Anyone been here ? Gear Reduction Starter Fitment

Never underestimate the power of vendors returning product to manufacturers because vendors' customers are pissed off because the product is not satisfactory to the customers.

It's an essential part of any product development process:cool:
 
I gained a little over 2mm of clearance just by replacing the supplied nut/washer on the starter stud with a shorter-profile nut & washer. Grinding off the now-exposed end of the stud would net that same amount of clearance between the tip of the stud & the engine block wall.

Following up on this, I ground the now-exposed end of the starter stud to where it's just about flush with the top of the replacement short-profile nut. I peeled back the blue silicone boot to expose the stud & nut, & then mounted the starter motor onto the bellhousing so I could see the actual clearance between the side of the engine block & the newly-shortened starter stud/nut. Net result is an increased clearance that I'm now completely comfortable with - looks to be at least 4mm-6mm or so. With the boot reinstalled, there should be nothing to worry about, & no need to relieve the side of the block or re-clock the starter motor mount.

040.jpg 041.jpg 042.jpg
 
You might want to have at least an inkling of actual experience with the item & the issue being discussed here before going off on a tirade against the part &/or the vendors supplying it. I have 2 of these IMI 850 starters, & have had 3 others in the past, & they ALL fit just fine. The issue being noted (close proximity of the solenoid cable stud end to the block) does NOT prevent the starter from being mounted, & it does NOT prevent the starter from working as intended.

Could the manufacturer modify the starter further to help alleviate buyers' concerns in this area? Most definitely. Simply changing the thickness of the hardware used would go a long way. Axially re-clocking the 2 main starter body-to-base bolt holes (&/or the 3 stud holes) might help a little more, but may also introduce unwanted access/interference issues with the 3 starter mounting studs (mainly the upper right one). It's highly unlikely that the manufacturer gets much customer feedback on this starter motor, let alone any improvement suggestions. This may be an opportunity for us as 850 Owners to do both.
Hello Jeff,

You did a bunch-o-work for sure.

Thanks from me and behalf of others who may be fitting this one.

Like your core sampling approach in lieu of a sledged slab. One can assume
that the block was unusable. Scrap? Metal salvage worth. Perhaps an anchor
if you have a large enough winch on your ship (smile).

-The measurement of the available web in that local area. Non machined surface(s).
Sandcast. Surface characteristic RA/RHR, difference in tooling through time, degradation
on internal jacket through erosion, buildup on internal jacket through deposit.

-The measurment itself. Tooling calibration. Normal to measured surface.

- As you know, more variables are available when determining thickness.
I've listed an applicable few off from the top of my head.

All this random thought?

What I'm pointing to is the tolerance on the actual measurement.
What do you figure.........plus or minus 2mm perhaps?


On a used block, a 5.4mm measurement suggests a of 3.4 - 7.4mm thickness given a 2mm tolerance.
Unknown what a factory fresh minimum wall thickness requirement is there.

Originally, I was thinking of carving to relieve the interference. Knowing the available thickness now,
I don't think that would be an optimal approach for me. One caveat. Sample size. One block evaluated. Is it
representative ? Enough for me to make a decision (smile).

Attached are some pictures of what I am encountering. It differs from yours on measured
clearance. Differences due to trimming ? Block variance? Axial placement of mounting holes on starter ?
Variance in bellhousing mounts ?

Shown, are my flat washer and nut dimensions.
I have a three thread engagement.
Also shown are the clearance dimensions.

I still encounter compressive force on the supplied silicone protective cap.
So much so that it is decreasing the original nominal wall thickness of that cap.

Notice the slope of the block compared to the face of the hexnut?
Not parallel. Parallel would yield an even airgap.
It is much closer on the bottom than it is on top.

A slight rotation of the assembly would yield a parallel relationship
and increase the airgap of the lower face on this one.

(And) as you voiced earlier....rotating the assembly might create an even greater hardship
mounting that assembly onto its holding studs. The attach hexnuts will begin interferring with the
motor body. I still think there is some wiggle room here though (wink).

thoughts?
 

Attachments

  • stackup.jpg
    stackup.jpg
    95.1 KB · Views: 84
  • thread.jpg
    thread.jpg
    41.9 KB · Views: 81
  • top.jpg
    top.jpg
    94.2 KB · Views: 92
  • bottom.jpg
    bottom.jpg
    95 KB · Views: 76
Last edited:
Hello Jeff,

You did a bunch-o-work for sure.

Thanks from me and behalf of others who may be fitting this one.

Like your core sampling approach in lieu of a sledged slab. One can assume
that the block was unusable. Scrap? Metal salvage worth. Perhaps an anchor
if you have a large enough winch on your ship (smile).

-The measurement of the available web in that local area. Non machined surface(s).
Sandcast. Surface characteristic RA/RHR, difference in tooling through time, degradation
on internal jacket through erosion, buildup on internal jacket through deposit.

