Wheel speed sensor questions - custom application

[Toilet Duck]

Afternoon all,

I'm running a heavily modified Ariel Atom and am currently seriously contemplating upgrading the ECU to a Syvecs S7Plus "plug and play" (link HERE). It's designed for the Honda Civic Type R FN2 which uses the same basic running gear as my Atom, my tuner that I normally use has confirmed it is compatible but may need some minor alterations to the loom adapter which comes with it. As the Syvecs ECU supports traction control, if I can get traction control working on my Atom then the man maths becomes easier to justify the expense of this ECU upgrade, and I'd be happier to bite the bullet and go for it

My Atom does NOT have any suitable wheel speed sensors for traction control (it only has a single magnet glued on to one CV joint for the speedo which won't help here

I have proven that, with some very minor "adjustment," a toothed ABS reluctor ring from a Chrysler Voyager will fit over the rear inner CV joints (offside and nearside) and I'm sure I can secure these in position.These rings each have 47 teeth.



My first question is, do the number of teeth used for speed detection have to be the same across front and rear axles? ie if I go with 47 toothed rings on the rear, do I have to come up with a solution that uses 47 teeth on the front? I would hope that an ECU is "intelligent" enough so that you just input the number of teeth front/rear and it works out the maths/algorithm to calculate a speed comparison across the two axles. The Atom uses 15 inch front and 16 inch rear alloys so the rolling circumference must be different anyway, so having the same number of teeth won't give the same readings surely?

Looking at the front axle I am very limited to what I can do. The upright is Atom specific and not a parts bin special used off another car. It has no provision for an ABS ring/magnet etc built in with the bearing housing etc

Here is a couple pics of a front upright (it is an older version, the newer ones as currently fitted have a bigger bearing):





Here is a currently fitted front upright:



I'm thinking that there are possibly three options to fit a trigger wheel

1 - The bit in the middle of the upright that rotates as the wheel turns is approximately 57mm in diameter and protrudes slightly from the face of the upright. If I could find a toothed ABS ring around 57mm external diameter, I'm wondering if it's possible to glue that in position? I'm hesitant as I don't know if a glue would be strong enough. The single magnet for the speedo sensor is fixed to a CV joint using Araldite, and that has never come free.

2 - I'm wondering if the bolts securing the brake discs to the bell could be used as triggers? The downside of this is there are only around 8 - 10 on the circumference, so I don't know if that's enough to give a high enough resolution? It would be 8- 10 pulses per front wheel revolution compared to 47 pulses per rear wheel revolution. I don't know if a difference that big would impact that traction control operation?

3 - The brake discs are vented, so between each of the two sides of the brake disc there are the vents/vanes sandwiched inbetween. I'm wondering if a sensor (assuming one would fit between the alloy wheel rim and brake disc) would be able to detect the vanes? That would give a better resolution that using the brake disc bolts. I haven't counted how many vanes there are yet but I'd guess maybe in the region of 30 plus, I need to take a wheel off and count. To clarify, I'm not referring to the grooves machined into the brake disc face where the brake pads sit.

The final question is the actual ABS/wheel speed sensors. Obviously I'm going to have to fabricate brackets to mount/hold them, but what type of sensor should I be looking at? I see that there are both "passive" and "active" versions, as well as other different permutations and standards. Can anyone help with regards to what sort of sensor I should be looking at? Not only must they detect the pulses, but they must output a signal that is compatible with the ECU.

I appreciate that all of this is quite a very specific issue and rather niche. If anyone knows of any other forums etc that may be able to help (if pistonheads can't come to the rescue) I'd appreciate any links etc

Many thanks

[stevieturbo]

Really you can use as many or as little as you want*


You will tell the ecu number of teeth at each wheel, so no big deal, but it can make some sense to keep them all similar for resolution purposes.

If the hubs you are using had an OEM option on some models, using that will also make a lot of sense.

You just want either a simple 2 wire VR, or preferred 3 wire hall sensor. GT101 was a common one, although Honeywell discontinued that, but other makers now copy it.

A lot can depend on space where you intend to mount it though. Most will want a straight shot at the teeth, unless the teeth are physically large enough to look at the side.
Fully support the sensor, do not rely only on that little bracket.

https://uk.farnell.com/honeywell/1gt101dc/sensor-3...

Haltech same https://www.haltech.com/product/ht-010609-gt101-st...
https://rrrshop.co.uk/products/hall-effect
https://www.speedingparts.co.uk/p/brands/ross/gt10...

