Some ABS questions...

<begin useless hydraulic/electronic engineering mumbo jumbo>

I don't have ABS, but after reading about the way some manufacturers are doing traction control with their brake systems (which I will term "ghetto lockers") by applying the brakes on the spinning wheels, I have some thoughts... I can't afford or justify lockers, but if I can hack up something that works from junkyard parts I might try it. Anyone know if the XJ OEM ABS system controls each wheel individually, or if it modulates the brake pressure to the whole system at once? I'm basically trying to figure out if I can install an ABS modulator from the junkyard and set it up with a control system I can use to individually apply the brakes to each wheel on command. From the pictures I'm seeing online of the HCU (hydraulic control unit, aka ABS modulator) the XJ unit only has 3 flare fittings, so my bet is no, but I'm hoping I am wrong.

It treats the rears as one unit - hence only 1 brakeline.

Hmm... I should have thought of that. Any idea on the fronts?

Lockers would be way easier

kastein said:

<begin useless hydraulic/electronic engineering mumbo jumbo>

I don't have ABS, but after reading about the way some manufacturers are doing traction control with their brake systems (which I will term "ghetto lockers") by applying the brakes on the spinning wheels, I have some thoughts... I can't afford or justify lockers, but if I can hack up something that works from junkyard parts I might try it. Anyone know if the XJ OEM ABS system controls each wheel individually, or if it modulates the brake pressure to the whole system at once? I'm basically trying to figure out if I can install an ABS modulator from the junkyard and set it up with a control system I can use to individually apply the brakes to each wheel on command. From the pictures I'm seeing online of the HCU (hydraulic control unit, aka ABS modulator) the XJ unit only has 3 flare fittings, so my bet is no, but I'm hoping I am wrong.

Click to expand...

Suzisamurai said:

Lockers would be way easier

Click to expand...

I can't afford or justify lockers, but if I can hack up something that works from junkyard parts I might try it. I figure if I can spend 50 bucks on hardware and come up with something that works "good enough", that's good enough for me.

I know I've seen ABS modulators with a LOT of hydraulic lines on them - anyone know anything specific? Any of the other engineer types on here ever worked on or designed some of those? I have no idea if they'd produce enough braking force to be worth it or if I'd be turning my rig into a brake-less deathtrap, so I'd like to stay with large SUV parts only, but if anyone has ideas, I'm listening.

My guess on the fronts is that it modulates them independently but I cannot say for certain. There are separate lines heading to each front wheel.

The easiest and lowest-tech way to do this might be to hook up a line lock to each of the front lines. When 1 tire spins: touch the brakes, lock that one line with slight pressure, and wheel out. Although two line locks is probably close to the cost of a used locker.

It appears as though the fronts are actuated independently, whereas the rears are on a single circuit. Section 5, p.33 of the 2000 FSM states the following: "The ABS system has three fluid pressure control channels. The front brakes are controlled separately and the rear brakes in tandem." This makes sense: if you follow the rear hydraulic circuit back from the master cylinder, it's the same as on a non-ABS model - just a line going back, a T-fitting, and a drum off at either side of that T-fitting. There would have to be extra hardware in there somewhere (either a second line to the rear axle from the master cylinder, or some sort of proportioning valve / secondary brake cylinder) on the axle to handle modulating each drum separately.

Awesome info, thanks for clearing that up. Wonder if I can rob some poor WJ or something of its ABS modulator... I'd really like to try this with all 4 wheels independently controllable.

EDIT: I'd obviously have to run a second brake line to the back, since I have an 8.8 going in "soon" I'll be working on the rear brake line(s) anyways... hmmm

EDIT2: it looks like the WJ (at least some years) has 5 flare nut ports on the ABS modulator, so it may be suitable for use in this harebrained scheme

kastein said:

EDIT2: it looks like the WJ (at least some years) has 5 flare nut ports on the ABS modulator, so it may be suitable for use in this harebrained scheme

Click to expand...

