Starting on my EPAS project (Toyota junkyard swap)

[Map63Vette]

So long ago when I first bought my car it had the factory Dodge power steering. Fast forward a bit and when I did my 5.7 Hemi swap, I changed over to a standard 24:1 manual steering box as there weren't really power options at the time, and I didn't really like the finger light Dodge style power steering anyway (Borgeson was still working on their PS box back when I did my engine swap, and it was expensive). Fast forward many more years and when I got my new job I started updating my Dart more as it was becoming my go-to daily driver. I was getting tired of having to sling the wheel around multiple times in a parking lot, so I swapped over to a 16:1 manual box with the thought that I might do an EPAS setup at some point. The 16:1 box isn't too bad to drive day to day, but it can get heavy at times and one handed driving around low speed corners can get interesting (car is a stick shift as well, so one handed driving is more common). This past Christmas I got a pretty complete EPAS column setup from a Toyota, so I bought a spare Dodge column to cut up to see how I could merge the two into one.

So far I've only really taken both columns fully apart to see what I have to work with and to do some measurements to figure out how I want to graft it all together. It actually looks like it might not be all that bad to get the two stuck together, though the input shaft is going to be the main one I'm still thinking about. I'll try to get some pictures this weekend to show all the parts, but I did measure them all up the other day and try to get them into CAD so I can do some layout work and get a better idea of where I might want to cut stuff down. I went ahead and 3D scanned the motor and column mounting bracket as well just for the fun of it and it actually lays over my crude model pretty good.

The general plan at the moment is to press the inner "shift" tube (it's just a dummy tube on my column since I have a console shift setup) onto the bore of the EPAS motor as they are pretty close in diameter. I'll also need to make some kind of spacer to go between the inner shift tube and the outer column housing to make them more solid. I'll have to cut down the Toyota and Dodge input shafts and weld/join them together in some way. I can go a couple of ways on the output side depending on whether I want to keep the collapsible part of the column or not. The "cheap" plan would be to just cut down the Toyota output shaft I have and weld it to a portion of the Dodge output shaft. That would be simple and straightforward, but would mean no collapsing anymore. If I want to maintain some amount of collapsibility it looks like I can buy U-joints and a telescoping shaft from Borgeson to join the two together. Drives the cost up quite a bit, but would potentially make for a safer setup in the event of an accident.

On the electronic side, my plan is to send the actual CAN messages the controller wants to see to the EPAS module so that I get variable assist depending on speed. I hooked things up on my test bench and was reading the CAN messages off the module to see what it thought was going on, but I don't have a ton of info on those at the moment. I can see the output of the torque sensor and some status messages, but I don't know what the actual status messages mean just yet. Once I get it in the car I'll see what kind of assist I get out of it and if the factory scaling is palatable or if I want to tweak it.

So plenty of work ahead and likely something I'll take my time with as I'd like to make sure it's done well, but figured some of you might find it interesting.

 

[Map63Vette]

So here are some pictures of the columns in pieces to give an idea of how they go together.

First up is the Dodge column:

I didn't break down all the parts at the end by the steering wheel, but you can see the main upper bearing housing and then a tapered shroud under it. On a column shift car that part rotates and is connected to the shift tubes inside the column. On this column there is a single filler tube that supports the shroud, but otherwise doesn't really do anything. It does have a collapsible section in it though. The shaft goes in the middle of all of it and also has a collapsible section. The column bolts to the mounting bracket with 4 bolts, and the mounting bracket bolts to the pedal support with 3 bolts.

Next up is the Toyota column:

It's harder to tell from the pictures, but it's mostly just an upper column tube and shaft and then a lower driveshaft style setup. This column is tilt/tele though, so its' got a little more going on. The tube with the white sleeve was originally pressed on the shiny portion of the motor assembly at the base of the input shaft. The upper column slid over it and the tilt and telescoping was controlled by the little black lever. Similarly, the upper portion of the input shaft rode in the upper column and was retained with a snap ring. The lower portion is just a typical driveshaft style setup with U joints at both ends, though this one also came with a solid shaft extender piece. Unfortunately, the U joints are just barely different sizes between the Toyota and Dodge. I believe the Dodge is 3/4 and the Toyota I think is 11/16. The column attached to the dash similar to the Dodge with two bolts on the upper column housing portion and one bolt through a bracket on the motor.

