• Hello Guest, welcome to the initial stages of our new platform!
    You can find some additional information about where we are in the process of migrating the board and setting up our new software here

    Thank you for being a part of our community!

TruTrac Install in a 1031 axle

Lando said:
Gary, in your write up you said you added a tooth to your tone ring and it didn't throw off your speedo. I cant quite figure out how this is possible, as each slot registers as a fraction of a turn. I would imagine that it would change your speedo reading by a factor of (n+1)/n. On your tone ring that would not make much difference because it has a lot of teeth. I would imagine on the 12 tooth tone rings, it will read about 8% faster than it should be. Furthermore, that error will be propagated through an integral when the odometer integrates the speed signal into position.

How did you determine that it reads correctly?

Based on my understanding of how hall sensors signals are conditioned to determine speed, I believe you would have to have the same number of teeth as OEM. It might even be better to just leave a gap than to add another tooth, because it just knows n amount of impulses are in a rotation, and the diff spins fast enough the speedo probably wont notice the delayed impulse due to the gap.
Click to expand...

they don't "count" the teeth, it's a frequency reading, the faster the pulse the sensor sees from the ring determines the speedo needle position.
 
volvoracer82 said:
they don't "count" the teeth, it's a frequency reading, the faster the pulse the sensor sees from the ring determines the speedo needle position.
Click to expand...

I'm pretty sure it needs some sort of position reference to measure speed, which is a change in position per unit time.

So lets assume were talking about a 12 tooth ring for easier math. The signal processor is programmed to assume that each tooth accounts for 30 degrees of ring gear rotation. Then the signal processor, also records the difference in time (deltaT) between impulses. 30*/deltaT will give you your angular velocity which you can then calculate vehicle speed based on wheel radius. If you add a tooth, then the signal processor will still assume 30 degrees of rotation per impulse, but in reality, it is only turning 27.69 degrees per impulse.
 
Last edited:
Lando said:
I'm pretty sure it needs some sort of position reference to measure speed, which is a change in position per unit time.

So lets assume were talking about a 12 tooth ring for easier math. The signal processor is programmed to assume that each tooth accounts for 30 degrees of sun gear rotation. Then the signal processor, also records the difference in time (deltaT) between impulses. 30*/deltaT will give you your angular velocity which you can then calculate vehicle speed based on wheel radius. If you add a tooth, then the signal processor will still assume 30 degrees of rotation per impulse, but in reality, it is only turning 27.69 degrees per impulse.
Click to expand...

These sensors arent that smart, yes on lets say a 2008 bmw where the sensor has 3 hall sensors within, along with a small processor , but this sensor, no. Frequency is all it needs, You need computing power for that application. Simply said, an ignition system/fuel system need a missing or twice the width tooth for the computer to know when top dead center occours. Vehicle speed sensor in this case has no computing power,i Just a master and a slave wit wires in between.
 
volvoracer82 said:
These sensors arent that smart, yes on lets say a 2008 bmw where the sensor has 3 hall sensors within, along with a small processor , but this sensor, no. Frequency is all it needs, You need computing power for that application. Simply said, an ignition system/fuel system need a missing or twice the width tooth for the computer to know when top dead center occours. Vehicle speed sensor in this case has no computing power,i Just a master and a slave wit wires in between.
Click to expand...

Speed is a change in position per unit time. It is impossible to measure speed electronically without both a position reading and a time signal. Even if the position is simulated mechanically in the speedometer, adding another pulse per rotation will throw off your measurement.

Also a crank sensor is an absolute encoder, because it needs to know it's absolute position to function properly. A speedo sensor only needs to be a relative encoder, it just keeps on counting, because it doesn't matter how the ring gear is clocked at startup to determine speed. That takes very little computing power, 8 bit computers can handle that.
 
Speed was verified before and after via the same GPS unit, both for speed and distance, variance was the same before and after within .5% before and after.

As Ryan mentioned, it's simply a signal generator. The signal these generate is a simple AC sine wave, which the speedo uses to calculate speed, via the time between pulses. With a missing tooth the signal wouldn't be a clean signal for the entire revolution of the diff. You are thinking of a setup that counts teeth, which this one does not. This simply measures the difference between the high and low parts of the sine wave.
 
gsellstr said:
Speed was verified before and after via the same GPS unit, both for speed and distance, variance was the same before and after within .5% before and after.