This is a pretty interesting idea; it opens up the possibility of traction control in addition to the traction enhancements we're already familiar with. Please don't take any of this as trying to dissuade you; I'm just thinking out loud.

Using the WJ master cylinder and running an extra line to the rear would probably be the easiest route to take. Conceivably you could keep the existing single-line setup and replace the T-fitting with something to handle modulating the split to the left & right drums, but how to package and control that then becomes the issue - and I really have no good suggestions for how or what to put in that T-fitting's place. It just doesn't seem like the best way of doing it when the 4-channel master cylinder is already there.

There may be hydraulic pressure issues with the XJ lines, most likely at where they enter fittings or joins. The WJ being heavier, it may have run a higher-pressure system, and that system was designed for a 4-wheel-disc system; the cycle time on drums between being applied and freed up may not be sufficiently fast to let the system work as intended.

You may want to check the differences between the XJ and WJ ABS units. It looks as though the XJ has the ABS controller on-board with the hydraulic controls, effectively making it a closed system wholly-separate from the rest of the stuff run by the ECU. If the WJs are the same, will it be necessary to replace the entire unit, and if so, will XJ sensors and tone rings provide readable pulses to the controller?

Again, not trying to tell you why it can't be done, just thinking out loud. Definitely want to see where this one goes.

casm said:

This is a pretty interesting idea; it opens up the possibility of traction control in addition to the traction enhancements we're already familiar with. Please don't take any of this as trying to dissuade you; I'm just thinking out loud.

Click to expand...

No offense taken... this is exactly what I was hoping for, people thinking out loud and hopefully coming up with something that helps.

casm said:

Using the WJ master cylinder and running an extra line to the rear would probably be the easiest route to take. Conceivably you could keep the existing single-line setup and replace the T-fitting with something to handle modulating the split to the left & right drums, but how to package and control that then becomes the issue - and I really have no good suggestions for how or what to put in that T-fitting's place. It just doesn't seem like the best way of doing it when the 4-channel master cylinder is already there.

Click to expand...

Yeah, that's my thought process also. Running another line is really no big deal, since it's just 3/16" brake line, been there, done that.

casm said:

There may be hydraulic pressure issues with the XJ lines, most likely at where they enter fittings or joins. The WJ being heavier, it may have run a higher-pressure system, and that system was designed for a 4-wheel-disc system; the cycle time on drums between being applied and freed up may not be sufficiently fast to let the system work as intended.

Click to expand...

Yeah, I was thinking about this, but I don't think it'll be an issue - it's all the same 3/16" brake line, and people do higher power master cylinder+booster swaps all the time, as well as disc brake swaps. I'm working on an 8.8 swap right now and it comes with discs, so that shouldn't be an issue, though the 8.8 has the tone ring on the carrier right next to the ring gear, I'm not exactly certain how that does anything other than measure driveshaft rotations.

I think I'm going to cruise the JY tomorrow morning if I wake up early enough, or saturday morning if I don't. A WJ ABS modulator (if I can find a WJ there... hahahaha right) is definitely on the list, or failing that, any SUV modulator with individual outputs for each wheel. Initially I was thinking that I'd have to just throw switches and select which wheel to apply the force to, but now I'm overthinking this whole idea and wondering if I can attach the OEM tone ring pickups to the proper wheels and build something that will determine when one wheel on an axle is spinning faster than the other, and use a "ghetto lock this axle" switch and the brake pedal to sorta-lock that axle by applying the proper brake until the pulse trains are roughly the same frequency again.

If anyone else wants to give this a try and post up results too, more power to you guys, I'm an ABS ignoramus so far and just starting to learn and lay no claim to this idea. Bear in mind though, I'm not taking responsibility for this kind of contraption drinking all the beer in your fridge, crashing your XJ, and sleeping with your wife, that's all your fault if you decide to run with this idea.

kastein said:

Yeah, I was thinking about this, but I don't think it'll be an issue - it's all the same 3/16" brake line, and people do higher power master cylinder+booster swaps all the time, as well as disc brake swaps.

Click to expand...

