Race Directors: Best way to measure/certify a course?

Keith Stone wrote:

A standard bike computer will never measure within tolerance. The problem is it's only taking one measurement per revolution, so you could be off something like 80 inches, and the computer missing a single revolution will mean you'll never hit tolerance

Right, plus that is magnified through the course.

The *best* a bike computer will come, with 4 magnets per wheel, is about 1:21 of a Jones Counter.

Like Scott I also use a calibrated bicycle cyclometer along with the Jones.

By the way I have several 1 mile loops, marked every 110 yards, and have often come within 1/2 a count per loop with repeat measurements on the Jones Counter. That's 2 inches per mile. The courses don't come out to exactly a mile but one of them is within 3 yards, so I have additional little marks for the mile points after each loop. Very convenient.

Wildcat34 wrote:

Scotth: what is your opinion on a typical GPS system's ability to measure a race course within acceptable error?

Cute, 1 or 2 guys taking after Keith w/a repeat post. Anyway...

Keith and John R have answered this query well. What isn't spelled out here is what's 'within acceptable error?' How much 'error' are we talking here? In the certification world, 2 measurements must come w/in .08% (about 4' per mile). Let's assume that's what's being asked about.

I have no experience using any of these measuring devices you can carry on the body. I've heard good things about them however. This does NOT mean I'd feel confident using them to measure for certification. If I'm after what I'd describe as 'an accurate measurement', something I'd feel comfortable w/while doing repeats, GPS, Timex and Nike devices would all be okay. A measuring wheel is cool too. I doubt GPS will be okayed for use to certify courses.

I've found when I have my mtn bike tires pumped up to max, my computer on it is remarkably accurate. The computer on my measuring bike, the one w/a Jones counter, is little better than assistance to alert me when the next mark is coming up. The computer on my road bike is generally accurate but varies subtly w/tire pressure & weather. The bike computers are functional but not accurate enough for certification purposes.

A standard bike computer will never measure within tolerance. The problem is it's only taking one measurement per revolution, so you could be off something like 80 inches, and the computer missing a single revolution will mean you'll never hit tolerance

You need to calibrate (just like you do with a Jones counter!), but there's no need to enter a calibration parameter into the bike computer. The wheel circumference is always set to 2.5m, and four magnets are placed on the spokes. That way, the odometer counts revolutions. Since you mark the starting point of the wheel, you can not only count whole numbers of revolutions, but fractions of them, based on the final rotation angle of the wheel (like using a measuring wheel). So the issue of the "resolution" is a non-issue -- you can guestimate about 1/3 of a spoke by eye, which is 2cm, not 8 inches. If you're off by one magnet, you'll know, because the counter doesn't turn over at the starting wheel position.

I'm not Scott, but I have used GPS units. On a course in a park with lots of turns, my GPS results differed by over 5% on a 1 mile course compared to the result I obtained from using a bike computer with a calibration run immediately prior to measurement. The GPS odometer was off by an enormous margin, even summing the distances on the track log underestimated distance. But on a longer, straighter course, the GPS did better. I don't think it's even close to course measurement standards though.

The *best* a bike computer will come, with 4 magnets per wheel, is about 1:21 of a Jones Counter.

Sorry, what does the number of magnets have to do with it ? As long as it's set up to count whole numbers of revolutions, it shouldn't matter a whole lot whether you have 4 or 8 or 1 magnet (though having more than 1 magnet does help tell you if you get a sporadic extra read or a skip). I can take readings to within 1/3 of a spoke (about an inch). Of course that doesn't mean that my measurement is accurate to within 1 inch, but the resolution of the computer itself is not an issue. Read the link that was posted earlier in this thread.

I can understand how you could calibrate a bike computer nearly, if not equally, as well as a Jones counter. The problem I see though, is that when you are out actually measuring the course, if for some reason your computer misses a magnet pass that's an entire revolution's error (or quarter revolution if you're using 4 magnets). If your Jones counter misses a click, it's like 2 inches.

The problem I see though, is that when you are out actually measuring the course, if for some reason your computer misses a magnet pass that's an entire revolution's error (or quarter revolution if you're using 4 magnets).

If you're using 4 magnets and it misses a click, then you will know about it because the odometer will turn over at (for example) spoke 9 instead of spoke 0. Or if you get a double-count it turns over at spoke 27. I suppose you could choose to correct this (you need to assume you made 1 error and not 3!) or just do it again until the spokes come out right.

So you can by electronic odometers that will actually show you the number of magnet passes? Or at least set their calibration so they increment their reading with each magnet pass (or every 4 magnet passes). If that's true then I could see how it would work with 4 magnets. With one magnet, I don't know how you would know that it missed a magnet pass.

So you can by electronic odometers that will actually show you the number of magnet passes? Or at least set their calibration so they increment their reading with each magnet pass (or every 4 magnet passes).