The plan going forward looks something like this:

I plan to press the inner shifter tube onto the motor housing like the small plastic wrapped tube that was there originally from Toyota. Then I need to make a filler piece to press between the inner tube and the outer housing. The steering shaft itself I'm still on the fence on. I was going to try to just cut the very end of both the Toyota and Dodge shafts off and weld the Dodge end to the Toyota end to keep the slip joint. I don't really need the slip joint though, since the upper column can't collapse, so I may just cut the Toyota input shaft at the base of the splines and attach the Dodge shaft there. Need to do some measurements to see what makes the most sense. On the output side, I still haven't done enough measurements to know what I want yet. The factory Toyota slip shaft might actually be long enough with the extension I have if I can find a yoke to swap out on the steering box end. I also need to figure out how I want to attach the lower portion of the housing to the motor. My current plan is to use the mounting holes on the motor and make a bracket that would bolt where the third bolt on the factory Dodge column mounting bracket goes. Strictly speaking, I wouldn't actually need a lower column housing, though I think it's easier to keep it to fill the hole in the firewall.

Here is a section view of the parts in my CAD software to give a better idea of the assembly I have planned:

And here's my plan for mounting:

Should hopefully make some more sense as I start cutting stuff up. I'll try to get more pictures along the way.

 

[Clelan]

Looking forward to seeing how this turns out! Good luck!

Cley

 

[72bluNblu]

Definitely going to be watching this one, I have 16:1 manual steering in my Duster and while I don't have a problem with it doing something like a variable boost EPAS has definitely crossed my mind more than a few times!

 

[Map63Vette]

Definitely going to be watching this one, I have 16:1 manual steering in my Duster and while I don't have a problem with it doing something like a variable boost EPAS has definitely crossed my mind more than a few times!

The variable part was the big one for me. I like a little weight to my steering when I'm cruising along, but would appreciate a little more help at low speeds and for things like turns from a dead stop. I'll be curious to see what the factory assist "curve" looks like. Might be interesting trying to scale it in a consistent way that doesn't just jump around. I don't know if the factory assist is linear or by some kind of power, though I'm not sure if I'll be able to determine that even after the install or not. I should have torque sensor values available to me on the CAN which I could log along with me speed to build up a curve, though it might be more complex than that. Either way I've got to get it in the car first, so I've got a ways to go before I have to worry about it.

 

[Map63Vette]

I started taking some measurements on the upper column to make my first cuts to start sizing things up. The inner tube I'm still going to have to flare out on the end to press onto the motor, but it's only maybe 1 mm off, so I think I can expand it. I found a socket that's the right diameter, so my plan was to get the end as hot as I could with a torch and try to hammer it over the socket to spread it a little bit. I could potentially just heat it and press it on that way, but I think that would be too tight a fit and it might crack. The outer column was simple enough and I just cut it at the start of the mesh part. I still need to work out a spacer to join the inner tube with the column housing, but that should be pretty straightforward. I might 3D print something just for a test fit. Here's a loose test fit just setting the column over the motor. So far it's looking pretty good.

Where things got even better was on the input shaft. I was scratching my head trying to figure out how I wanted to join the two as I was a bit worried that any sleeve might not fit inside the inner tube. As it turns out, the upper Toyota input shaft that has the female splines is entirely hollow. Even better, the ID of it is a near perfect fit for the OD of the Dodge steering shaft, aside from it being fully round and the Dodge being a double D.

So at this point I think I'm just going to drill through the full assembly and put a roll pin through it, then probably do a bead of weld at the joint as well. So the weld should take most of it with the roll pin as a backup just in case. I've got ~1.25" of insertion between the two shafts, so I could even just do two roll pins and that might even be enough to hold it without any weld. Suffice to say I'm rather pleased with the way it's turning out so far.

 

[Killer6]

The variable part was the big one for me. I like a little weight to my steering when I'm cruising along, but would appreciate a little more help at low speeds and for things like turns from a dead stop. I'll be curious to see what the factory assist "curve" looks like. Might be interesting trying to scale it in a consistent way that doesn't just jump around. I don't know if the factory assist is linear or by some kind of power, though I'm not sure if I'll be able to determine that even after the install or not. I should have torque sensor values available to me on the CAN which I could log along with me speed to build up a curve, though it might be more complex than that. Either way I've got to get it in the car first, so I've got a ways to go before I have to worry about it.

