gemellocattivo.com

Gee, you’re nothing if not perseverant. The strange thing is that I have been following all this time and one of my girls is married now. At least it drove for a bit. Not much longer!

This has got to be the stupidest thing I've ever conceived......

More steering progress. Going from angled rack to straight I wanted the pinion shaft as far left as possible and I over-shot a tough so the bearing won't install because the clutch cable boss is in the way. Fortunately I no longer have a clutch cable so the boss can go.



And with that I was able to play connect the dots





I set the EPS as high and as far from the pedals as possible so nothing is welded yet. Tomorrow when my head is fresh I will replace the old shot rubber bushes that hold the rack mounts with aluminum and get it mounted solid. Recheck everything and tack the slip[ joints....and I should have steering again.

The steering ECU I have is very easy to hook to the car ECU and make the steering speed sensitive and adjustable....but not as adjustable as I was thinking. I can tell the EPS unit to deliver max assist from 0-100% based on speed but the slope of the strain v assist curve is fixed so if say max assist is set to 50% then from 0-50% ir will feel exactly the same as if it was set to 100% and that makes me sad. I'm working with the vendor to make a new ECU with 2 adjustments that will make the unit truly fully tunable and get the steering feeling exactly the way I want it to feel. I have a price but not a link to place the order, hopefully soon.

Can’t you link it to a normal output and put a curve in indexed by speed? Hope I am making sense here…

Yes, that is exactly how this unit is setup to work. It has a PWM input and anything below 20hz is 100% and drops to 0% at 800hz. My ECU reads front wheel speeds plus rear axle speed so no problem to send a variable freq signal that goes though a single table....x kph = y hz. and my buddy was going to make a little box with a phone app so people without a fancy ECU can do the same....like use the signal to the electric speed as the input.

What we realized doesn't change is the input strain to output tq relationship in the EPS. The speed input limits max output tq but at any strain below the output limit, the relationship of strain to output is fixed. I could probably just ask to have that slope be a function of speed so its always 0-100% but the 100% point varies with speed......but I can't help but think they set it up the way they did on purpose....I'm not sure but having that 2nd input lets me create quite a range of steering feel I think

Yes, thought about the latter - should be good enough.

I started the day by dropping in the pedals and checking foot clearances.....it SOOO much better than the old setup. It seems like a non-issue



Honestly this new unit feels like its designed for this car. Its got a smaller housing OD than any of the OEM unit and then it slips right into the space between the foot-well and dash box. I'm not sure sure how I could ask for better.


Next up was connect to the ferrari upper u-joint that uses a taper pin. I marked it with a transfer punch....my life changed when I learned about transfer punches. Then a taper reamer.



Next the rack mounts attach to the car with rubber bushes.....that gets the WRONG stamp so I make aluminum replacements. check the angle quick, setup the lather and bushings




The new rack is too small for the OEM bellows and all kinds of wrong for the bellows it came with so I made sleeves. They aren't in any load, the just hold the bellows so I shrink fit them on...1-2 thou interference, heat it up, drop it on, done



Then weld all the joints and put it back together....its king of late....I haven't made any king of torque arm for the EPS but it is mostly trapped and I really wonder how it works.....not my finest wiring but functional


and with the EPS at its default 100% output I have near 0 effort turning, noticeably lower effort than the old unit provided. And a quick check it s 1.6-1.7 turns lock to lock. Win! Tomorrow I'll figure out the torque arm then probably start working on a new cover to seal the cabin and all the linkage needs paint so still a bit of work but its getting close.

This morning I've mostly been thinking about EPSA tuning and what I should be asking the vendor to do for me. I had asked to a couple changes and they said at least the end of the month, maybe a bit longer before they have time to look at it.  I'm thinking I should be using that time to actually understand what I want.

I read thought a thread here

https://grassrootsmotorsports.com/forum/grm/what-happened-to-steering-feel/191112/page2/

And found talk on hotrod forums of a similar issue to what I had, failure to self return.  Some said adding caster fixed it other said it didn't.  My car didn't have the issue, at least not that I recall,  when the EPSA was installed but broken or maybe I just didn't notice because I was focused on how hard it was to steer. 

