Turning Radius: Another installment from the WA MSM laboratory

['95MSM]

Mr Hyde has installed a de-powered NB rack in his MSM (and posted a very nice write-up to update the NA specific instructions on the Flyin Miata website).  Now that he has an MSM rack that isn't attached to an actual MSM, my curiosity about the wider turning radius and what Mazda did to the MSM rack has been piqued once again.  Let me start by giving all appropriate props to Hyde for allowing me to partially dis-assemble his power steering rack.  There is a fair amount of stuff to mess up on one of these if you really tear into it, and he has given me the green light, trusting that I'll put it back together properly. Kudos also to SMR for the original research into MSM specific parts in the steering rack, and for providing the exploded drawing and parts list so I could dig a little further.

It turns out that there are two differences in the MSM rack: the actual steering rack shaft and the threaded collar (Mazda calls it a "holder") that threads into the passenger end of the housing and holds the rack shaft.  I first thought the MSM shaft was going to be the same as the NB, and that the collar was going to be longer, limiting travel by sticking further into the housing.  If that was the case, getting our turning radius back would be a very easy fix, but when I removed the collar it appeared to be nearly identical to the normal NB collar, particularly the length of the collar.  

With my first hypothesis shot down in flames, I decided to re-assemble the rack and see just how much different the throw is between the MSM and NB racks.  Hyde brought his car over and I verified that the normal NB steering rack has almost half an inch more throw than the MSM rack.  We know from the parts list that only the rack shaft and collar/holder are different between the MSM and normal NB.  I observed that the MSM rack pulls the tie-rod knuckles in until they touch the housing, so it seemed likely that Mazda put a narrower rack shaft in the MSM.  Hyde provided fairly conclusive anecdotal proof: when he removed the MSM rack, he carefully counted how many turns the MSM tie-rods went into the tie-rod ends, so that he could re-assemble the NB tie-rods the same depth and get his toe-in fairly close.  When he finished installing the NB rack to the same thread count, he had massive toe out.  The steering arms weren’t changed, the tie-rods and steering rack housing are the same, so the difference is in the width of the steering rack shaft.  The inner tie-rod knuckles are about ½” closer together on the MSM rack, a change Mazda made to increase the turning radius so the 17” wheels and tires wouldn’t rub.  

That makes the MSM rack arguably the worst steering rack ever installed on a Miata.  Why? There is only one correct tie-rod length for a specific set of control arms, tie-rod height & angle and steering arms.  I’m trusting that Mazda got it right for the NB inasmuch as they made a semi-big deal about the improvement in bump-steer when it was released in ’99.  By my measurements, the MSM working tie-rod length is about ¼” longer than the normal NB. This might not be enough to notice on street cars, but I would not be at all surprised to hear that MSM track cars just don’t feel as precise as NB track cars.
 
There may be hope.   If I had to make a limited number of MSM steering racks, I would look hard at starting with a normal shaft, then machining it narrower.  If that is the case, our shafts will have all of the necessary gear teeth machined into them to give us normal throw and turning radius if we extend the length.  I can confirm this by removing the tie-rod knuckles and checking to see if the end of the shaft can be drawn into the housing.  If it passes that test, I’ll evaluate the knuckle threads to see if a ~¼” shim still leaves sufficient thread engagement and if the seals are far enough inboard that the shims would not be drawn into them at full lock.  I’m relatively confident about the seals and I think the knuckle threads are long enough, so it may come down to how Mazda made the MSM steering rack shaft.  If they found that they could save $.25 on each shaft by machining fewer gear teeth into the MSM shaft, our only solution will be to retrofit normal NB racks.

I’m still at a loss as to the difference between the MSM and non-MSM collars.  The part is actually an assembly, consisting of the diecast threaded collar, an o-ring that seals to the rack housing, a seal that seals to the rack shaft and something that looks like a sleeve in the parts drawing.  I didn’t dis-assemble the collar to the point where I saw the sleeve or understood its purpose.

I welcome any comments, questions, etc, and especially new ideas about why the collar is different turbo vs non-turbo.