I know it's a different brand, but the upper option Hyundai's had a steering wheel button which allowed the amount of assist to be selected, (3) levels were available. I believe called out as 'Sport', 'Standard', & 'Comfort'. Might investigate if Toyota had a similar feature..

 

[Map63Vette]

I know it's a different brand, but the upper option Hyundai's had a steering wheel button which allowed the amount of assist to be selected, (3) levels were available. I believe called out as 'Sport', 'Standard', & 'Comfort'. Might investigate if Toyota had a similar feature..

Yeah, I think I know what you're talking about. My dad had a C7 Vette for a while that was similar and I believe had steering settings that affected the weight and feel. I haven't seen anyone mention it on the Toyota stuff, at least as far as CAN messages go, but I'll have to look into it. I could basically build that same idea into mine though, which is an interesting thought now. I could have a multi position switch that could chance scalings on the fly.

I need to do a little more checking on my module as well. By most accounts, people say the internal relay usually takes a few seconds to come on if it's not seeing CAN traffic, but mine came on right away. I always had it hooked to CAN, but I wasn't sending it any messages originally, so I need to pull the part number and make sure I don't have a non-ABS unit that wants a discrete VSS signal instead of a CAN one.

 

[Red383]

Not an A-body, but I installed EPS in my Old Dodge:
Electric Power Steering install - Sweptline.ORG

 

[DionR]

Not an A-body, but I installed EPS in my Old Dodge:
Electric Power Steering install - Sweptline.ORG

How does it drive? Any issues with returning to center?

 

[Map63Vette]

Not an A-body, but I installed EPS in my Old Dodge:
Electric Power Steering install - Sweptline.ORG

That's looking really similar to what I had in mind and followed a lot of the same thought process I have at the moment. Nice to see that it worked out well. I like your idea for dimpling the motor housing for the set screws too. I was trying to think of a way to reinforce that joint a little better and was having a harder time since it's a steel to aluminum thing.

I did finally manage to get some work done on mine again this past weekend. I found a socket that was the same size as the motor hub to press into the end of my inner tube to try to expand it out to make it slip over the motor. I originally tried heating the tube up with some of the propane bottle torches and hammering it over the socket, but I got nowhere with that. I don't think I could get enough heat in it to be useful, so I ended up just taking it over to my parents' house to use my dad's little Harbor Freight hydraulic press. He had a couple of other sockets that were just a bit smaller so we could step up the size as well instead of doing it all in one go. Didn't come out quite as nicely as I had hoped since it ended up flaring off-center, but I think I can still make it work okay.

Truthfully speaking, I really don't even need the inner tube and am debating whether I keep it or not. I was originally thinking I needed it to help keep the shift shroud in place, but the shroud has reduced diameters on either end to pilot into a ring that's located in the outer housing and a bore that's in the upper bearing housing, so it really isn't bringing much to the table. I would still need a filler spacer thing between the outer housing and the motor bore one way or the other, and it would probably be easier to make one if I'm not trying to make it fit over the inner tube now that it's slightly off center.

I was going to 3D print a test filler piece to go between the outer housing and the motor to get an idea of what I'm looking at. It would be really easy to have someone make it since it would just be a tube with some pretty controlled dimensions for the press fit. Would love to have a lathe to do it myself, but I don't have the room. I do have a friend that has one though, so I'll have to do some looking.

My biggest thing at the moment is trying to decide what I want to do with the lower half of the setup. The "easy" route would be to get the Borgeson coupler and just cut my stock shaft and be done with it, but I like the idea of trying to keep some collapsibility if I can. I think the Toyota pieces I have will span the distance between the motor and the steering box, but I need to swap out the yoke on one end for that. I did some searching the other day, but finding steering sized yokes is more difficult since most people deal in driveline sizes. The factory u-joints are also staked in, so I'd have to do a little delicate surgery to remove the staking so they could be disassembled. There is a driveline shop just down the street from my job though, and I was in there talking about steering yokes for a project for work not that long ago, so I figured I might take it to them and see what they could do.