I have no idea what the fiesta unit did exactly but I requested and got the strain (Nn) v output(A) curves for the new controller:



This is interesting.  I commented that last night when I powered on the EPSA I was near 0 effort.  The max output torque for the unit I have is about 46Nm and I think I measured 28 or 30Nm to turn the wheel parked so nothing over 30Nm is relevant...and that seems like an important point to understand at different speeds.

Pause - I decided it was important enough to go remeasure with all the new stuff.  At about 22Nm it moves, normal turning speed is closer to 25Nm.  Then with the front wheels up its hard to measure....3Nm? then car on the ground, EPSA on I see..1-2Nm?  I should be using about 1/2 the max output and that would mean 2.25ish on the graph?  so that kind of agrees.

The friction steering in the column above the EPSA (column of rack mounted) is effectively 0...just a shaft in a couple ball bearings so that seems irrelevant and it looks like there is 0.25Nm offset baked in the controller to deal with it. Which I guess all just brings me back to friction is evil and the caster is not adequate to overcome the added friction so nothing I do to the controller sort of active steering to center is going to sort it.  I read 1 account that the prius unit when its got its CAN signals and has the steering position center correctly set does return to center but not in default mode. 

I'm kind of curious if anybody with a rack mounted EPSA has measured the wheels up force and can share?  I guess same question about any hydraulic PS setup?  the feel has got to be all about the friction I'd think?

Still not sure what changes I'm going to ask for exactly.....

Very little actual work today. I cleared out the heater parts and hoses making it impossible to to really access the steering unit then pondered for quite a bit what the enclosure will look like, how I might make it and how I might install it. I'm pondering a trash bag and spray form like they done for packing stuff....maybe. I also realize the foam seals inside the air duct valve are long past functional, one of the heater valves appears to be leaking and I don't have the blower fans anyway so none of it works but that is a project for another day.

I've kind of settled on the torque arm to the inside I think.

I remembered I bought a steering position sensor....but I'm not sure how it works, maybe a bad investment. I need to look at it.

Last I ordered a cheap signal generator so i can send a speed signal to the steering unit a bit without needing to connect to the ECU just yet as I don't know the final setup and there are no output channels left just waiting for a purpose so $14 to amazon is just easier.

A few years back, a late night encounter with a car pulling out in front of me killed my E46 330ci. Since I had a 120 mile round trip work commute, and was now using my FZJ80 Landcruiser with 35" tires, I needed a new "fun" car for the commute to work. Finding no M3's or similar in my price range, I started looking at used GTI's. The first one belonged to a kid who's dad showed it to me after a 3plus hour drive [he was the mechanic he informed me and had done the all important water pump/tensioner/timing belt service]. It was lowered as a lot of the mk4 community does, but as I am now 63 I didn't really fancy anything that was gonna jar me constantly. On the test drive this GTI exhibited the characteristics you describe ie: no return to center, very disconcerting and definitely not desirable for a 120 mile a day commute, so I passed on that car. I did end up with a much nicer GTI which I have modded a little, but I like to think tastefully, including 1/2" drop springs, and poly bushes as well, and do not experience that no return to center with these mods.

I've been getting about a dozen fake registrations a day (mostly russian) I had to delete so I changed the registration challenge question and back to 0...nice.



Yeah....I suspect there maybe some alignment magic that could be done....I posted the question to the youtube guy "suspensions explained" hoping he'll solve my problem

The steering ECU uses a pwm signal to read speed...its designed to hook to an electric speedo I guess. Rather than mess with hooking it to my ECU for basic testing I bought a $13 signal generator with a knob that lets me easily adjust the steering assist and it seems to do exactly what its supposed to do...the steering goes smoothly from weirdly easy to you really need both hands/o-assist. I think its good and I need to get it wired to the main ECU insert a steering tuning table.



The bad news is with me really paying attention I noticed some odd resistance when the steering wheel is about about +/-90deg. It seems the lower ujoint that really shouldn't go past 30deg is at about 45deg.....I will need to replace that with a double ujoint or drop the steering column lower about 3" so a double ujoint seems the answer....that means I need another mount or support but I should be able to handle that with EPS mounts