In the meantime, if you track your car and have had thoughts about the steering getting a little odd under heavy braking or when hitting a bump, we might now have a very sound idea why this is happening.  And if you see a couple of classified ads on miata.net for “high performance MSM steering racks – buy the rack Mazda designed for their best NA/NB series Miata”, that would be Hyde and myself.

[mr_hyde]

That makes the MSM rack arguably the worst steering rack ever installed on a Miata.

Umm... I thought we were going to keep that a secret until we sold our MSM racks?   

'95MSM has made a solid discovery here.  In short, the MSM rack is not optimal for bumpsteer compared to a regular NB rack.  It's an easy fix though - just pick up a used NB rack and install it.  Sell your 'high performance mazdaspeed special' rack to some poor guy at club roadster (sorry Karma  ) to recoup the costs.

That said, I am happy to donate my rack to Mark's quest to 'fix' the MSM rack in lieu of replacing it.  Knowing what we know now, I won't be installing it status quo in any of my cars plus I'm quite happy with the depowered NB rack I'm running now.  Stay tuned!
-h

[VagaXt]

I've read the whole OP, but I'm still somewhat lost.  Is this whole deal about the decreased turning radius and simply you (95MSM) suggesting to switch to a normal NB rack to fix this?

IMO, I like my MSM rack as-is because of the faster ratio (supposedly) the MSM rack has over the normal one.

[SilverMiataRacer]

I have heard about the lesser turn radius but I've never heard anything about a different ratio.

[VagaXt]

I have heard about the lesser turn radius but I've never heard anything about a different ratio.

Google searched it:  http://www.supercars.net/cars/2729.html

From 5th paragraph, near the end of it.

Mazdaspeed MX-5 Miata also receives a quicker steering ratio since the rack stroke was reduced more than 7mm from Miata's.

Also here: http://www.autotrader.com/research/article/16267/2004-mazda-mazdaspeed-mx-5-miata.jsp

It's all here, plus a quicker-ratio 2.3 turns lock-to-lock steering rack, in place of the standard 2.7-turn rack, just in case the steering seemed a tad sluggish before. Bet you won't find a faster steering rack in many aftermarket catalogs.

[SilverMiataRacer]

Never saw it but it doesn't sound like the ratio is faster it's just that it takes less turns.  Also the first article states we have a BB turbo so how reliable can that info really be.   

[newold_m]

Yep, compared to a regular NB rack there is less turns lock-to-lock because the turning is wider, not because it takes less turns for a given angle. 

Cool discovery and nice write-up '95MSM!

[VagaXt]

Well, another reason why I like the MSM rack is that it automatically prevents my 15x9 6ULs' inner lip from scraping my front lower control arms at full lock.  For normal NA and NB racks, you need to put a 3mm spacer to prevent this, or simply don't be full-on drifting at full-opposite lock all the time. 

['95MSM]

VagaXT: Yes, I undertook this because I would like to have my old turning radius back.  I thought there was only one part involved (the collar/holder) and if my initial guess as to how Mazda limited the steering radius had been correct, I have the equipment necessary to modify it.  For me, it would have been a labor-only upgrade (always attractive to the unemployed).  I also thought there might be an opportunity to modify collars on an exchange basis.  It would be fun to have my fix on a whole bunch of MSMs, but if there is no easy way to mod the rack, the solution would be to retrofit an NB rack. 

When I saw how Mazda limited the rack and compromised bumpsteer, that spoke to the racer in me.  My Sports 2000 is bumpsteered to .001" change in toe from full compression to full droop.  I remember all too well how small the adjustments were to get it dialed-in.  Assuming NB bumpsteer is good, moving the inner tie-rod pivot nearly 1/4" is a huge change.

You could get back to normal NB bumpsteer w/o "improving" the turning radius.  Your spacers would need to be larger than the rack housing, so they would continue to restrict rack travel just like the narrow MSM shaft does.

As others have posted, the ratio of the rack is unchanged.  The parts list doesn't offer an MSM pinion gear and the lost stoke vs. "missing" turns at the steering wheel works out close enough.   

