Wheel speed sensor

[jstoezel]

Hi:

My car came with a Hall effect type wheel speed sensor, which I believe is for an AIM system. I haven't really used it because I'm using a GPS based data logging system that provides the actual travelling speed at 20Hz.

Is there any benefit using 1 wheel speed sensor over the GPS speed? Or even combining the 2?
I understand it might help correct the GPS speed (and calculated travel distance), but at the same time it's the speed of only 1 wheel, that can lock etc (as opposed to averaging or using 4 wheel sensors).

Thanks,
Jean

[fvracer27]

Is there any benefit using 1 wheel speed sensor over the GPS speed?

Accuracy

You want to use a wheel speed sensor and you will only need 1

[smsazzy]

I would stick with gps. It is not prone to lock up issues that way. Unless you are racing in a forest or something, you will not have accuracy issues with a good gps receiver.

Also, if you abandon gps altogether, you will have to get a beacon also for lap times.

[BLS]

Perhaps a stupid question: What is the reason for speed readings? Just for interest? Or is there some usefullness to the data that RPM cannot satisfy?

[Mystique Racing]

I don't like the GPS speed because it seems to lag when the car is in the corners. However, I will admit that the GPS speed is nice to have as a backup when the wheel speed sensor fails which I have had happen more then once.

In the picture below the red line is wheel speed and the black is GPS.


[ external image ]

[fvracer27]

I don't like the GPS speed because it seems to lag when the car is in the corners. However, I will admit that the GPS speed is nice to have as a backup when the wheel speed sensor fails which I have had happen more then once.

In the picture below the red line is wheel speed and the black is GPS.


[ external image ]

From the picture you can see how much more info you get with the sensor.

[fvracer27]

Perhaps a stupid question: What is the reason for speed readings? Just for interest? Or is there some usefullness to the data that RPM cannot satisfy?

There is a ton of reasons to have speed readings best thing to do is take a class on Data. I just took 1 and found out so mush for so little, the stuff can be so helpful if you know how to read it. I took the clas because everytime someone told me something someone else said something different and Data is really interesting to me.

[jstoezel]

Hi Scott:

Thanks for the illustrated reply. Do you combine the 2 speed readings at all? If so, what logic do you use?

Thanks,
Jean

I don't like the GPS speed because it seems to lag when the car is in the corners. However, I will admit that the GPS speed is nice to have as a backup when the wheel speed sensor fails which I have had happen more then once.

In the picture below the red line is wheel speed and the black is GPS.


[ external image ]

[SR Racing]

In Scott's graph it appears the main difference on the straights is simply the claibration of the wheel speed sensor. (It always a bit lower than the GPS). In the corners there will always be some variation as the GPS is reading speed from its antenna location. The wheel speed sensor is from either the inside or outside wheel depending upon its mounting point. GPS is a ok back-up and you don't have to worry about the track beacon, but other than that, the data you get from either is fine. A change you might see in wheel lockup is hardly a worry overall. (and with the Wheel speed you can easily see a wheel lockup (assuming it is the sensor'd wheel) With GPS you can't see it at all.

In regards to requiring the track beacon. We have had 2 to 3 cars on the track for every race for the last 3 years and have never put a beacon out. Someone else always has one out there.

[tiagosantos]

There are lots of strategies as far as speed signals go, they all have pros and cons.. I guess the ultimate system would be 4 wheel speed sensors, averaging the speeds and discarding values significantly lower than the rest through filtering.. Hmmm. Expensive and complex! I believe the wheel speed sensors are more accurate, but wheel lockups can become a problem. You'd also need to keep measuring the tire circumference - maybe not much of a problem with your radial street tires, but with bias ply slicks, it can change quite a bit from tire to tire.

My GPS signal seems ok for what I need it to do (track mapping), and like someone else said above, I tend to use the RPM channel for comparison more often than speed. I don't have a wheel speed sensor on my car, but I would love to have all 4 wheels to check for lockups. I don't feel the expense is justified though, in my case at least.

[smsazzy]

Scott,

It looks like your tire circumference calibration is a little off. That is the reason for the diffence. If you look closely, it always seems to be the same amount off except for a couple other differences likely due to slip angle, inside vs outside tire, etc.

[jpetillo]

Scott, which wheel is the speed sensor on and what is the course direction - clockwise/counterclockwise? John

[Mystique Racing]

My wheel speed sensor is in the right front wheel. The course, Thunderhill Raceway, is CCW. There could be many reasons for the delta in the speeds including tire flex, tire circumference, air pressure, etc, etc. With that being said I still prefer to use the actual wheel speed for session Vs session and lap Vs lap speed comparison. I am no expert in GPS technology but I am a bit skeptical that the GPS can measure speed changes without a bit of lag when the car changes direction, such as going through a corner, as accurately as the wheel speed sensor. Certainly having two wheel speed sensors would be better then one. A simple speed averaging math channel could be created like Average speed = ((Speed_1+ Speed_2)/2)

[SR Racing]

but I am a bit skeptical that the GPS can measure speed changes without a bit of lag when the car changes direction, such as going through a corner, as accurately as the wheel speed sensor

Correct. Typical GPS sample locations are slightly better than 1 per sec. Any rapid change in direction or speed will not be correct. (The GPS will just average at the <1 per second sample rate.) Of course, your wheel speed is really only an average also (one rotation of the wheel), but it will typically be more responsive than the GPS.
If you wanted more accuracy(responsiveness) on the WSS, you could use 2 or 4 magnets. There is a doppler function available in the GPS strategy that can make speed changes instantaneous for all practical purposes, but not implemented in generally available GPS units.