Good points, and typically all that's needed for the rear disc swap is to replace the proportioning valve to the rear hydraulic circuit (I'm running a Wilwood adjustable one on my D44 rear discs). Sounds reasonable.

I'm working on an 8.8 swap right now and it comes with discs, so that shouldn't be an issue, though the 8.8 has the tone ring on the carrier right next to the ring gear, I'm not exactly certain how that does anything other than measure driveshaft rotations.

Click to expand...

Thinking about it, this sounds like a somewhat simplified version of how the XJ ABS works in practice: even though the XJ measures speed at each rear wheel, it can't actuate the rear brakes independently of each other, so the end result (in braking terms) is probably very close to if not the same as just measuring driveshaft rotations. What I'm getting at is that while there might be a more accuracy in the XJ setup, the advantage of that is lost by not being able to actuate the rear brakes separately.

Just one thing regarding the tone rings: you'll need to make sure that the overall ring diameter, tooth count, and tooth spacing are the same on whatever tone rings you're sampling from as what the ABS controller expects to see - pulse width & frequency will be off otherwise, and it could cause it to work unpredictably or go into failure mode. Obvious, I know, but thought it may be worth mentioning.

I think I'm going to cruise the JY tomorrow morning if I wake up early enough, or saturday morning if I don't. A WJ ABS modulator (if I can find a WJ there... hahahaha right) is definitely on the list, or failing that, any SUV modulator with individual outputs for each wheel.

Click to expand...

I wonder if an LH-series car ABS unit may not be the same. Dakotas might be a possibility, too, though I seem to recall that the ABS system in those only operated on the rear wheels.

Initially I was thinking that I'd have to just throw switches and select which wheel to apply the force to, but now I'm overthinking this whole idea and wondering if I can attach the OEM tone ring pickups to the proper wheels and build something that will determine when one wheel on an axle is spinning faster than the other, and use a "ghetto lock this axle" switch and the brake pedal to sorta-lock that axle by applying the proper brake until the pulse trains are roughly the same frequency again.

Click to expand...

Or build an inline controller to do exactly that. Feed the ABS sensor outputs to the inline unit and have it operate in two modes: regular and traction control. From there, the inline unit either passes a straight signal from the sensors ('regular' mode) back to the ABS controller, or modifies the signal ('traction control' mode) before sending it back out.

There's also the option of modifying the software in the ABS unit's on-board controller to handle this, but I don't know how feasible or appealing that is.

casm said:

Thinking about it, this sounds like a somewhat simplified version of how the XJ ABS works in practice: even though the XJ measures speed at each rear wheel, it can't actuate the rear brakes independently of each other, so the end result (in braking terms) is probably very close to if not the same as just measuring driveshaft rotations. What I'm getting at is that while there might be a more accuracy in the XJ setup, the advantage of that is lost by not being able to actuate the rear brakes separately.

Click to expand...

Yeah, I always found that a little odd. The XJ system with two sensors can actually measure wheel slip I guess (if you assume that the wheel turning faster is not slipping, valid only on surfaces with a certain minimum amount of traction and if the ABS controller catches on before both wheels on the axle start slipping) but the Exploder system literally seems entirely useless to me, since the wheels could be spinning at 80mph in different directions and the sensor would read absolutely zero rotation since it's on the carrier/ring gear. It really seems like a complete duplicate of the vehicle speed sensor to me since it's directly linked via the ring gear, pinion, and driveshaft.

casm said:

Just one thing regarding the tone rings: you'll need to make sure that the overall ring diameter, tooth count, and tooth spacing are the same on whatever tone rings you're sampling from as what the ABS controller expects to see - pulse width & frequency will be off otherwise, and it could cause it to work unpredictably or go into failure mode. Obvious, I know, but thought it may be worth mentioning.

Click to expand...

Yeah, good point. I guess if I do this with the rear axle I'll have to find a way to install XJ tone rings on the 8.8 axleshafts along with XJ reluctor sensors.

casm said:

I wonder if an LH-series car ABS unit may not be the same. Dakotas might be a possibility, too, though I seem to recall that the ABS system in those only operated on the rear wheels.