-The measurment itself. Tooling calibration. Normal to measured surface.

- As you know, more variables are available when determining thickness.
I've listed an applicable few off from the top of my head.

All this random thought?

What I'm pointing to is the tolerance on the actual measurement.
What do you figure.........plus or minus 2mm perhaps?


On a used block, a 5.4mm measurement suggests a of 3.4 - 7.4mm thickness given a 2mm tolerance.
Unknown what a factory fresh minimum wall thickness requirement is there.

Originally, I was thinking of carving to relieve the interference. Knowing the available thickness now,
I don't think that would be an optimal approach for me. One caveat. Sample size. One block evaluated. Is it
representative ? Enough for me to make a decision (smile).

Attached are some pictures of what I am encountering. It differs from yours on measured
clearance. Differences due to trimming ? Block variance? Axial placement of mounting holes on starter ?
Variance in bellhousing mounts ?

Shown, are my flat washer and nut dimensions.
I have a three thread engagement.
Also shown are the clearance dimensions.

I still encounter compressive force on the supplied silicone protective cap.
So much so that it is decreasing the original nominal wall thickness of that cap.

Notice the slope of the block compared to the face of the hexnut?
Not parallel. Parallel would yield an even airgap.
It is much closer on the bottom than it is on top.

A slight rotation of the assembly would yield a parallel relationship
and increase the airgap of the lower face on this one.

(And) as you voiced earlier....rotating the assembly might create an even greater hardship
mounting that assembly onto its holding studs. The attach hexnuts will begin interferring with the
motor body. I still think there is some wiggle room here though (wink).

thoughts?
Loosen the bellhousing bolts a bit and see if the engine will rotate slightly in relation to the bellhousing. It might be that your trans (engine actually) is missing the locating dowels/sleeves? that ensure the trans is clocked correctly in relation to the engine.
They should be on the lower (2) bellhousing bolts.
 
Loosen the bellhousing bolts a bit and see if the engine will rotate slightly in relation to the bellhousing. It might be that your trans (engine actually) is missing the locating dowels/sleeves? that ensure the trans is clocked correctly in relation to the engine.
They should be on the lower (2) bellhousing bolts.
Hello Gene

Thanks for chiming in !

What you express is germane and a consideration.
Made me go back and take a peek at the whole project (smile).

All dowels are pressfit on both mating halves.

It is hard to establish a natural index on the bell housing.
(But) I think I found one. Shown on the attached image below.

This is the bottom-most inboard fastener through bell housing to block.
Notice the natural intersect ?


I am now inclined to start trimming the other available hexnut on the
electrical stud. Effectively reducing its height also. There is some risk there as that
trimming will start moving the electrical busbar closer to ground. The risk is slight though
as there is plenty of margin there to work with and it too has its own insulative sleeve.

An addition stud protrusion trim to follow accordingly.

The quest continues.........
 

Attachments

  • Index.jpg
    Index.jpg
    137.5 KB · Views: 88
  • trim2.jpg
    trim2.jpg
    87.3 KB · Views: 80
Following up on this, I ground the now-exposed end of the starter stud to where it's just about flush with the top of the replacement short-profile nut. I peeled back the blue silicone boot to expose the stud & nut, & then mounted the starter motor onto the bellhousing so I could see the actual clearance between the side of the engine block & the newly-shortened starter stud/nut.
A slightly better view of the improved clearance between the engine block & the replacement starter stud/nut:

045b.jpg 046b.jpg

I now have an actual (measured) clearance of 2.24mm between the engine block & the lower left side of the starter stud nut (as pictured), with ZERO chance of the 2 making metal-to-metal contact. The blue silicone boot does still contact the block when installed, but that's really a non-issue at this point.

More clearance could be possible if an even shorter-profile nut were to be used, the one shown is just something I scrounged up out of a box of random hardware I had on hand. I will likely try a shorter one on my other (new) IMI 850 starter - might be worth another 1.0mm-1.5mm to be gained?
 
A slightly better view of the improved clearance between the engine block & the replacement starter stud/nut:

View attachment 52514 View attachment 52515

I now have an actual (measured) clearance of 2.24mm between the engine block & the lower left side of the starter stud nut (as pictured), with ZERO chance of the 2 making metal-to-metal contact. The blue silicone boot does still contact the block when installed, but that's really a non-issue at this point.

More clearance could be possible if an even shorter-profile nut were to be used, the one shown is just something I scrounged up out of a box of random hardware I had on hand. I will likely try a shorter one on my other (new) IMI 850 starter - might be worth another 1.0mm-1.5mm to be gained?
Hello Jeff,

Thanks for your continued effort on this one.
Much appreciated !

Much clearer pics.

Your pics now Illustrates the slope relationship to the face of the
hexnut well.

