XT. If you look carefully, you’ll also notice that the red trace (from the QStarz 818XT) looks smoother than that from the Garmon Glo. You can see this better in the detailed view below.

ValeTrace

The QStarz 818XT delivers a data point to RaceChrono every 1/5th of a second (hence 5Hz) without fail. The Garmin Glo is supposed to deliver a data point every 1/10th of a second (ie 10Hz) but fails to do so, often skipping one or more data points within each second – look for the gaps in the purple dots on the blue plot. Over the course of the 3 minute lap, the Glo delivered a data point every 0.133 of a second resulting in an average frequence of 7.5Hz – about 25% below the quoted 10Hz. As noted in my previous article, this could be a flaw in the Garmin or more likely the Bluetooth stack in the Android operating system used to run RaceChrono.

Comparing the two traces, it looks as though the QStarz 818XT (red) has come kind of filter or smoothing algorithm because of the smooth nature of its trace. The Garmin Glo’s trace (blue) shows some “jumping” between two consecutive data points; almost as if the bike jumped one bike width’s to the side. Rather than being a flaw, this indicates to me that the Glo doesn’t attempt to compensate, smooth and adjust fixes but just reports the fix as calculated.

TraceLuffield1Over the course of the lap, both devices report pretty similar traces, except for the trace recorded by the QStarz 818XT at Luffield. You can see this in the red trace that leaves the track completely for a while before finally rejoining the circuit. A more detailed view is show below.

TraceLuffield2

It’s pretty obvious that the QStarz 818XT was unable to accurately locate itself possibly due to a temporary loss of view to the overhead satellites. It takes the device almost 8 seconds to re-establish its position back on track. This could have occurred because the QStarz 818XT was mounted in a position on the front of the bike where my helmet (and upper torso) could have been shielding the device from a decent view of the satellites. I checked the laps before and after and this problem didn’t recur so it’s likely to have been a one off that lap. What’s interesting is that a seemingly large error like this has relatively impact on the overall lap time and both devices report lap times within 1/10th of a second of each other for that particular lap.

Lap QStarz 818XT Garmin Glo Difference
1 3:11.55 3:11.52 0.03 3/100ths
2 3:06.13 3:06.12 0.01 1/100th
3 3:05.64 3:05.64 0.00 Same
4 3:02.12 3:02.02 0.10 1/10th
5 3:03.94 3:04.00 0.06 6/100ths

With the exception of the excursion off track described above, both devices show fairly similar traces around the circuit. In terms of position reporting, the Garmin Glo looks to be more accurate than the QStarz 818XT but that could be because the QStarz 818XT was located less favourably than the Glo. If you’re looking to plot your lines around the track, the Glo would be a better unit to use than that QStarz 818XT. If however, you’re just interested in your lap times, then both devices do a great job reporting very similar lap times as you can see above.

In my next article, I will analyse the speeds reported by the two devices at different parts of the circuit.

Silverstone GP 2 Jun 2014

SilverstoneGP-circuitI was really looking to riding Silverstone’s GP circuit again; the last time I rode the longer GP circuit was nearly a year ago. An added bonus was an email the night before from someone I met in Cartagena asking whether I would be at Silverstone too. He had traded in his Ducati 848 Streetfighter for a Panigale 899 and no doubt would be just as fast as the last time we rode together.

This track day was organised by Silverstone circuit and cost £175 (US$ 280) for the day. Although pricey, at least we would be offered free instruction, food vouchers and the use of the expansive Wing pits and facilities. The GP circuit is 3.6 miles (5.7km) long and feels like riding the International Circuit twice back to back. Newcomers are often confused about which part of the circuit that they are on as the two pit straights (Wing and National) are easy to mix up – this can lead to disastrous results if you confuse the slower and tighter Copse corner with more open and faster Abbey turn at the end of the straight! It’s definitely worth spending some time learning the track before you go balls out through some of those turns.