But really, how accurate do we have to be in FV?

[brian]

My latest system is GPS based and I really like it. In addition to not having to rely on beacons, which can move from session to session, I can use the same start finish point as the AMB system uses making my lap times virtually the same as T&S. As to accuracy, I appreciate what Scott and Jim are saying but don't think anything that's happening in less than a second is repeatable by a driver like me.

[Speedsport]

My latest system is GPS based and I really like it.

I liked your old one - when we had that drag race up the hill on the last lap at Road America a few years back heading to the June Sprints flag I remember looking over and the lights were so bright on your dash I was watching them to see when you were going to shift.


I think the GPS based systems also suffer on tracks with elevation change. The relative speed to the sattelite changes when going up or down hill.

[jstoezel]

Aren't new GPS based units more accurate though? My old GPS unit was home built and I didn't massage much of the GPS data. GPS update rate was 10Hz, speed consistency was ok (no jumps when the unit was acquiring 5 satellites or more) but I had nothing to compare against.

I'm getting a new unit that claims a 20Hz update. I believe they fusion data from the accelerometers (and even gyros when the option is enabled) to improve the position and speed accuracy. Unit is waiting for me to pick at UPS so I haven't tried how accurate it is just yet.

There was a question earlier in the thread about the utility of a speed sensor. Not sure if this was mentioned in an early reply but speed compared against RPM can help detect wheel spin, which helps to diagnostic clutch/transmission issues and whether the driver makes a good job of shifting/downshifting.

but I am a bit skeptical that the GPS can measure speed changes without a bit of lag when the car changes direction, such as going through a corner, as accurately as the wheel speed sensor

Correct. Typical GPS sample locations are slightly better than 1 per sec. Any rapid change in direction or speed will not be correct. (The GPS will just average at the <1 per second sample rate.) Of course, your wheel speed is really only an average also (one rotation of the wheel), but it will typically be more responsive than the GPS.
If you wanted more accuracy(responsiveness) on the WSS, you could use 2 or 4 magnets. There is a doppler function available in the GPS strategy that can make speed changes instantaneous for all practical purposes, but not implemented in generally available GPS units.

But really, how accurate do we have to be in FV?

[jpetillo]

My wheel speed sensor is in the right front wheel. The course, Thunderhill Raceway, is CCW.

Scott, this should give you an idea of how much of the difference could be attributed to which tire the sensor is on. It seems like under 0.5 MPH for most of your run. This assumes that it's the outer tire. John

Cornering at 1.0 G and doing....
50 MPH, your speed is off by 0.62 MPH (1.3%).
75 MPH, your speed is off by 0.42 MPH (0.6%).
100 MPH, your speed is off by 0.31 MPH (0.3%).

Cornering at 1.4 G and doing....
50 MPH, your speed is off by 0.87 MPH (~1.7%).
75 MPH, your speed is off by 0.58 MPH (~0.8%).
100 MPH, your speed is off by 0.44 MPH (~0.5%).

[tiagosantos]

I think what John is really trying to say is, just drive faster and it won't matter!

[Mystique Racing]

John, I am confused by your post???? Not sure what your point is........

[brian]

I understand it to say that the outside tire is going faster than the inside tire, due to radius differences.

[hardingfv32-1]

GPS Data | Databytes | Data
By Luke // February 16, 2009
GPS offers a number of new ways to analyse and validate data. This article will give examples of the quality of data that can be expected to be gained, together with how this data can be used effectively. It will also outline what cannot be done, and how caution should be exercised when using GPS data.

TYPICAL GPS PLOT
Here we have a typical GPS plot from a 20Hz GPS system installed on a formula car. We can see the four wheel speeds, together with the speed generated from the GPS unit. Note that as the car is negotiating a series of corners, the effective radius the inner and outer wheels travel is different. Generally speaking, the GPS signal can be expected to be somewhere between the two. We can also see that the GPS speed is not susceptible to instantaneous accelerations on the wheels caused by hitting kerbs, shifting gears etc.

What a single GPS receiver
can do:
• Give an accurate measurement of straight-line speed. Ideal for cars with large aerodynamic download as it allows measurement of the heading velocity of the vehicle, independent of tyre squash / growth. In the same vein, it is perfect for using during aerodynamic straight-line testing for an accurate speed source as an input to a closed loop speed control.
• Given careful consideration of the data, driver lap overlays can be used to see the relative lines drivers take around a track. A series of positional accuracy channels are included and, as long as the values of these are taken into consideration and show the data to be of good quality, then one lap can be overlaid with another and a good estimate of relative driving lines can be seen.
• Create track maps with a more accurate representation of the circuit driven on.
Caution should be applied when using a single GPS

[jpetillo]

John, I am confused by your post???? Not sure what your point is........

I was trying to figure out how much you can expect your wheel speed sensor would differ from the GPS if everything was calibrated right.

What I did was calculated the tightest radius of turn given a velocity and coefficient of friction. Given that, you can calculate the velocity of the wheels with respect to the vehicle velocity to see how much it differs. It would tell you the most the wheel speed sensor could differ from the car's speed if you were not sliding, skidding or spinning the wheels. Any more or less than that as seen on your graphs between the GPS & wheel speed sensor would be attributed to the other things you had mentioned. I'm not sure I answered your question?

[jpetillo]

I think what John is really trying to say is, just drive faster and it won't matter!

Or just go straight!

I understand it to say that the outside tire is going faster than the inside tire, due to radius differences.

Yes!