Driveshaft Speed Sensor

[jayb]

I'm in the middle of making a few changes to my race car, and thought that this one might interest a few people.  Last year the car was running way to much RPM, about 7700 through the traps during Drag Week, where power peak on the engine is around 6600-6800.  I wanted to bring the RPM down some, so the first step was to change out the 4.29 gears for some 4.11s.  In addition, the street tires I'm running are only 28" in diameter, and with the radiused wheelwells they look just a little small when they are on the car.  So, I'm upgrading to a set of 30" tall tires.

These two changes taken together make it practical for me to drive the car on the street without the Gear Vendors overdrive.  So, along with these changes I decided to remove the GVOD to save some weight, plus add an aluminum driveshaft to save a little more.  The transmission is an ATI Superglide, but when I bolted the normal tailshaft housing onto the trans in place of the GVOD unit, one problem immediately became apparent:  The ATI transmission tailshaft housing has no provision for a speedometer.

With as much street time as this car sees I didn't want to be running around without a speedometer, so I spent a little time investigating my options.  I have a complete set of Autometer gauges in the car, including the speedometer, and when I checked Autometer's web site I found that they had an electronic speedometer available in the same style as my other gauges.  After downloading the directions for the electronic speedo, I figured out that I could use the same kind of magnetic sensor as a pickup for the speedometer as I'm already using for cam and crank sensors on the engine.  So this looked like a pretty good fit.

The byproduct advantage of the whole issue was that I would get a driveshaft speed sensor installed in the car.  This could work with the speedometer, and the output could also be datalogged by my EFI system.  A driveshaft speed sensor is really a nice sensor to have, because it lets you determine if the car is spinning the tires at the track, and also gives good information to help you dial in your clutch or torque converter.  I've been threatening to install one for quite a while now, but there was going to be a lot of fabrication required, so I hadn't gotten it done.  This weekend I decided to do it.

My business background is in magnetic sensors, and I know that the gear or "target" that the sensor has to pick up is critically important for good performance.  The sensors that I'm using, which are Cherry GS1001 Hall effect sensors, give a suggested gear tooth profile in the data sheet, so I used that as a starting point and designed a six tooth gear that would clamp in place around the driveshaft yoke.  I wrote a CNC program to machine the target in two parts, and then after cutting them out of cold rolled steel I drilled and tapped them so they would bolt together around the yoke.  Here's a picture of the target:



Next I had to design and build a bracket that would hold the sensor.  Attaching it to the chassis didn't make a lot of sense to me, because the transmission mount is rubber, and the air gap between the sensor and the teeth of the target is pretty small, around 0.060".  The logical place to mount the sensor was with a bracket attached to the tailshaft housing of the transmission, and fortunately for me this is a machined section on the ATI Superglide.  I made up a drawing of what I wanted then machined the first part of the bracket out of aluminum.  As often happens with my first time machining programs, a couple of minor machining errors crept in, but the part was still usable; picture below:



Note the raised portion on one of the bracket halves; I needed to machine another small piece to bolt to that, and hold the sensor.  The sensor and the cable I'm using are shown in the photo below:



I like this sensor because the cable screws onto it; in case of a sensor failure, it can be disconnected from the cable and easily replaced.  Also, the head of the cable has two LEDs built into it.  One is a green LED that indicates the sensor is powered.  The other is a yellow LED that lights up when there is no tooth present in front of the sensor, and goes out when the tooth is present.  This makes it really easy to diagnose any sensor problems (or problems with the cable, for that matter). 

The body of the sensor is 12mm in diameter (about 0.470"), so I designed a bracket that would bolt onto the first one I made, and had a hole in it for mounting the sensor.  Here's a pic:



Finally I assembled the whole thing on the car, while adjusting the sensor position for the desired 0.060" airgap.  The mounting appears very solid.  I have not yet wired it up, but I don't expect any issues with making it work.  Here's a picture of the whole thing assembled on the car:



This whole project took about 14 hours over the weekend, and is probably overkill, but I'm confident that it will survive in the vehicle.  I'm anxious to try this setup out in the next month or so, when I start driving the car again.  I will report back with any issues I find - Jay

[My427stang]

Nice work!