I'll try to grab some more pictures the next chance I get. Once I get the input shaft drilled and pinned the upper part of the setup is really pretty much done. I still need to make a bracket for the rear of the motor to pick up the third mounting point, but that's going to be another day of taking steering columns in and out of my car so I can hold it up in place and get an idea of how I want to clock it. I kind of wish my car wasn't my daily so I could leave it disassembled, but it really only takes 5-10 minutes to take the column out at this point, so it's not that bad.

 

[Red383]

How does it drive? Any issues with returning to center?

It drives fine, no issues with return to center. I am just running it in the default mode (no Controller), I have thought about buying one of the controllers fro ebay to see if I can up the power while sitting still, but really, it is not that bad to even worry about.

 

[Map63Vette]

So finally got around to doing a little more work on this over the weekend. The more I thought about it, the more I preferred the idea of using the original Toyota inner tube setup as it's a fair deal thicker than the inner tube of the Dodge column, and I know it's good and centered to the steering input shaft. If the Dodge tube would have expanded more centered I do think that would still be a viable option, but this felt like the better way to go given the circumstances.

I 3D printed a couple of rings to center the column housing over the motor and inner tube setup. The lower one is just a plain ring, nothing special there:

The upper ring I went ahead and put a tab in to line up with a notch that was in the tube from the factory. I probably don't need it that bad, but depending on how tight of a fit I want, this would help keep it in place when the outer column is slid over it:

I had to print the upper ring several times to get the fitment right. The last print I did was still probably a little looser than I'd like, so I may yet print another to try to snug it up. I'm debating whether I use this as my final part as opposed to trying to make it out of metal. Plastic is actually probably the right choice, but maybe something like a tough UHMW or Delrin or something like that. It would be easy to machine (so I could keep the tab), without being too expensive. The column housing isn't perfectly round or smooth either, so that would give it a little give when installing. The main goal of the upper spacer is just to keep the housing aligned more than anything. It doesn't have to be a super tight fit.

Here's the setup with both installed and the housing ready to slide over:

I also went ahead and cut down the column mounting bracket to fit around the motor. I just kind of eyeballed this one and cut it right behind where the bracket flares out. As it turns out, this was just about perfect and left a little room for the wires from the torque sensor while still offering some protection from a wayward foot:

At the moment this is all still a pretty loose fit. I think long term I might make the lower spacer out of metal and bolt the column housing to it. I liked the idea of drilling through the spacer into the motor housing just a bit so that some set screws could get a bit of a bite into it to help offer some torque reaction and keep the two attached in a more secure way. However, before I do that I need to get this setup in the car and figure out the motor clocking. I figure I'll set this all in the car, clock it where I want it, then mark the column hosing and motor with a sharpie to get an alignment. Then I can drill holes/dimples to keep it all there and make a rear mount for the motor to pick up the third mounting point I cut off of the original column bracket.

While I'm there, I will also see about taking some measurements for the lower column housing and steering shaft. With any luck I can get everything I need with just one test fit, then work on getting parts made/modified in the meantime. I still need to figure out the control module mounting as well, but I figured I need to get the motor in the car first for space claim and to see where the harness will reach. I know I can extend if it absolutely necessary, but I'd rather not if I can get away from it. There's really not a lot of length to work with though, so we'll see.

 

[gzig5]

Really neat project and something that I was wondering about doing on my car. I'm definitely doing a Borgenson upgrade or something because the stock power assist is too much and the stock gearbox needs help. I'm concerned about the electronic aspect and if it will work in a stand alone mode. Interested to see what you find.
If it comes to it, I can probably help you with making stronger bushings if you don't have a lathe.

 

[Map63Vette]

Really neat project and something that I was wondering about doing on my car. I'm definitely doing a Borgenson upgrade or something because the stock power assist is too much and the stock gearbox needs help. I'm concerned about the electronic aspect and if it will work in a stand alone mode. Interested to see what you find.
If it comes to it, I can probably help you with making stronger bushings if you don't have a lathe.