As Ryan mentioned, it's simply a signal generator. The signal these generate is a simple AC sine wave, which the speedo uses to calculate speed, via the time between pulses. With a missing tooth the signal wouldn't be a clean signal for the entire revolution of the diff. You are thinking of a setup that counts teeth, which this one does not. This simply measures the difference between the high and low parts of the sine wave.
Click to expand...

I understand that the sensor only puts out a wave form signal, but all of the mechatronics in the speedometer (galvanometer?) is calibrated for a linear response to frequency.

For example: lets go with a car traveling 40mph on a 26 inch tire with a 12 tooth tone ring.

Assume the wheel is always rotating at the same speed as the ring gear.
Unit conversion: 40mph = 704in/s
Circumference of 26in tire is 81.68in

(704in/s)/(81.68in) = 8.61 rotations per second

The ring gear and tone ring rotates 8.61 times per second, and there are 12 peaks in the signal per rotation.

(12r^-1)*(8.61r/s) = 103.32 s^-1

The frequency the speed sensor will read is 103.32 Hz.

For the same car with a one extra tooth per rotation:

The ring gear and tone ring still rotates 8.61 times per second, but now there are 13 peaks in the signal per rotation.

(13r^-1)*(8.61r/s) = 111.93 s^-1

The frequency of the speed sensor will read is 111.93 Hz

So for the same car but with one extra tooth the frequency is increased by 8.3%. How does the instrument cluster account for this? I don't believe that it does.

My car has a 12 tooth tone ring. Adding a tooth to the tone ring would cause a noticeable amount of error. Which is why I asked. I suspect that since your car has a 48 tooth tone ring that the error would only be 2%, which is not noticeable and also below the uncertainty of most GPS units.

I will be waterjetting a part, and I will report back to find out if my math and understanding of hall sensors and signal conditioning is correct.
 
Last edited:
Lando said:
I understand that the sensor only puts out a wave form signal, but all of the mechatronics in the speedometer (galvanometer?) is calibrated for a linear response to frequency.

For example: lets go with a car traveling 40mph on a 26 inch tire with a 12 tooth tone ring.

Assume the wheel is always rotating at the same speed as the ring gear.
Unit conversion: 40mph = 704in/s
Circumference of 26in tire is 81.68in

(704in/s)/(81.68in) = 8.61 rotations per second

The ring gear and tone ring rotates 8.61 times per second, and there are 12 peaks in the signal per rotation.

(12r^-1)*(8.61r/s) = 103.32 s^-1

The frequency the speed sensor will read is 103.32 Hz.

For the same car with a one extra tooth per rotation:

The ring gear and tone ring still rotates 8.61 times per second, but now there are 13 peaks in the signal per rotation.

(13r^-1)*(8.61r/s) = 111.93 s^-1

The frequency of the speed sensor will read is 111.93 Hz

So for the same car but with one extra tooth the frequency is increased by 8.3%. How does the instrument cluster account for this? I don't believe that it does.

My car has a 12 tooth tone ring. Adding a tooth to the tone ring would cause a noticeable amount of error. Which is why I asked. I suspect that since your car has a 48 tooth tone ring that the error would only be 2%, which is not noticeable and also below the uncertainty of most GPS units.

I will be waterjetting a part, and I will report back to find out if my math and understanding of hall sensors and signal conditioning is correct.
Click to expand...

Get a tire that will return an 8.3% lower rotational velocity and you are set. It's as easy as 2(pi(r))=s. Then again LH2.4 electronic speedometers tend to read fast anyway. I think mine at least read ~5-8% faster than my gps readout at 80 mph.
 
harribert said:
Get a tire that will return an 8.3% lower rotational velocity and you are set. It's as easy as 2(pi(r))=s. Then again LH2.4 electronic speedometers tend to read fast anyway. I think mine at least read ~5-8% faster than my gps readout at 80 mph.
Click to expand...

Old school speedometers can also be adjusted, but I prefer hardware fixes to "software" fixes. It will cost me maybe $10 bucks to waterjet a new tone ring, and a speedometer adjustment would cost over 4 times that.
 