['95MSM]

Here is today's update, and it is definitely good news / bad news:

The rack shaft seems to have all of the normal teeth cut into it.  I could pull the end of the shaft at least 1/2" further into the housing on the passenger side (we only need 1/4", maybe a little less).  I didn't bother testing the driver's side, as I can see about 3 full teeth cut into the rack that are full of grease - they've never seen the pinion gear.  So more than enough gear stroke is available. 

The passenger side seal is deep enough that it will never be an issue, and there is no seal or holder on the driver's side of the rack, so no problems pulling the rack further into the housing than Mazda intended on the MSM.

The locking tab system, however, is a problem.  The end of the rack shaft has two slots cut into it, sort of like a partial castle nut.  The locking tab washer is keyed to these slots, so that when the tabs are bent down over the knuckle flats, the knuckle cannot turn in relationship to the shaft.  It would be possible to machine spacers with male tabs that engage the slots on the shaft and female slots for the locking washer, but at that point it is probably cheaper to simply buy an NB rack.  For the true DIY guy, it might be possible to find some heavy walled tubing (Aircraft Spruce is always a good source if 4130 chrome moly will work) and shape it with hand tools, but it would be a lot of work.

Another alternative is to rely on permanent Locktite or something similar, or maybe install the knuckles with the spacers, mark how they line up with the castle-nut like slots on the end of the shaft and then drill them for a roll-pin.  I may try the Loctite route, using a primer to get a thorough bond.

[DropTopDrifter]

Interesting. I've known for years that MSM rack has a larger turning radius compared to the standard NB and had guessed that it was a last minute deal because of the 17" wheels and the fear of them rubbing...rather than coming up with a different offset wheel to prevent this. It's easy for me to compare a NB to a MSM back to back since both are sitting in my garage. I remember the first time I went to do a u-turn in Jenn's MSM and was quickly greeted with not being able to make the turn in the space given. Right after, I jumped in my car and did the same thing but was able to make the turn. I know a few people that were drifting with MSMs and not doing so well and know that the steering rack difference is a big difference and am surprised that more MSM owners don't swap them out for a normal NB rack. Anyone who knows NBs well will always wonder what Mazda was (or wasn't) thinking with the MSM. I wonder why Mazda didn't follow what they did in Australia with the older SE Turbo versions that used parts/advice from a reputable aftermarket tuner (AVO).

2002 Turbo SE (Aus.)

The Australian MX-5 SP Turbo is a local development which is essentially a turbo kit mounted to the existing 10:1 compression BP-VE engine. The SP Turbo package employs a nickel-alloy cast iron exhaust manifold, a ball-bearing Garrett turbocharger, front-mount air-to-air intercooler, a carbon-fibre airbox, Bosch blow-off valve and large diameter exhaust. Supporting mods include bigger injectors, different spark plugs and an upgraded coolant radiator. The factory ECU has also been re-mapped to suit. With around 0.5 Bar boost, the turbocharged MX-5 SP cranks out 150kW at 6800 rpm and a massive 280Nm at 4600 rpm – that’s 39 percent more power and 60 percent more torque than a contemporary atmo version! This engine is second only to the rally-bred Familia GTR spec BP turbo.

[scotty b]

My Dakota can do a pretty tight u-turn, despite it's large turning radius, as long as Johnny Law isn't around   My MSM could make a pretty tight u under the right conditions, just don't forget to countersteer...

[Davex3]

Wow, interesting with the msm steering rack. It always makes me wonder why Mazda sometimes took a step backwards with the msm on some parts.

And the infamous SP. That carbon fibre airbox will cost you $5,000 au new from Mazda. 

['95MSM]

I've never seen a decent pic of the SP airbox before - SWEET!  Love the inlet airhorn.

[DropTopDrifter]

I've never seen a decent pic of the SP airbox before - SWEET!  Love the inlet airhorn.

It's almost like someone knew that a turbo-red engine needed a bigger air box inlet and even noticed the fresh air source under the headlight...imagine that!