Some people like threaded sensors like this as it gives a bit more flexbility.

https://www.haltech.com/product/ht-010610-red-sing...
https://motionraceworks.com/products/rife-drivesha...

etc etc etc etc etc etc

Also worth considering....most OE hall effect crank sensors would work just fine too. Not always cam, but most of them would too.

You can do VR as well, but these may not respond just quite as fast from standstill.

You can also use the brake disc bolts you show in the photo. I've used these on one of my cars for years and they work fine with the GT101, 8 bolts on the driven wheels.

Other car I just tapped into the OE VR ABS wires to pickup speed

[stevieturbo]

And if it makes it a little easier, you only really need one undriven wheel sensor ( unless your wheels go airborne a lot )

And for the rear, even if you had a pickup on the diff/crownwheel that you could use as an averaged rear wheel, that too will work.

As ultimately even with 2 driven wheel sensors, you'll average them anyway for the "driven speed" you will use for the TC system

[Toilet Duck]

Stevieturbo, thank you very much indeed for all the info and links, that's a massive help to me!

At the moment, only the front discs have separate bells and therefore the mounting bolts. The rear brakes do NOT have those same bolts. If I use those 8 or so bolts as position indicators for the front wheels, would it be too much of a difference if I used the 47 tooth ABS ring on the rear? Or would I be better grinding some of the teeth off to close the gap between front and rear trigger numbers?

I will go with hall effect sensors as per your recommendation. Can I test the operation using a multimeter or would I need access to an oscilloscope? The threaded sensors you linked to will probably be best as I can see them being easier to set up in terms of proximity to the teeth/bolt trigger, I imagine you can adjust the gap to a high tolerance with the threaded bit.

I'll get some sensors ordered and then fabricate some sturdy brackets. If that's all successful then I'll bite the bullet and order the Syvecs

Again, many thanks indeed

[GreenV8S]

stevieturbo said:

You can also use the brake disc bolts you show in the photo. I've used these on one of my cars for years and they work fine with the GT101, 8 bolts on the driven wheels.

I know somebody who was getting a fuzzy signal from that, until it occurred to them to grind the '8.8' off the bolt heads.

[Toilet Duck]

Me again

I've been researching those sensors that stevieturbo kindly linked to. The toothed ABS ring that I've proved will fit my rear inner CV joints has teeth that are approx 3.4mm wide, a gap of approx 3.5mm between each tooth, and teeth that are approx 2.3mm in height. Not all the hall effect sensors list data on the specs of toothed trigger wheels that are compatible, but the ones that do suggest that my trigger wheel will NOT work.

This is from the Haltech website for their GT101 sensor:



This is from the Farnell website for the Honeywell 1GT101DC sensor:



As you can see, both of these suggest they will not work with my ABS trigger wheel. They should however work fine on the front axle if I use them to pick up the bolts for the brake disc mounting. The rear discs are currently one piece so do not have mounting bolts. I could change those discs to floating versions exactly the same as the front axle (they are the exact same size discs front and rear on my car), but that's going to cost circa £600 and there is nothing wrong with the current rear discs in terms of wear etc, so it's rather a lot of money just to get some bolts for triggering purposes...

Are these hall effect sensors really that fussy about tooth spacing/size/height etc? I'm assuming they must be as the manufacturers wouldn't publish such strict criteria otherwise? Or is the problem the ABS ring I've picked, does it have some weird type of tooth sizing/spacing? I have next to no experience with ABS trigger wheels and sensors.

This is the equivalent ABS ring I've been playing with, is it a typical design?:

[vw_99]

https://www.southwalesprops.co.uk/propshaft/consum...

Might be worth a look for options on abs ring sizes.

[witko999]

Can you not get your own toothed discs designed and then laser cut? Laser cutting costs next to nothing really and saves you searching for something that might fit.

[Toilet Duck]

vw_99 said:

https://www.southwalesprops.co.uk/propshaft/consum...

Might be worth a look for options on abs ring sizes.

Thanks very much for the link. I've had a look, there is one ABS ring listed that is a "possible". The inner diameter would need increasing by around 2mm, but the number of teeth is 27 which is 20 less than my current ABS ring, so that should give a better tooth profile for the sensor to pick up.

[Toilet Duck]

witko999 said:

Can you not get your own toothed discs designed and then laser cut? Laser cutting costs next to nothing really and saves you searching for something that might fit.

I hadn't considered this, thanks very much. I will have a look into this as it's not something I have any prior experience of.

