[Mikko Nassi]

Starting a new topic for a discussion that started in an unrelated thread:

Any reason why you have your speed sensor on the rear? Or do you have one on the front as well?

Only have one speed sensor.....

Why the rear? well, it shows where my wheelspins are....from poor throttle control! and how I brake....

I use hard compounds.....which is another reason to observe the above.

why the recommended front from Unipro? ( their manual did not suggest rear at all....)

I think if you pour over the manuals of most data logging systems you'll find that they also recommend to put it on the front wheel in a kart with no front brakes.

The trace that you get with the wheel speed sensor is your speed trace and to measure your distance. With the front wheel in non-front brake karts this is extremely accurate - it doesn't lock up, so the only time when it really starts to lie is in extreme understeer conditions or if you're on a track with a few jumps .

When you have just the one sensor and you put it on the rear wheel you don't get an accurate speed trace and distance measurement anymore. This is absolutely critical when comparing laps - with rpm you don't get very accurate comparisons at least on low rpm when your clutch is slipping around.

Sure you can see lock-ups and wheel-spin... but you can also very easily see wheel spin as well as lockups to an extent by overlaying your rpm trace and front-wheel speed trace.

With at least the unipro software it actually automatically makes a slip-graph for you as well so you can see your wheel spin.

So you gain being able to easily see lockups - which you should very easily feel on track as well - but you lose the ability to overlay speed traces accurately.

Most systems are also able to split the lap into segments based on distance, which is incredibly useful to analyze all sorts of things, but is only accurate if you use the front wheel to measure distance.

Whaddaya think?

[gp8]

When we started out we used a rear wheel sensor primarily as it was easier to install, but also for the reasons Jules mentioned. Once the driving had at least become consistent (ly slow...) we swapped to the front wheel as this allowed us to analyse mid-corner speed with greater accuracy. this in turn helped us investigate the affects of weight distribution, chassis setup and driving technique on entry deceleration, mid-corner speed and exit acceleration without having to allow for wheel spin etc. Oh, and you get something that looks much more like a track map out of your MyChron from a front wheel data feed...

[Mikko Nassi]

When we started out we used a rear wheel sensor primarily as it was easier to install,

Yeah I've read the "easier to install" explanation a few times. ... I'm not familiar with the MyChron front speed sensor - what makes it difficult to install?

With the Unipro the front speed sensor is very similar to the rear, only takes a few minutes to install. Like on the rear axle where a collar is used that goes around the axle there's a collar that attaches to the front wheel hub.

Rear:


Front:

[gp8]

Mychron (AIM) do not produce any brackets or mounting kits to aid a front wheel installation. We've had some made up from laser cut aluminium to suit that look a lot like the Unipro front wheel sensor and installation bracket. A rear wheel installation is just a small angle with a hole drilled in it hung off a bearing cassette bolt - much easier!

[RocK]

for AIM user, GPS is an option. AIM said its the best form to log speeds.
i used it, no complains so far..

[Mikko Nassi]

RocK do you have a wheel speed sensor as well? You could check with that to see how accurate the GPS sensor is.

I read a comment from a Pi Research employee saying that the GPS was 4-7kph off at times and even over 10kph off at some points. This was on the MyChron4 system. He claimed that GPS is not quite there yet for karting applications - you'd need a military-spec GPS (mega-expensive) to get the accuracy there apparently.

gp8 sounds like there's a market for someone to make front wheel installation kits for MyChron and other units that don't sell a suitable kit.

[RocK]

Yea i plan to have a 2nd speed sensor (rear i thk), to log wheel spin also to compair vs gps. only then we can see how accurate gps are.

[gp8]

Mikko, yes there is a market for installation kits for MyChron wheel sensors. Originally I was just going to buy the UniPro kit, but the price in Australia was eye popping. The installation bracket for the stub axle was a comfortable AUD$20, but the front wheel sensor disk (just a disk with 6 magnets that fixes to your front hub with grub screws) was AUD$180

Also, I couldn't figure why I'd want 6 magnets sending 6 impulses per wheel revolution when the sample rate is only 10hz. Not sure if I'm right, but I figured that with a tyre diameter of 875mm the maximum speed I'd be able to read would be about 52.5km/h, being the speed travelled when each of six impulses per rev is 1/10th of a second apart...