We're zeroing in on this one. The math is adding up.

I suspect that your holdnut on the stud is around 1.5mm in height
different than mine.

I also suspect that your are experiencing this angulation relationship
of the starter to the bell housing. Please reference the attached rough pic
shown below.

Do you have the freetime to confirm this relationship of
the assembly to the bell housing edge?
 

Attachments

  • clock.jpg
    clock.jpg
    212.7 KB · Views: 96
That engine and transaxle are entirely too clean. Must be a simulation. ;)
Hello dnudelman,

Ah, C'mon now.....you can clearly see debris in the thread root of the axle flange mounting studs shown in this picture.

Nice catch. I just might chase them....ha ha ha ha
 
Re: IMI starter assembly sold by our favored vendors.

Mounted on the 850/903cc block. Notice the riding condition
on the electrical stud/nut which has been expressed by MANY, repeat
MANY people who have fitted this starter.

I'm not happy with the riding condition. (Even)
with the protective silicone cap squished against
the cast iron of the block acting as insulation.

Vibration, fatigue, and degradation of the elastomer boot comes to mind.

What I have done to help mitigate the riding condition is
reduced the height of the interferring electrical stud by 4mm,
and trimmed the height of flanged hexnut on that stud by 3mm.
Clamping force has still been maintained on that stud. It is now
at its barest minimum though.

Still fitting too tight to the block for my liking after the stud/nut trimming.

One vendor has recommended to relieve (locally grind) the block
in that area to reduce this riding condition. This seems a good
course of action providing that we don't thin it too much in that area
and create stress risers.

Anyone relieved the block in that area ? How much material is available there ?


Thanks upfront into any insight on this one.

lezesig
'72 850
-FINAL-

Obvious to me that the designers for this fitment on the 903 block concentrated on this riding condition-clearance issue.

Indicated by a shorter bakelite insulative cup on one side of the starter versus the other.
(Also) indicated by the use of a shaved hexnut on the interferring post. Surface condition of
that nut indicated either it was shaved from stock, or a special nut was fabricated from hexstock.

The presence of an O-ring for sealing and compressed at the same time as seating on the base
of the cup indicates focus in this area. Kudos to the team.

I didn't want to carve into the block for clearance. Thanks Jeff for the providing the web thickness on
the block in this area.

What I did to provide additional clearance......

The gap on the bottom, which is the closest the stud gets, is now .092", or 2.3mm.
This was achieved by trimming four components. Reference the attached pics.

The TOP has a much wider gap as compared to the BOTTOM, as shown in the attached pic.
Reclocking the starter slightly would optimize this gap and produce a much better result.

I am satisfied with the clamping force exerted by each trimmed hexnut now but they have now
been trimmed to their barest minimum IMO.

Note: The blue colored elastomer protective cap was not modified. The two original flat
washers were re-used.

out for now,
lezesig
'72spi
 

Attachments

  • BOTTOM.jpg
    BOTTOM.jpg
    111.4 KB · Views: 91
  • REVEAL.jpg
    REVEAL.jpg
    180 KB · Views: 89
  • TOP.jpg
    TOP.jpg
    110.4 KB · Views: 93
  • TRIMMING.jpg
    TRIMMING.jpg
    152.5 KB · Views: 94
-FINAL-

Obvious to me that the designers for this fitment on the 903 block concentrated on this riding condition-clearance issue.

Indicated by a shorter bakelite insulative cup on one side of the starter versus the other.
(Also) indicated by the use of a shaved hexnut on the interferring post. Surface condition of
that nut indicated either it was shaved from stock, or a special nut was fabricated from hexstock.

The presence of an O-ring for sealing and compressed at the same time as seating on the base
of the cup indicates focus in this area. Kudos to the team.

I didn't want to carve into the block for clearance. Thanks Jeff for the providing the web thickness on
the block in this area.

What I did to provide additional clearance......

The gap on the bottom, which is the closest the stud gets, is now .092", or 2.3mm.
This was achieved by trimming four components. Reference the attached pics.

The TOP has a much wider gap as compared to the BOTTOM, as shown in the attached pic.
Reclocking the starter slightly would optimize this gap and produce a much better result.

I am satisfied with the clamping force exerted by each trimmed hexnut now but they have now
been trimmed to their barest minimum IMO.

Note: The blue colored elastomer protective cap was not modified. The two original flat
washers were re-used.

out for now,
lezesig
'72spi

Great work! Perhaps I'm an idiot, but I didn't much notice this issue with mine. So I crawled under my coupe this weekend to do a bunch of maintenance and took the opportunity to take a pic of this situation on my car.

1633365299571.png


Now I'm concerned I should take it our and make sure the blue boot isn't being crushed and worn through. I've had this starter on my car for a couple of years now. Hmmmm...
 
Back
Top