Measuring truck's wheel speed for feedback to the system

Thread Starter

Ishan Jain

Joined Nov 12, 2016
100
Hi folks,

I am working on a project and for that I need to measure the speed of the truck wheel and initially I used an IR sensor but it is not reliable and accurate so is there any other way to do that which is reliable as well as easy to mount on the truck wheel.

Thank you.
 

Alec_t

Joined Sep 17, 2013
14,280
Does your truck have an ABS system? If so, it already has a wheel speed sensor you could tap into.

Edit: See K1W1's post below. A secondary sensor would be safer. But, as with any modification to a vehicle, you should check that it won't invalidate your insurance or contravene any local vehicle-related regulations.
 
Last edited:

K1W1

Joined Jan 13, 2018
13
DO NOT TAP INTO THE ABS SYSTEM.

This is really bad advice!
ABS is a safety system that could save your life someday. Tapping into a wheel sensor could cause the ABS ECU to malfunction when really needed.

If ABS is fitted there is usually a reluctor ring on the wheel for the ABS wheel sensor. You may be able to mount another sensor to pick up a signal for your project.
 

K1W1

Joined Jan 13, 2018
13
Yep, even with all the new-fangled safety safety systems in today's vehicles, we still kill thousands of people every year on our roads.
I wonder what the death toll would be without seatbelts, airbags, crumple zones, ABS, ESC etc.
 

crutschow

Joined Mar 14, 2008
34,280
Don't know how we ever lived so long without ABS. :rolleyes:
Be snarky about if you want but it likely saved me from an accident when I entered a freeway blind off-ramp a little too fast in the rain and realized too late that it was backed up nearly to the freeway.
I clamped on the brakes and felt them shudder as they pulsed the brakes to keep me straight, and I did manage to stop inches short of the car in from of me.
Without the ABS I'm pretty sure I would have either slide into someone in the next lane, or hit the car.
I know they tell you to manually pump the brakes in such a situation but I couldn't have done it nearly as fast as the ABS.
 

strantor

Joined Oct 3, 2010
6,782
Is calculated wheel speed ok? If so, you could pick up RPM of the driveshaft by putting some magnets on it read by namur prox, although that may throw off balance. Putting reflective stickers on the driveshaft read by IR sensor would work until the stickers got dirty. An inductive prox could pick up RPM of the universal joints but the resolution would be lackluster.
 

shortbus

Joined Sep 30, 2009
10,045
I am my own ABS.
As am I. My 2004 Pontiac Grand Prix, came from the factory with no ABS. ABS and other things like it are just one more dumbing down of the driving public. I was taught how to control a car and pay attention when driving. One of the big things that still resonates with me to day was taught in driver ed back in the old days. "always look three car lengths ahead when driving." Snarky or not it still applies even with ABS.

Now you kids get off my grass.
 

shortbus

Joined Sep 30, 2009
10,045
Wow. I thought ABS has been mandatory for the last 20 years. Nope! Next year backup cameras will be mandatory but still not ABS.
It gets even weirder on the Grand Prix. Instead of having the master cylinder having a front and rear circuit, it has them crossed. The right front and left rear are on one hydraulic circuit and the other has left front and right rear on it.
 

#12

Joined Nov 30, 2010
18,224
pick up RPM of the driveshaft by putting some magnets on it read by namur prox, although that may throw off balance.
After-market cruise control kits come with 2 equal magnets which are strapped on the driveshaft 180 degrees apart with stainless steel wire. Add a hanging down piece of metal and attach a coil to detect the magnetic pulses.
 

strantor

Joined Oct 3, 2010
6,782
After-market cruise control kits come with 2 equal magnets which are strapped on the driveshaft 180 degrees apart with stainless steel wire. Add a hanging down piece of metal and attach a coil to detect the magnetic pulses.
TWO pulses per rev is enough for cruise control?!

Let me do some math real quick...
Assuming 30" tires and 3.73:1 gears:

(30"*Pi)/12" = 7.85' travel distance per wheel rev
5280ft/mi /7.85ft = 672 wheel revs/mi

672*3.73 = 2,508 driveshaft revs/mi
2 pulses per driveshaft rev, 5,017 pulses per mile.

assuming 60mph (2508 driveshaft RPM), that's 5,017 pulses per minute
5017/60 = 83.6Hz

Sure, uh, I guess that's "good enough" HAHA. In my world, VFDs & Servo amps control motor speed ("cruise control") using 2,000+ pulses per revolution encoders on shafts going 1,800+ RPM for 60kHz. At the very least, a 512PPR encoder for 15kHz.

It completely blows me away that this can work with such a low frequency. Thank you for dragging me outside of my bubble.
 

crutschow

Joined Mar 14, 2008
34,280
As I recall, the cruise control I added to my front wheel drive car used 4 magnets attached to the front wheel axle with 185/70R14 tires, which gave 834 revs/mile.
That gives a frequency of 834*4/60 = 55.6Hz @ 60MPH.
It worked well.
 

Thread Starter

Ishan Jain

Joined Nov 12, 2016
100
Thank you guys for your suggestions, the truck doesn't have an ABS system so please tell me how can I get the wheel speed keeping in mind the mounting difficulties I will face when placing each sensor on or around the wheel.
 

LesJones

Joined Jan 8, 2017
4,174
Also how many pulses do you require per revolution. A hall sensor would probably be more reliable than IR. You may be able to drill holes, machine notches or mount small magnets into holes drilled into the hub assembly. There may even be some feature on the hub assembly that could be detected with a hall sensor.

Les.
 

AnalogKid

Joined Aug 1, 2013
10,986
assuming 60mph (2508 driveshaft RPM), that's 5,017 pulses per minute
5017/60 = 83.6Hz
I guess that's "good enough"
Sure it is.

What is important is how stable the sensor output transition edges are with respect to the angular velocity of the shaft. Even one magnet would be good enough as long as it doesn't wiggle. Using the same math as deriving the frequency of a signal by measuring its period, you easily can get 3-digit resolution of the shaft velocity.

In your example but with only one magnet, 83.6 / 2 = 41.8 Hz = 24 ms. The difference between 60 and 61 mph is 4 us per period, so a 1 MHz, 100 ppm clock is more than twice the precision needed for 100 mph resolution *instantly* (as in only one revolution). But nothing in a car happens instantly, or even in 24 ms. Early systems were analog (!) and used an R-C filter to get a short-term average of the shaft velocity and prevent "hunting".

ak
 
Last edited:
Top