[stevieturbo]

Toilet Duck said:

Stevieturbo, thank you very much indeed for all the info and links, that's a massive help to me!

At the moment, only the front discs have separate bells and therefore the mounting bolts. The rear brakes do NOT have those same bolts. If I use those 8 or so bolts as position indicators for the front wheels, would it be too much of a difference if I used the 47 tooth ABS ring on the rear? Or would I be better grinding some of the teeth off to close the gap between front and rear trigger numbers?

I will go with hall effect sensors as per your recommendation. Can I test the operation using a multimeter or would I need access to an oscilloscope? The threaded sensors you linked to will probably be best as I can see them being easier to set up in terms of proximity to the teeth/bolt trigger, I imagine you can adjust the gap to a high tolerance with the threaded bit.

I'll get some sensors ordered and then fabricate some sturdy brackets. If that's all successful then I'll bite the bullet and order the Syvecs

Again, many thanks indeed

You can sort of test with a meter, crank/cam,but same odds. ( as said, there will be many crank sensors you could use too ) https://www.youtube.com/watch?v=TXKf96wqaoU

But testing at low speeds, may not always offer the same result at high speeds.

Really though, if push came to shove you could just grind half of the teeth off ( assuming even numbers...not quite sure with odd numbers...or maybe grind off a multiple to ensure even spacing )

What sort of ABS sensor does the vehicle use that you're intending to get the trigger wheel from ? If 3 wire, you can be pretty sure it will be a simple hall, so just use that ?

If 2 wire, it's not so clear as you can get VR and MR. MR is not usable directly with Syvecs, although you could get their X4 expander which can take in MR signals ( although that's usually piggy backing onto existing setups, not sure about 100% standalone )

Although I say VR can be a little slower to pickup from standstill ( as it requires velocity of movement )....even on my car using the OE ABS rings, I can get a good signal immediately rotating the wheel by hand.

Various types covered here.

https://www.youtube.com/watch?v=MTWvdIOcvRg

I hear what you're saying about the GT101 but I haven't heard of anyone really saying theirs didn't work on whatever setup they use, even people using it as a crank trigger wheel on what appeared less than ideal ring.

But there are literally hundreds, thousands of sensor options. But if you can find out what is normally mated to that ring of yours, it will be a good starting point..

[stevieturbo]

And again....does your gearbox have a speed pickup at the crownwheel ? You should be able to use that

And no, different tooth counts on each wheel does not matter, you will tell the Syvecs what each wheel has.

Obviously there will be a point where too few is too little resolution, not sure if there is a physical reality where you can have too many teeth.

I've one car with 11 at the front and 8 at the rear ( front originally a Subaru tone ring that had 44 teeth, but back when I used DTA it couldn't accept that many teeth so I ground a load off )
I just retained the same with Syvecs.. And other car at present just uses OE bearing/sensor setup with 29T and VR sensor.

Subaru rings aren't bad actually as they can be mounted to the rear of the driving flange if you can get it drilled etc.

eg ( rear, front is different )

https://www.importcarparts.co.uk/product/genuine-s...

[stevieturbo]

witko999 said:

Can you not get your own toothed discs designed and then laser cut? Laser cutting costs next to nothing really and saves you searching for something that might fit.

He could reach out to Graham.

He's done a few for himself and other guys. 3d printing a plastic housing to house metal teeth, or just a custom ring itself.

https://www.zetecinside.com/xr2/shop.shtml

[Toilet Duck]

Stevieturbo, many thanks for posting again

stevieturbo said:

You can sort of test with a meter, crank/cam,but same odds. ( as said, there will be many crank sensors you could use too ) https://www.youtube.com/watch?v=TXKf96wqaoU

Thats a good link, thank you

stevieturbo said:

What sort of ABS sensor does the vehicle use that you're intending to get the trigger wheel from ? If 3 wire, you can be pretty sure it will be a simple hall, so just use that ?

It looks like THESE are the ABS sensors to go with the ABS ring I'm currently experimenting with



It states its Hall effect, but also that its only 2 wire.

stevieturbo said:

And again....does your gearbox have a speed pickup at the crownwheel ? You should be able to use that

I don't know is the honest answer. The speedo display uses a magnet glued onto a CV joint, but the current (standard) ECU gets its speed reading from elsewhere. Not sure if there is another speed pickup or if the ECU works out the speed referencing RPM against what gear you're in.

stevieturbo said:

He could reach out to Graham.