Click to expand...

I'll have to take a look if I don't find something else suitable first.

casm said:

Or build an inline controller to do exactly that. Feed the ABS sensor outputs to the inline unit and have it operate in two modes: regular and traction control. From there, the inline unit either passes a straight signal from the sensors ('regular' mode) back to the ABS controller, or modifies the signal ('traction control' mode) before sending it back out.

Click to expand...

Pretty much what I was thinking - though I don't have factory ABS and intend to keep it that way, so I was just thinking a standalone controller.

casm said:

There's also the option of modifying the software in the ABS unit's on-board controller to handle this, but I don't know how feasible or appealing that is.

Click to expand...

ohhhhhh so tempting. I've got far too many reverse engineering projects going on already though, and I'd almost be willing to bet that Chrysler chose yet another microcontroller architecture/ISA I am not familiar with for the HCU, since they did that with the TCU, ECU, and OHC microcontrollers already.

I figure first step is to find hardware that'll work and build a proof of concept that requires me to flip a lot of toggle switches and read a lot of gauges to get anything done, then get fancy with it afterwards, otherwise it'll never go anywhere.

Didn't manage to get to the junkyard today, had to head for work by 8AM and only just recently got out of the office. Tomorrow and Friday are looking the same, so Saturday it is...

why not just install a hand operated turning brake system like what alot of sand cars use? it taps into the rear brakes individually with no interuption when useing the foot pedal. it has two side by side levers that can be pulled by hand. or there is a single style aswell that you either pull back for one side, or push forward for the other. i prefer the two side by side levers myself. basically it would be like having two seperate parking brake levers, but do not lock in place. they are very easy to install. just tap into the hydrolic lines as the seperate to each rear wheel cylinder. you can purchase them from most any buggy shop. CNC or NEAL are most common brands.

2stix: I'd do that, but it doesn't involve as much overengineering and playing with electronics, and I'd have to use them manually... this idea I have has the potential to work automatically once I make it smart enough.

Update on info - got my WJ ABS modulator at the JY today, it's got 5 ports but two go to the MC The rear is still only one hydraulic circuit. So it looks like I may need to look in other vehicles on my next trip, or use two modulators, each set up for the front wheels with the rear wheel line capped off. Also pulled 3 wheel pickup sensors (the 4th was destroyed.) Next I need to get tone rings that'll fit my 8.8 (or make them fit), and start hacking up the ABS modulator - the sensor wires go directly to it, so I think it's got the controller built onto the side. This might make it more difficult to repurpose.

Costs so far:
ABS modulator out of a WJ (part number 5604 1022AG-A): $40
ABS reluctor pickups (3, need one more, if I'd gotten that one it would have cost the same): $20

Expected future costs:
* tone rings ($10 for the rings, no idea how much beer to get the shafts turned to install them)
* maybe another modulator
* whatever electrical magic I need to pull to make this crazy idea work

EDIT: some specifics on the modulator wiring harness for any other EE/vehicle electronics types to pick over...
Aside from the 4 twisted-pair reluctor pickup harnesses (white = front right, yellow = rear right, I have not checked the other two but they are red and green), I have:
* four heavy gauge wires. two are BLK/ORG, one is RED/GRN, one is RED/BLU. I suspect these are power.
* nine light gauge wires. BLU/PNK, WHT/BRN, BLK, YLW/PNK, PNK/ORG, WHT/ORG, YLW/PRP, LBRN/WHT, GRY/BLK. Not sure what these are. I seem to recall BLU/PNK being related to the TCU on an XJ, not sure what it does on a WJ though.

I was thinking about how stupid it is to have one line to the rear brakes, but that only matters in 2WD. In part time 4WD, a front wheel has to spin also to let a rear wheel spin. So won't it work anyway?

You've got a point there. I hadn't thought of that...

Of course, if (say) you broke a front shaft, you'd be pretty screwed with this setup. I'm trying to make it behave as much like a "real" differential locker as possible.