Thanks! I keep going back and forth on the bushing setup trying to figure out where I think the loads are really going to be on the system and whether my thinking is overkill or not. I keep referencing the original Toyota setup and trying to understand how they did it to reassure myself. For the most part, it mounts in a very similar way where the upper column bolts to the dash with 2 bolts, then the motor bolts to the dash with a separate single bolt. The inner and outer columns really aren't attached to each other at all, they are just held in alignment by the slip fit of the inner and outer tubes, and it can't be a super tight fit because that's where the telescoping of the Toyota column is.

I do think I'm going to want to make some kind of lower column housing setup though as it would help support the motor as well as the mounting bracket I'm going to have to come up with. I'll get a better idea once it's in the car, but it has been coming together better than I thought it would as a lot of stuff just kind of "fit" or was close enough to not require a ton of rework. Looking forward to actually getting it in the car one of these days.

 

[Map63Vette]

Have some good weather this weekend, so trying to get as much done on this as I can since weekends are about the only time I have enough long blocks of free time to dig into things vs a spare hour here and there.

Got the half column mocked up in the car using the front mounting studs to see where the rear of the motor is going to sit relative to the rear mounting hole on the original bracket:

It's looking pretty promising so far. I cut the rear mounting tab off what was left of the bracket I already cut up and my plan is to make a flat plate that will bolt to the steering motor that I'll weld to the remainder of the factory bracket. I do wish the third mounting point on the pedal box was a little more robust, but I think this should be okay. It will hopefully have even more support if I can get the lower half of the housing attached as well.

Here's an idea of what I'm talking about with the plate:

I'm 3D printing a copy as I type this to do a test fit and hopefully use as a template to cut up some steel.

Generally speaking I'm pretty happy with the fit so far. It looks like it hangs lower than it does in some of these pictures. It would be nice if I could rotate it even more so the motor points more upward, but there's just too much going on under the dash.

I think I'll probably try to mount the control module just above the motor, possibly using the same mounting bolt as the rear column mount.

I measured up the steering shaft and the lower housing distances as well to get an idea of what I'm looking at there. Unfortunately, I'm not going to be able to use the original Toyota lower shaft setup. It's just not long enough. It's ~21" from the face of the output shaft on the motor to the face of the input shaft on the steering box. The Toyota shaft I have is only ~9-10" long fully compressed and maybe 12" fully extended. I have a ~6" extension shaft that also came in the bundle, but that still leaves me 3" short. Extra unfortunately as well, what I have left of the stock lower shaft is also too short to reach all the way.

I've still got a couple of options here, but not sure which way I want to go just yet. I'm debating whether I just buy some tubing and cut the end off of the steering shaft and weld it onto the tubing to lengthen it or whether I want to see about making a u-joint style shaft. I'm likely either going to buy a coupler adapter from Borgeson to adapt straight 1" shaft to the Toyota output spline and retain the original Dodge "u-joint" on the steering box input, or I could just buy telescoping shaft with matching u-joints from Borgeson to make a whole new shaft with modern needle style u-joints on either end. The latter is likely going to be pricier, but I'm not sure if there's any other real benefit to it. It would make the support bearing at the end of the lower housing more difficult to deal with as well, though I could pull the telescoping shaft apart to deal with that. Strictly speaking I don't need a bearing there, but it would be nice to have something to keep the air from the engine bay from blowing in there.

 

[Map63Vette]

Got a little more done yesterday to try to button up as much of the top end of the setup as I can. Welded up the bracket for the motor and drilled and pinned the two upper steering shafts together. I was struggling a bit trying to figure out how to position the rear mounting tab and ended up going with a pretty low tech solution after all the 3D printing I was doing to try to mock things up. I took the stock bracket off the column currently in my car and just screwed together some 2x4s to get the spacing and offset right. Then I screwed the cut down pieces to the same holes to line everything up so I could tack things together.

Not very pretty welds, but pretty sure they should hold together fine:

I did print some little filler pieces for the upper shafts to fill the gap between the flats on the Dodge "double-d" style shaft and the plain round Toyota one. Just helped to keep things tighter and better aligned. There isn't much, but there is a gap between the Dodge and Toyota stuff, so it's possible this might be just barely out of alignment. I haven't tried to spin it or anything, but the upper bearing is in a rubber isolator, so I should have a little compliance there anyway if things aren't perfectly aligned.