He's done a few for himself and other guys. 3d printing a plastic housing to house metal teeth, or just a custom ring itself.

https://www.zetecinside.com/xr2/shop.shtml

I'll reach out to this chap. That's pretty clever how he's designed them with a bit of flex to "stretch" over what you are mounting it too, a clever idea! I will re-measure to get some exact measurements and ping him an email. Thanks for the link, I would never have come across that myself!

Once again, really appreciate the help!

[stevieturbo]

Just look at the gearbox, there either is a sensor or there is not.

Can you view the speed reading on the ecu ? does it stop if you unplug the single tooth sensor ?


And a quick google and I get the impression that sensor is an MR type, not something you can easily work with.

Speak to Graham, easiest option will be just get him to cut you some wheels that fit nicely, say 20-30 teeth or so, well defined with nice gaps, and that will keep most VR and Hall type sensors happy.

His little GT101 bracket is quite nice too, offers good support and adjustability for gap.

[littleredrooster]

As an interested bystander (I have absolutely no expertise in this field), but purely as an observation/suggestion, would a sensor ring from a motorcycle be of any use? Most are perforated stainless steel discs riveted to the wheel with the sensor mounted on the forks/swing arm/brake caliper etc.

[Toilet Duck]

stevieturbo said:

Just look at the gearbox, there either is a sensor or there is not.

Can you view the speed reading on the ecu ? does it stop if you unplug the single tooth sensor ?


And a quick google and I get the impression that sensor is an MR type, not something you can easily work with.

Speak to Graham, easiest option will be just get him to cut you some wheels that fit nicely, say 20-30 teeth or so, well defined with nice gaps, and that will keep most VR and Hall type sensors happy.

His little GT101 bracket is quite nice too, offers good support and adjustability for gap.

Hi again,

Just had been googling and it appears that there is a speed sensor on the gearbox, and there has been a few revisions over the years as to the type. I've had a look myself and confirmed a sensor is present.

I will try and disconnect it tomorrow and drive it with the ECU data logging to see if the speed is lost.

I've emailed Graham about his custom rings , he's already been back in touch so that's promising. He has mentioned that he can supply NPN 12 volt barrel induction sensors to go with his rings. I've asked if they can produce a signal from minimal rpm the same as hall effect sensors. I like the idea of the ECU getting an input without the need to first travel a distance

[Toilet Duck]

littleredrooster said:

As an interested bystander (I have absolutely no expertise in this field), but purely as an observation/suggestion, would a sensor ring from a motorcycle be of any use? Most are perforated stainless steel discs riveted to the wheel with the sensor mounted on the forks/swing arm/brake caliper etc.

Interesting idea, but I'm quite constrained with space. I think one of those bike sensor rings would be relatively large and potentially difficult to mount in the area I've got to work with

[stevieturbo]

Toilet Duck said:

Hi again,

Just had been googling and it appears that there is a speed sensor on the gearbox, and there has been a few revisions over the years as to the type. I've had a look myself and confirmed a sensor is present.

I will try and disconnect it tomorrow and drive it with the ECU data logging to see if the speed is lost.

I've emailed Graham about his custom rings , he's already been back in touch so that's promising. He has mentioned that he can supply NPN 12 volt barrel induction sensors to go with his rings. I've asked if they can produce a signal from minimal rpm the same as hall effect sensors. I like the idea of the ECU getting an input without the need to first travel a distance

He uses a Life ecu, same as Syvecs really. It's not so much distance as such, but as a VR sensor gets stronger with rotational speed, say if you were to literally crawl along, the voltages seen may be quite low. Although in such circumstances, it hardly matters.

The only issue where it really matters is when using it for launch control and you use this motion to switch from staged launch to the launch ramp. But even then, as this will happen pretty abruptly, it still shouldn't be an issue.

But Hall will give full strength waveform at any speed for every tooth.

Even here with my OEM ABS, simply turning the wheel by hand, you can see first tooth or two is not as strong as those once speed picks up. But they are still distinct enough to be seen.



If there is a sensor in the gearbox, great, use it. Although it would also do no harm to add wheel speed sensors too, that way you can see when one wheel is spinning much easier.
And for TCS you still need an undriven front wheel speed anyway.

[stevieturbo]

littleredrooster said:

As an interested bystander (I have absolutely no expertise in this field), but purely as an observation/suggestion, would a sensor ring from a motorcycle be of any use? Most are perforated stainless steel discs riveted to the wheel with the sensor mounted on the forks/swing arm/brake caliper etc.

This would seem a little odd, as usually all sensors need a ferrous material, ie magnetic in order to be seen. Maybe they use a low grade ? or a special sensor ?