It all got way too confusing, so I just had a single magnet wheel sensor disk made up in aluminium with tungsten inserts to ballance out the weight of the single magnet...

It still makes my brain hurt just thinking about it

[gp8]

Oops...I mean "tyre circumference" not "tyre diameter" of 875mm. I must remember not to post comments before my morning coffee

[Mikko Nassi]

Also, I couldn't figure why I'd want 6 magnets sending 6 impulses per wheel revolution when the sample rate is only 10hz. Not sure if I'm right, but I figured that with a tyre diameter of 875mm the maximum speed I'd be able to read would be about 52.5km/h, being the speed travelled when each of six impulses per rev is 1/10th of a second apart...

Unipro explains this, and makes it all seem very smart.

The measurement of speed can be done in different ways. Some Laptimers use a method where the speed is calculated using the RPM combined with input about the gearing and wheel circumference. Other Laptimers counts the number of pulses from a wheel sensor in a certain amount of time, and then calculates the average speed during the period. The main disadvantages of both methods are that they are not very precise, and have a low update frequency. These methods are not very suitable for data logging.

Instantaneous method
The instantaneous method counts the time between pulses from a wheel sensor, and then calculates the speed at specific instants in time. We use this method in the Unipro Laptimer. Our wheel sensor has six pulses per wheel revolution, which results in a very accurate speed measurement. The total precision is 0.48% from 0 to 350 KMH combined with a very fast update rate. The method is perfect for a real data logger and enables the PC Analyser program to do things like a slip graph showing you if the wheel is spinning or dragging and how much!

So the data is calculated a lot more than 10 times a second, but it is only stored in the logger 10 times per second.

[gp8]

Now that is SMART! I've just read my Mychron's user manual as well (when all else fails read the instructions...) and note that the speed channel is configurable for up to 199 impulses per wheel revolution...

I have written a query to MyChron Mike's loose wires blog at AIM on how the MyChron 3 and 4 actually caulcuate speed and will post the reply when I receive it.

AIM also now produce what I'd describe as a partial front wheel speed sensor installation kit comprising a bracket to attach the sensor to the stub axle. You are still left to figure out how you are going to mount the magnet(s) to your front non-magnetic wheel rim.

[gp8]

According to "Mike MyChron" the MyChron 3 and 4 use the same technique as the UniPro, being time between impulses. The exact response to my query about the technique of calculating speed was

"Time between pulses, of course it needs the tire circumference to finish the calculation for speed so the more accurate that info is the better. Remember that due to tire growth at higher speeds and as inflation pressure rises due to tyre warming, the measurement is not ever going to be 100 percent accurate but should be used as a baseline for all your changes. Just be sure to periodically verify your circumference for greater accuracy"

Mike suggested a maximum of 10 impulses per revolution and noted that due to the tyre growth issues the accuracy of wheel speed sensors vs their GPS module was near identical.

[Jules]

Well Mikko & guys,

Thanks for an interesting thread....which in some ways started from one of my earlier replies.

Anyhow, I do agree that logging the front wheel speed does seem to be more advantegous but the rear wheel speed also do tell a few things.

Having said that, I think we can conclude that the more magnets, the more accurate the data will be and I do think this is true for any wheel u are placing this.

Now, like gp8's post - the datalogger I have does not have a front wheel magnet bracket.....so it's either I gotta get the one from Unipro or do what the earlier versions of data loggers do....which is to stick ONE magnet on the rim! Which we will all agree gives data......but perhaps not accurate enough.

On my present set-up, i have a rear bracket for the 50mm rear axle and that carries 4 magnets.....

well, let me think of something and maybe make a replica of the AUD180 kit...ahem!

[Mikko Nassi]

well, let me think of something and maybe make a replica of the AUD180 kit...ahem!

The magnet-disc thing is listed as only 34 Euro + Tax on uniprolaptimer.com.

The 180 AUD would probably be complete with the sensor as well.

[gp8]

Actually, the quote I received from the distributor here in Australia was AUD$160 for the sensor disk (each) only and AUD$20 for the fitting bracket (each) only. It didn't include the sensor. It must be due to a strange exchange rate or taxes or something other than margins of course... Euro 34 is quite reasonable.

How much are the same pieces in Malaysia?