So at this point I could attach the whole thing into the car and I'd just need a lower shaft to be able to drive with the new column setup. I still want to do something about the lower column housing though to help give the motor some more support as it is fairly heavy. I also got to thinking about it this morning and I think a brand new telescoping shaft with u-joints likely makes the most sense as I realized that would be the only way to keep the collapsible feature of the column. I went through the trouble of keeping the breakaway mounting for the rest of the setup, so I might as well see it the rest of the way through.

 

[Map63Vette]

Went ahead and ordered parts from Borgeson to make a shaft. Got a 3/4-36 to 3/4 DD u-joint for the steering box and an 11/16-36 to 1 DD for the motor output side of things as well as a matching 3/4 DD to 1 DD telescoping shaft to join them. Should make for a pretty much bolt-in setup between the two. Still need to figure out how to mount the lower column housing to the motor, but hopefully I'll get a little time this week to mock things up. I took some measurements when I had the setup in the car over the weekend, so with any luck I can get close before then and possibly have electric steering this weekend.

 

[Map63Vette]

One step closer this past weekend, but it rained most of the day Saturday and I had dinner plans Sunday, so I didn't get a chance to get anything in the car. I did get the lower column housing cut down and a plate welded to the end of it. I also went ahead and put some short beads on steering shaft I pinned together for some piece of mind. I think the roll pins should be fine, but figured it couldn't hurt to have a little extra.

I've got the lower steering shaft and u-joints on hand as well, so I just need to cut it to size and I should be able to install everything into the car. I could potentially cut it now, but I wanted to bolt everything up and recheck my measurements to be sure. I know I'll have some give and take since the shaft telescopes, but I'd like to err on the side of having more compression in the shaft than extension.

I still need to do some electrical and coding work to get the whole system up and fully running, but I might tackle it in stages. Supposed to be a pretty cold weekend coming up instead of the nice 60 degree weather we've been having here lately though, so less fun working in the driveway. My hope is to get everything installed in the car and at least have it running in failsafe mode. Should be easy enough to work on programming after that if it's all hooked up at least.

 

[Map63Vette]

Success! Got everything in the car this weekend and took it for its maiden voyage. It's certainly interesting, but really quite impressive. It's basically silent and quite light in the failsafe mode. The steering feels maybe a little bit "gummy" for lack of a better word. It's got a little resistance everywhere, but it's consistent no matter where you are in the steering wheel travel. Return to center is a bit on the slow side, but I wonder if some of that might have to do with it being in failsafe mode. I have HemiDenny upper control arms on the car now using the center position spacers, so I have a moderate amount of caster. Will probably swap over to the full caster setup to see if that makes any difference. It feels very stable driving around though, so I'm very pleased with it.

I still plan to dig into the coding side of things to try to get it working as it would have in the Toyota. I did set up the Megasquirt to read the torque sensor CAN messages and can see what the unit is doing, so that's pretty fun. It has status messages as well I'm trying to get it to read, but I think I need to keep playing with things. The MS is a bit weird with how it handles CAN variables and sensors, but worst case I think I can just turn off the torque messages and swap over to the status ones. I really don't need to know what the torque stuff is doing anyway.

Tried to grab a couple of pics along the way. Here's a setup of the full column assembly ready to go into the car:

The upper column housing isn't attached to the motor at all, it just slides on. I'm still considering whether to replace one of my plastic spacers with an aluminum one that I drill and tap for some short screws to hold them together, but they don't really need to be attached. It just makes installation a little easier when you aren't fighting with the motor spinning around.

Here's a couple of pictures of it fully installed in the car:

It looks like it hangs down moderately far, but sitting in the driver's seat, you really don't see too much of it. It's also close enough to the bottom of the dash that it's nowhere near the throttle pedal for someone with big feet. I reused the lower portion of the column housing to keep some collapsibility in the system and to help reinforce the motor mounting. I had to swap the set screws and jam nuts that came with the Borgeson u-joint to shorter ones and Loctite on the shaft side of the joint to have clearance inside the housing tube, but I think it should be fine.

I didn't grab any pictures, but I did have to cut back the engine bay side of the housing as well. The Hemi valve cover sticks into the same spot as the column by a little bit. The old column shift column I had had a cutout where the shift levers went that perfectly lined up with the valve cover before. The console shift column is a full round tube all the way to the end, so I just cut it back closer to the firewall to clear the engine like some others have done here.