Yes. Use 4 magnets, set the circumference to 2.5m so that's 10m or 0.01km every 4 magnets), and set the display to metric. When you do this, the odometer increments by precisely 0.01km with each full revolution. So if the odometer reads 2.27km, that's 227 turns of the wheel. Then count of spokes to get the remaining distance.

elflord wrote:

So you can by electronic odometers that will actually show you the number of magnet passes? Or at least set their calibration so they increment their reading with each magnet pass (or every 4 magnet passes).

Yes. Use 4 magnets, set the circumference to 2.5m so that's 10m or 0.01km every 4 magnets), and set the display to metric. When you do this, the odometer increments by precisely 0.01km with each full revolution. So if the odometer reads 2.27km, that's 227 turns of the wheel. Then count of spokes to get the remaining distance.

Sorry in advance for my ignorance, but I am trying to set up my bike computer (sigma 500) using the four magnet system after reading the article on the website and am completely unfamiliar with bike computers

1) Does it matter if the magnets are attached at the end of the spokes? the only way I can get them close enough to the sensor is by sliding them down towards the center.

2) If I've marked off the spokes (36), is it necessary to make further divisions between them? I'm not looking to certify a course, just looking for greater accuracy than the standard bike computer.

3) Once I have the number of revolutions, I assume that I need to compare that with how many revolutions are needed to go over a fixed distance (e.g. a mile), yes? Is using a track acceptable for this?

thanks in advance,

dave

1) Does it matter if the magnets are attached at the end of the spokes? the only way I can get them close enough to the sensor is by sliding them down towards the center.

Your priorities are correct. More important to get them close to the sensor. No need to have them at the end of the spokes. The clips are very effective.

2) If I've marked off the spokes (36), is it necessary to make further divisions between them?

I use this for a similar purpose to you. I think it depends on how much accuracy you want, but I'm pretty happy with about 1/3 of a spoke (2cm or so), which I can do by eye.

3) Once I have the number of revolutions, I assume that I need to compare that with how many revolutions are needed to go over a fixed distance (e.g. a mile), yes? Is using a track acceptable for this?

For your purposes, I'd say yes and yes. You need to calibrate it otherwise you're not going to do much better than a standard setup.

I use a standard track, this is the easiest way to do it.

One of the problems is that the track is not straight. So you can either use the 100m course or deal with the turns (keep in mind that the 400m is measured 6in from the line!) I'd recommend just using 4 rides of the 100m course instead.

Dealing with the turns introduces a lot of complications.

It is best to first have some experience with course measurement and certification before giving advice on these things.

> 1) Does it matter if the magnets are attached at the end of the spokes?

It doesn't matter as long as they are spaced equal distance apart, i.e. every 9 spokes etc. You are measuring revolutions, not circumference.

2) If I've marked off the spokes (36), is it necessary to make further divisions between them?

As has been said, you can count the spokes. However, you are working outside of the margin of error or the cyclometer.

3) Once I have the number of revolutions, I assume that I need to compare that with how many revolutions are needed to go over a fixed distance (e.g. a mile), yes? Is using a track acceptable for this?

YES it is vital to calibrate the cyclometer, which means riding the bike over a calibration course several times.

NO a standard track is NOT useful for this purpose. The only part of the track that is useful is the straight. However I would only use the straight with a Jones counter and not with a cyclometer. For that purpose you need to have at the very minimum a 400 meter and preferably an 800 meter straight. You can NOT use a curved surface for calibration and expect accuracy.

The Jones counter measures 84 counts per revolution of the wheel. You can also tell 1/2 and 1/4 counts etc. A cyclometer with 4 magnets measures only 4 counts per wheel. That is a ratio of 1:21. A cyclometer is certainly no closer to a Jones counter than this. You also have the glitches of where the wheel is when the magnets go by, and when the cyclometer decides to start the readings etc.

Also I've noticed that my cyclometer gives slightly different readings from time to time, especially at the start but also during a course.

The Jones counter is ALWAYS the same.

There is no comparision between the accuracy of these.

Also you can and should calibrate the internal workings of the cyclometer.

As to measurement, the most important part of the process is measurement of the calibration course, which only needs to be done one time, and then accurate checking of the measurement device -- every time -- before and after it's use.

It is best to first have some experience with course measurement and certification before giving advice on these things.

I do not have experience with certification, but I do have experience with measuring courses for my own purposes, using various methods (GPS, footpods, bike computers). So in my opinion, I am qualified to give advice in this limited scope.

I am prepared to cite supporting references to any claims I make that are beyond the scope of my personal experience.