Think about this though...yes, you are adding in a tooth and gap (total of approx. 5/16"), but you are also moving the teeth into a larger radius ((approx. 3/16") so they will be moving slower being further from the axis. Your spacing is still the same so the variance on something with the number of teeth plus the added diameter thus slowing things down counteracts the added tooth.

Again, same GPS, same roadway, netted .5% variance tops.

BTW, this would be better suited to a different thread, rather than in the how to install the diff thread...
 
I think the obvious solution is if you have a 12 tooth ring, adding an extra 5/16 gap at one tooth is going to matter a whole lot less, especially because you don't have ABS for it to confuse.
 
What hell is going on in here?

Lando, you're on the money. gsellstr, volvoracer82, I think both you guys need a lesson in dimensional analysis and polar coordinate systems. You should try actually reading Lando's posts in here. He lays it out pretty perfectly.
 
Karl...did YOU see the part where the speed difference AND mileage distance is within .5% with the added tooth? I'm not going to dig into the details, but the real world results are what I'm seeing here. I do understand what Lando is discussing here honestly, however the end result from the user standpoint is more my concern.

Not trying to get into a pissing contest with anyone, just saying this is what the actual end result shows, and for the masses here, that .5% at most isn't enough to make a difference, especially given most here have different tire sizes or even different brands of tires than the car came with 20-40 years ago, which will throw it off more than the one added tooth onto a larger diameter tone ring.

Given the fact that 99.5% of the masses here don't have access to the CAD equipment and waterjet or laser cutting equipment to re-engineer the tone ring with .0065" wider gaps and teeth, then cut one out, this solution is viable, cost effective on the TB budget, and works with no discernible negative aspects. If you want to continue picking nits, please feel free, but please move to a new thread to keep the 'how to install a diff' thread free of the banter that isn't nearly as critical to the masses as it is to the engineering types. ;-)

/end ******* ranting
 
Gary, it's a good write up, I was checking it out because I have a trutrac I'm going to be installing soon. I just posed the question because I didn't feel yor solution would work for my vehicle, I was hoping that getting the right information out there would avoid confusion, because I'm sure that many others with the 12 tooth ring are wondering the same thing. Didn't mean to muddy up the thread.
 
No worries Lando. :)

Honestly for yours I would maybe take the stock tone ring and cut it in half, tach that on, and call it a day. You would have 2 sides splitting the 5/16" so the 3/32" extra gap would be of minimal issue in terms of resolution. Heck, split it 4 ways and the extra gap is down to 3/64".

Part of the reason I added the missing tooth as well was to eliminate the possibility of the ABS system seeing random changes in speed when the tooth went past as well. The last thing I need is random ABS cycling. Had enough of that when the ABS modules died on the R. lol
 
gsellstr said:
I do understand what Lando is discussing here honestly,
Click to expand...

I'm pretty sure you don't, actually.

gsellstr said:
Think about this though...yes, you are adding in a tooth and gap (total of approx. 5/16"), but you are also moving the teeth into a larger radius ((approx. 3/16") so they will be moving slower being further from the axis. Your spacing is still the same so the variance on something with the number of teeth plus the added diameter thus slowing things down counteracts the added tooth.
Click to expand...

... and this post proves it. Adding a tooth is counteracted by the larger diameter? That's a total nonsense statement.

I get that adding a tooth to your trigger wheel had very little real world impact, but it would in Lando's case. Be clear with what you're trying to convey, and arguing a patently false point doesn't do anyone any favors. It causes confusion and is likely to mislead other members.
 
Karl Buchka said:
I think both you guys need a lesson in dimensional analysis and polar coordinate systems.
Click to expand...

Why the polar bro? Isn't the reading a simple linear rate of what passes the sensor? Such that the extra tooth is changing the frequency as Lando has stated (omega constant, divisions increased)... The sensor doesn't care that the tone ring is rotating or translating.
 
Gimme a little credit Karl, yes, I DO get what Lando is talking about. I'm not the average high school dropout you may think I am. There are 2 schools of thought...the engineering side that every last .00001% is critical, and the real world side where .005% works just fine. There are various ways to come to a similar solution, or an acceptable solution that works for 99% of the masses.

As for Lando's case, with the 12-tooth tone ring he has far more room to work with in terms of spacing the teeth out, compared to the 48-tooth tone ring I have, so the solutions are easier to come up with.

Just sayin'.
 
You must log in or register to reply here.
Back
Top