 

[Killer6]

That would need to be -0- calibrated, 1st the alignment has to be 'on' & wheel straight once that is completed, then cal -0-'d. It will 'remember' where there wheel is supposed to return to, hopefully that will improve the feel.

 

[Map63Vette]

That would need to be -0- calibrated, 1st the alignment has to be 'on' & wheel straight once that is completed, then cal -0-'d. It will 'remember' where there wheel is supposed to return to, hopefully that will improve the feel.

I've read that before, but I've never confirmed exactly how to do that or whether it applies to all the different units out there. Some said the Toyota units don't care like the Saturn units do. I actually have a Toyota VCI and software so I could in theory connect to it and try to read stuff from it, but I haven't actually tried to do that.

One place said to just ground one of the CAN lines for something like 7 seconds.

I've been trying to find a Toyota service manual to see what the factory recommended procedure is for replacing an EPAS unit/controller in the hopes it would have info like that. I know that the cars have steering angle sensors as well (not sure where they mount them), but I've never seen anyone mention whether the EPAS units care about that or not (aside from just not throwing faults if they don't see the message).

 

[Map63Vette]

I got to thinking about this some more on the way home yesterday and I think the first thing I should probably do is just disconnect the power to the system to see if that changes anything. The steering unit itself offers some drag to the system as it's not something that you can turn the shaft by hand on. I'm wondering if maybe it's just that extra drag that I'm noticing.

The more I thought about the "zeroing" idea, the more I'm also not entirely sure it makes sense. From what I can find at least, there is no angle sensor in the motor itself, so it wouldn't have any way to track any kind of zero anyway. I do think it seems a little odd that it doesn't take some kind of angle input for the controller though, as that feels like it would make a lot of sense to cut assist once you get to the stops or stuff like that. Maybe they don't because they use the same unit in multiple cars though, and each car might have a different trave limit, so each controller would have to be programmed for each car model.

 

[Map63Vette]

It's fully functional! I had some good time this past Saturday with really nice weather to sit out in the car and mess with the CAN communication and messages to try to get the Toyota stuff emulated correctly. Had a couple of struggles getting the right libraries to work with the Arduino I have (it's the oddball Nano Every, which doesn't quite line up 1:1 with the regular Nano), but finally managed to get it broadcasting. I set up my Megasquirt to read the messages as well based on the Toyota dbc files people have built up online to confirm what was coming out looked correct. The stars finally aligned when I noticed that I didn't actually have power assist while sitting there doing testing. Normally in failsafe mode the box provides assist even without the engine running, but it's designed to only start assisting when there is an RPM signal. So it's expecting to see a message from the engine with that value. Low and behold, when I got it all set up and running and it was sitting there broadcasting 0 rpm, the assist stayed off. Fired it up and took it for a drive and noticed the steering was even lighter in the driveway and from a dead stop. Harder to tell how it lets up at speed since I don't have a lot of high speed corners to test with, but the CAN values I was broadcasting with the Arduino were lining up with what the MS was showing as well.

Also did a little testing with the return to center this morning on my way to work and it is noticeably better, so I think the problem is really the failsafe mode. I'm thinking about setting up a dial to be able to different levels of assist as well. The plan there is just to scale the speed value by some number based on the dial setting to make the steering think the car is going faster than it is. That should take away some assist, but I don't know what the limit for that is. I would assume it has minimum assist by typical highway speeds. My understanding from the CAN message is that the speed is limited to 115 mph by the variable and scaling. Not sure what would happen if you try to send it more than that.

All told I'm quite happy with the setup though. It's maybe still a touch lighter than I would like, but it's miles better than all the cranking I had to do on the wheel in parking lots before. It still kind of just feels like magic how nicely it assists without making any kind of noise or notice that it's doing anything. It feels very natural in that regard. I think it actually makes the car feel more stable as well. It seems to soak up and auto adjust those little back and forth movements I used to have to do on the highway to keep the car centered, especially on windy days. It even feels better around corners, though I haven't really tried any autocross style driving with it.

 

[bjkadron]

Very nice job! I'm considering doing this. If I were doing this without EFI, would you think just interpreting a tach signal with an Arduino would work for the rpm signal? Also any plans on how to adjust the assist level? I don't want mine too light.