Spots of rain started falling from a cloudy and grey sky just as I left London at 6am. The forecast was variable for the day and I hoped for better weather at Silverstone, 80 miles to the north. Although I do want to try my Michelin Power Pilot 3’s on track in the rain, I wanted to delay that experience for another day. Luckily as I drove on, the sun started to break through the clouds and my anticipation of a good day on track increased along with the warmer and brighter weather.

There were three groups this day; novice, intermediates and fast – I was booked into Inters group. Having been to this track quite a few times now, I know many of the instructors by sight and name and so during sign on, Neil, one of those instructors, offered to ride with me later on that day. I was quite looking forward to this because Neil’s good fun and I’ve always enjoyed riding with him on track.

Of the three groups, the inters one was the biggest with 73 (yes, you read that right!) bikes. There was a fair mix of abilities in the group ranging from some ex-racers to a few less confident riders. The first session of the day comprised three sighting laps followed by a return to the pits before heading out on track for the remainder of the session. The session started well during the sighting laps but the day soon deteriorated into carnage. This seems to be becoming a bit of a familiar story! I’ve never been on a track day with so many fed flags and hold ups. It seemed as though our sessions were almost always delayed by five minutes due to to a red flag or some other mishap in the previous fast group session, and that there would almost inevitably a red flag incident during our sessions too.

LuffieldCrashProblems was caused by several factors. Firstly the size of the group was such that you’d often come across a bunch of bikes which would take time to get through and which inevitably led to frustration and some spectacularly bad overtaking. Another factor was the mismatch in speed and agression between the slower riders (I include myself) and the faster ones. The final issue was that a minority of riders who seemed to have little regard for their own well being or anyone else’s on track. Even one of the instructors I was talking to during the day was worried about being t-boned by some idiot going into a corner.

Despite the problems, I still enjoyed the day although this was possibly the worst managed event I’ve been to at Silverstone in a while. One of the contributing factors is that Silverstone now offer paid instruction which had the effect of reducing the number of floating instructors out on track to help keep things in check. In fact one instructor even complained that he felt ever so slightly “used” because he was booked for paid instruction every session and was hardly going to get a break all day nor help anyone else on the track.

During the third session, I managed to make a 10 second improvement in my lap times compared to my last outing on the GP circuit the previous year. Part of that improvement came from my determination to try and focus on my body position and lines pretty much to the exclusion of all else in order to try and achieve higher corner speeds. This was one of the things picked up by an instructor at Brands a couple of weeks earlier.

Although Neil was too busy to ride with, I did manage to spend half a session riding with another instructor, Owen, in the first session after lunch. I followed him for a while trying to learn his lines and then he followed me for a bit. Watching the video back of that session, I can see that my lines and my consistency for keeping those lines is definitely and slowly improving.

My improvements, however were no match for Anders on his beautiful red 899 Panigale. He was devastingly fast – the result of a combination of natural ability, youth and just the right amount of aggression… oh, and horsepower. While the Panigale was fast through the corners, it was unbelievably so down the straights. However, the downside of that power was its fuel consumption because Anders was using around 5 litres of fuel per 20 minute session on track – an average consumption of around 20mpg! Frustrated by the inters group, he moved up to the fast group for the last two sessions of the day where he reported fewer bikes on track and a much better standard of riding. He also felt that the group wasn’t particaurly quicker than the many of the quicker riders in the inters group.

Here are my times for the day.

Session Laps Fastest Comments Video
1 - - Didn’t time this session
2 5 3:06.36
3 5 3:02.12 youtube
4 - - Failed to start timer, doh!
5 5 3:07.03 Following instructor youtube
6 3 3:08.61 Red flagged
7 4 3:06.54

The problems with the number of bikes in our group and the perception of a lack of supervision marred what would have been a truly brilliant day. Despite that I did enjoy it and really felt that I’m continuing to make slow improvements in my riding. One of my big Achilles Heel’s is still a lack of corner speed and I’ll be taking a specific course in the next few weeks to try and work on that particular aspect of my riding.

How good is Silverstone’s GP circuit? Good enough that I’m booked to return in four weeks time!