[cwhitney67]

Nice work Jay

[Drew Pojedinec]

I just installed an aftermarket powerglide tailhousing, it had a roller bearing for the yoke and lacked the speedo.  I thought it had a boss tho that could be drilled/tapped.

Anyhow, now that you have a speed sensor on the driveshaft, have you considered using it to sense tire slippage and possibly change timing to retain traction?

[jayb]

That's exactly what I plan to do at the track, although I think I will be focusing on suspension adjustments to eliminate any tire slippage first, before any timing adjustments.  Between this sensor and a Go-Pro looking at the suspension under the car, I think I ought to be able to dial in the suspension pretty quickly...

[Drew Pojedinec]

Makes sense.
Tho with a proper setup on the sensor, it could sure make you dangerous and surprising on street tires

[machoneman]

Jay, always the innovator! 

Just watch, his next step will be a algorithm linked to the driveshaft sensor and the EFI controller that, by sensing tire slip, instantly kills some timing or like a stutter box, drops some cylinders to kill power!.   

[Falcon67]

Good engineering!  Overdone is just right for reliability IMHO.  I might try that sometime on my dragster using my "CNC" (Chris Noodling the Controls)  I would just comment that the Falcon ate the speedo cable around 1976 or so.  All those "contest of speed" things on Trinity Blvd I think.    I just chunked it and drove with the tach until I lost the motor sometime in 1980.  That was daily driving to and from work.  Actually, It's a street/strip car now and - it still doesn't have a speedometer.

I've been having some fun with 1/4 aluminum sheet what with motor plates and such.  I thought "lets try this" and bought a 1/2" bi-metal blade for my 10" C-man wood bandsaw.  I don't cut much wood but boy that saw will carve AL now like a boss. 

[shady]

could it double as a rev limiter just in case the tranny or rear breaks? that would be cool so as too not take the engine against a distributor limiter.

[cjshaker]

Between this sensor and a Go-Pro looking at the suspension under the car, I think I ought to be able to dial in the suspension pretty quickly...

You beat me to it. I was going to suggest the use of a GoPro. The Hero 4 allows for some very nice slow motion video, and that can be extremely useful for suspension diagnosis. I'm beginning to become a little bit of a GoPro addict since my Son bought me one to use on my car. And nice work on the speed sensor.

[cjshaker]

could it double as a rev limiter just in case the tranny or rear breaks? that would be cool so as too not take the engine against a distributor limiter.

I don't think that would work since it only senses driveshaft rpm. The motor could be at 10,000 rpm while the driveshaft is only reading 40mph. And it would still have to limit the motor RPMs by way of some sort of ignition cutout.

[BruceS]

Jay, does the sensor also dictate the number of teeth in addition to the profile of the target gear?  If not, how did you arrive at that?

Thanks,
Bruce

[machoneman]

Between this sensor and a Go-Pro looking at the suspension under the car, I think I ought to be able to dial in the suspension pretty quickly...

You beat me to it. I was going to suggest the use of a GoPro. The Hero 4 allows for some very nice slow motion video, and that can be extremely useful for suspension diagnosis. I'm beginning to become a little bit of a GoPro addict since my Son bought me one to use on my car. And nice work on the speed sensor.

Yes, YouTube has some vids of a drag car's suspension and how it reacts. Pretty cool stuff. I just wish we had same years ago while racing when all we had were video cameras someone held near the starting line to record the car coming off the line. Btw, this was very helpful but not nearly as good as seeing the suspension's movement from under the car.

[57 lima bean]

Hemi Joel probably had an old Direct Connection unit lay'n around.Remember those?

[jayb]

Jay, does the sensor also dictate the number of teeth in addition to the profile of the target gear?  If not, how did you arrive at that?

Thanks,
Bruce

I wanted to put as many teeth on the target as possible, because the sensor can handle frequencies of up to 10 KHz (10,000 cycles per second).  At 8000 RPM, the driveshaft is spinning 133 times per second, so 6 teeth is only 800 cycles per second.  What limited me to 6 teeth is the requirement to have the target in two halves, so it would bolt together and clamp on the driveshaft yoke.  The bolts would get in the way with more than 6 teeth on the target, so I settled on that.  That number is actually larger than some I've ready about; for example the driveshaft sensor sold by FAST uses two magnets on the yoke, essentially.  It probably works fine, but I figured more resolution was better so I went with six.