On the other hand, you make sweeping (and often, erroneous) criticisms of using a bike computer, despite the fact that you quite clearly do not understand the method that you are criticising. Now, please do read this URL before commenting further, so that at least you understand what it is that you are commenting on.

http://home.earthlink.net/~caverhall/newrevcounter/abstractcontents.htm

.

While I don't have experience with course certification, the author of this website does.

As has been said, you can count the spokes. However, you are working outside of the margin of error or the cyclometer.

Incorrect. Read the URL.

NO a standard track is NOT useful for this purpose. The only part of the track that is useful is the straight.

I agree that the straight is useful and the curves are problematic.

However I would only use the straight with a Jones counter and not with a cyclometer

Why ?

You can also tell 1/2 and 1/4 counts etc. A cyclometer with 4 magnets measures only 4 counts per wheel. That is a ratio of 1:21.

Incorrect. There are 36 spokes on a wheel, not 4. Read the URL.

You also have the glitches of where the wheel is when the magnets go by, and when the cyclometer decides to start the readings etc.

Not if you're careful. Read the URL.

Also I've noticed that my cyclometer gives slightly different readings from time to time, especially at the start but also during a course.

Yes. Because you just used a standard bike computer setup, and didn't take care to position the wheel correctly at startup. You'll get better consistency if you're more careful.

Also you can and should calibrate the internal workings of the cyclometer.

Incorrect. Read the URL.

As to measurement, the most important part of the process is measurement of the calibration course

VYes, but it's a bit of an overkill for just measuring ones daily routes or intervals. For that purpose, I think a track does pretty well.

thanks for the advice. as i stated earlier, i'm not certifying courses or the like, I'm just looking to find some distances on local roads, and parks.

if the curves of a track aren't good (i understand why), is one one-hundred meter straightaway long enough to calibrate it? i have no intention of shelling out for a jones counter, and am just looking for the most accuracy I can find.

Why doesn't USATF come up with a second level of certification? You wheel off a quarter mile, calibrate your bike computer to it, and then go measure your course. Presto, B-level certified.

I think there are a lot of race directors that would be willing to do this.

Right now there are so few races that have a certified course, many race directors don't view it as something they even need to consider. If USATF made a B-level certification with an easy procedure, many more courses would be certified and it would become the norm. Race directors would be expected to at least B-level certify their course. Any race that wasn't at least B-level certified would be assumed to be inaccurate.

You raise some interesting points dukerdog. First, USATF would never be party to a system of certification where the value is nebulous. The 'B' status that would be accorded fails to spell out how close to accurate the measurement is. Bike computers and operators are not created equal.

Your point about the relative lack of certified courses is a good one. I don't know the answer. Directors are put off or scared off by various things connected w/certification. The sport would be less recreational and more professional if the most fundamental aspect, course accuracy, were a given.

scotth wrote:

You raise some interesting points dukerdog. First, USATF would never be party to a system of certification where the value is nebulous. The 'B' status that would be accorded fails to spell out how close to accurate the measurement is. Bike computers and operators are not created equal.

But I would know that someone at least put a minimal effort into measuring the course. USATF would need to make it very clear that a B-level certification doesn't guarantee anything about how accurate a course is. It only certifies that someone calibrated a bike computer and then rode the entire course. If I had that guarantee in every race I ran I would be a happy camper.

Right now we have a situation where you know the course has been meticulously measured following the USATF-approved proceedure, OR, as is usually the case, you know nothing about how the course was measured. We need something in the middle.

if the curves of a track aren't good (i understand why), is one one-hundred meter straightaway long enough to calibrate it?

For your purposes, yes. It's probably not adequate for the certification gurus posting here, they'd want 400m. I'm not a certification guru, I just measure my courses for my training/intervals, so it's good enough for me.

Put it this way -- 1 spoke is about 5cm out of 100m, which gives you a "calibration resolution" of one twentieth of a percent. By doing a 4 or so runs, hopefully you

reduce random errors and get a decent calibration. Again, it won't be up to the standards the ceritification gurus demand when measuring courses (they do multiple runs of a straight 400m course before and after), but it will be much more accurate than a GPS or footpod doodad.

In terms of internal consistency, I got about 5 spokes difference (less than a foot) over a 0.9 mile course with a fairly hasty and careless measurement. This is not as good as the accuracy that the other guys in this thread get with their Jones counters, but for my purposes, it's pretty damn good.

Elwood,

I appreciate your enthusiasm but you would benefit by getting some experience so you understand what you are talking about.... first.... then talking about it after that.

4 magnets per wheel = 4 counts per wheel.

You don't have 36 counts per wheel unless you have 36 magnets per wheel.

Understand?

The Jones counter has 84 counts per wheel.

Dukerdog,

Many race directors like to have their courses accurately measured but not certified, which I have also done for them quite a few times, and have also shown others how to do this. It's also been a lot of fun measuring my own courses, and running on them, the courses and marks lasting for years.