EDIT: the case on the HCU (bolted to the side of the modulator) is held shut with E5 or E4 external torx screws. My set only goes down to E6. DAMMIT CHRYSLER :twak: use normal bolts like the rest of the world.

The whole "one circuit into the back axle" is a holdover from early truck/car solid axle suspension design, and a continuing problem with truck brakes.
The early brake systems only had 1 circuit to each axle. They upgraded to separate circuits to the front tires when they went to IFS. Most car manufacturers went to individual rear circuits when they switched to IRS. Most trucks never went to IRS.
Then ABS came out.
Early ABS was used in trucks to solve a brake bias problem that cars don't generally have. Brake bias is set at 60-80% front bias based on the weight distribution of the vehicle. A truck's weight distribution changes so radically from unloaded to loaded, that setting the bias right is impossible. Either it's set for the unloaded condition, and the rear brakes under-perform when loaded, or it's set for the loaded condition, causing rear wheel lockup when un-loaded.( The rear brake load sensor on the MJ was one of the mechanical solutions to the problem.) The first trucks with ABS had rear wheel only ABS. They set the bias for heavy loads, then used the ABS to keep the rear tires from locking under unloaded conditions. Even when they went to 4-wheel ABS, the rear sensors were still used to mainly control bias.

The Explorer suspension/axles/brakes were taken from the Range pickup and it has the "legacy" single tone ring on the dif.
The XJ never had a legacy pickup truck in it's past(the MJ was based on the XJ, not the other way around) so when they finally got around to putting ABS in, it got a 4-sensor system, even though it never got a dual rear brake circuit.

Hey Kasten, if you haven't seen it, check this out:
http://www.naxja.org/forum/showthread.php?t=998638
Fitted factory tone rings and sensors on a D-44. 'Give you an idea on what it might take to get them into an 8.8(if it'll even fit)

Totally un-related info: 4-wd Dodge pickups clear up into the early 80s had a single front circuit, with a single flex line from the dif up to the frame.

tbburg - yeah, seen that thread. I've actually recommended it to a few people, got some great info in it, it's one of the threads that sparked me thinking about this whole project.

kastein said:

EDIT: the case on the HCU (bolted to the side of the modulator) is held shut with E5 or E4 external torx screws. My set only goes down to E6. DAMMIT CHRYSLER :twak: use normal bolts like the rest of the world.

Click to expand...

Solved this problem with a cutting wheel in my Dremel. The HCU case is a little worse for wear...

Seems the modulator is fairly simple - there is the pump motor on one side, and then there are a number of sealed valves mounted into the aluminum valve body (shocking, right?) I'm at work and I forget how many there were, I believe there were either five or six. They look like stainless steel silos hanging out of the side and each goes into the centerbore of a small solenoid, so I think they are magnetically driven.

Popped the HCU itself open, sadly it was very well weather-sealed and I pretty much destroyed it in the process. The PCB inside is laminated to a heavy (~1/8") aluminum backing plate, I'm assuming for rigidity and fibration resistance, then "potted" into the housing with clear silicone rubber. The cover is glued on around its entire perimeter with black RTV. Almost all of the ICs on the board bear custom Chrysler or ATE (the makers of the whole unit) part numbers, I haven't really looked into what they are as I doubt I'll be able to find datasheets. I'm pretty much planning on gutting the thing, ripping out the control board, and driving the solenoids and motor myself. Might have to fully disassemble the valve body as well, since I have no real idea (other than guesswork based on valve and flare fitting location) which solenoids control which brake lines.

You know a 8.8 spool is very very cheap and easy to install............

That's entirely missing the point of this whole thread. I'm trying to get as close to having a selectable locker as I can without spending more than, say, 100 bucks. So far I've spent 60 and I expect to spend about another 40-60 on assorted hardware and electronics components before I'm done. Yes, I know it won't perform like a real locker. No, I don't think this justifies me installing a spool/lincoln locker or dropping the extra 800-1000 on a selectable or 200-500 on an auto locker.

EDIT: also, it provides me with another electronic device to tinker with. :compwork: