The Top DSM Community on the Web

For 1990-1999 Mitsubishi Eclipse, Eagle Talon, Plymouth Laser, and Galant VR-4 Owners. Log in to remove most ads.

Please Support JNZ Tuning
Please Support ExtremePSI

Setting bearing preloads

This site may earn a commission from merchant
affiliate links, including eBay, Amazon, and others.

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

bwhite

20+ Year Contributor
1,039
2
Mar 5, 2002
Brook Park, Ohio
Ive changed my shift forks and wanted to set shaft preloads. I used the fsm and it says to use .06 solder which I got and measured with but the center diff and the input shaft are not crushing the solder at all. Should I just use the larger 3/32 solder and remeasure? The intermediate shaft shim is. 038 and the solder measurement came up with .042. Does that mean I need a .012 thicker shim to get my .008 of preload?
 
i USED 3/32 AND .062 AND THEY both gave about the same reading but the rosin core lead solder (soft stuff) for plumbing wasthe most consistant, and gavethe best results,

you add to the solder to get new shim size, so with .042 and wanting to shim .008 you need a shim that's .050" thick, I have taken apart a lot of trannies lately and found that there's a lot of shims in bearings that are different thicknesses and im so damn poor that i've had to take and cut them and bend themsmaller or open up and silver solder in an extension from another one the same size instead of buying them all that i need, good thing too since i went through two trannies in under a month
 
So I remeasured with the 3/32 and got some better readings but the input shaft doesnt seem right. My original shim was .040, my solder measures .057. So even with the thickest shim I found its still only at .0005 preload. Does that seem right or ok preload? The way I see it is if I had .017 in of endplay, thats within factory spec so the bearings cant be bad. Is there different shims for running that high of preload?
 
No, tha'ts not right i'd be willing to bet from what i've read on here of tim giving advice and similar results myself

Here's afew thigns that i found REALLY matter..get SOFT 3/32 solder, cut the pieces SHORT like only 10mm and place them EXACTLY where the book says/points with the lines in the manual... make sure the bottom of mid case is tight to bell housing, torque it all with the rear cover on (some people talk of leaving it off, i don't have luck like that)

I check each one i have done 4-5 times between both solders.. worsst case put solder on one shaft at a time so there's not a bunch taking the load.. i have seen with used cases and new bearings readings as follows, input is usually in the mid .040's like .043. 046. intermediate is about the same and the center diff is usually around .051 -.053 I had to make customshims from shim stock i have formachining to go under the input shaft one to get it right at .007" preload, i put my int. shaft at .007 -.008, and the CD at .060 shim which is about .007... That's tighter than the first build but although i coudln't turn the input by hand with it in gear, the trans is awesome now that the cluch it broken in

Also, try putting the solder in and leaving it for an hour or 35-40 minutes, , and remember get the soft plumbers stuff, cut pieces SHORT (less area to absorb the crush load ) and tighten everthing to spec and wait, or as i said do one shaft at a time, shim each one you do as you prepare the next one so it gives accurate account of what the trans wil be like assem,bled (my own thoughts there)

But mine clicks into 2md and third when cold at low RPM, higher irev easier it shifts, then once warm no clicking and still shifts great at over 8k RPM when i lost traction playing in a wet parking lot :D

Sorry so long to reply, my OS registry got f**ked yesterday and i had to format and boot up a second SATA drive into a dual OS to be able to save much of my needed stuff, just finihed tonight
 
Yes, your input measurements are common for a trans that has been untouched. 99% of the time, we have to add a .010" thicker shim in order for the input shaft to be snug enough. Depending on the year of the unit, some are worse than others.

I wouldn't look at it as a problem. Due to the fact that many DSM trannies have the input shaft way too loose, your .0005" preload should be fine for the average car. If you are going to run a lot of power, I would try to tighten it up a bit more, though.

Another option is to replace the input shaft end bearing at the bell-housing side of the case. Mits has a new revision of that bearing which helps move the shaft up a bit to help counter the loose preloads.

Keeping the input shaft loose allows for faster and a better feel when down-shifting. So for the average street car with 16G power, keeping it a little loose so as to allow for better shifting might be worth it. Too tight and it can start to feel a bit notchy if you road race the car.

Good luck!

Jack

So I remeasured with the 3/32 and got some better readings but the input shaft doesnt seem right. My original shim was .040, my solder measures .057. So even with the thickest shim I found its still only at .0005 preload. Does that seem right or ok preload? The way I see it is if I had .017 in of endplay, thats within factory spec so the bearings cant be bad. Is there different shims for running that high of preload?
 
While i won't dissagree with what you're saying Jack, becuase you've build more than i have in trannies.. there's a thread titled "shimming gurus' needed) on page 2 or 3 in this section where some one was getting large measures like this, and i wasgetting them too until using the proper solder and technique of making sure i wasn't using too long of pieces since it spreads load just like a full face clutch meaning the force won't crush it all as far as if shorter peices were used, I was just saying double check because i got those readings, then did it right and go measure that were in the "normal range" but then again i put new bearings on both input and intermediate shaftts when i built mine, if he has old bearings being re-used that might be a real measurement and not an errrorbut it's easy to make an error is all i was getting at
 
Well I remeasured again using 3/32 on the center diff and the .062 on the input and intermediate shafts. I repeated that three times in a row and came up with .051 on the input, .039 on the intermediate shaft and .057 on the center diff. This is a 92 trans with the early fork design. I think my main problem was that when setting the solder in I was setting to far towards the inside and the bearing race was smashing the solder into a triangle. My factory shims were input shaft-.040, intermediate shaft-.038, and center diff-.055. Now hows that sound?
 
Sounds way better. Now get the proper shim for your desired use of the transmission. I still put around 0.005"-0.007" preload on drag racing transmissions, and around 0.003" preload on street car daily driver transmissions.

It still sounds like worn tapered roller bearings that should be replaced, but here are the input shaft shim values and part numbers for reference if you are intending on simply reusing your current parts.

Part Description Part Number ID Number Thickness (in)
Shim - Input Shaft MD727658 56 0.0220
Shim - Input Shaft MD727659 65 0.0256
Shim - Input Shaft MD727660 74 0.0291
Shim - Input Shaft MD727661 80 0.0315
Shim - Input Shaft MD720937 83 0.0327
Shim - Input Shaft MD720938 86 0.0338
Shim - Input Shaft MD720939 89 0.0350
Shim - Input Shaft MD720940 92 0.0362
Shim - Input Shaft MD720941 95 0.0374
Shim - Input Shaft MD720942 98 0.0386
Shim - Input Shaft MD720943 1 0.0398
Shim - Input Shaft MD720944 4 0.0409
Shim - Input Shaft MD720945 7 0.0421
Shim - Input Shaft MD710454 J 0.0433
Shim - Input Shaft MD700270 D 0.0445
Shim - Input Shaft MD710455 K 0.0457
Shim - Input Shaft MD710456 L 0.0468
Shim - Input Shaft MD700271 G 0.0480
Shim - Input Shaft MD710457 M 0.0492
Shim - Input Shaft MD710458 N 0.0504
Shim - Input Shaft MD706574 E 0.0516
Shim - Input Shaft MD710459 O 0.0527
Shim - Input Shaft MD710460 P 0.0539
Shim - Input Shaft MD706573 - 0.0551
Shim - Input Shaft MD710461 Q 0.0563
Shim - Input Shaft MD710462 R 0.0575


You can always stack shims to get the desired preload you want.
 
Last edited:
Thank you. I used your list from another post you made. Thanks again.
 
I may be showing my ignorance here, but it seems as though everyone is doing the solder crush test without RTV application, when it seems it would affect shim spacing. Shouldn't the crush test be done using the same installation methods as the final installation? Seems that the addition of RTV afterward would add endplay in the amount of the "crushed RTV".

There doesn't seem to be any kind of tutorial on this process, and I'd like to get it right. Yes, I know it's an old thread, but better then opening up a new on the same subject.

Another thing.. what temperature should the measurements be done at? When measuring with this amount of precision, you'd think temperature would count especially with multiple types of metal with different thermal expansion rates.
 
Last edited:
Guys, pretty sure most of you take apart your trans weekly. Please Help.

So for preloads, I'm used to preloads being a specified torque value, not thickness. What I *think* you do to get the correct spacer is take the exact measurement and add the desired "preload" or thickness to get your total shim thickness.

For my build I decided on

Input Shaft --> 0.005" - 0.007" PRELOAD
Intermediate Shaft --> 0.007" - 0.009" PRELOAD
Center Differential --> 0.005" - 0.007" PRELOAD
Output Shaft --> 0.003" - 0.005" PRELOAD
Front Differential --> 0.002" - .0067" ENDPLAY

This is what the FSM tells me, it doesn't say anything about RTV application, but I still feel not using it is going to skew the measure thickness!

You must be logged in to view this image or video.


Anyone I can call to figure this out? I want to get this figured out today!!
 
Last edited by a moderator:
You're over thinking this. RTV isn't going to add any extra clearance. Twicks and Jack do this daily. Trust what they say. Call them if you're not sure. They have their numbers on here.
 
I called Jacks, can't get a hold of TMZ I think he's moving his shop, and voice-mail says out of country so I emailed Tim.

What I gathered by calling Jack's:
They go by feel, and they change out various shim thicknesses until it feels right.

They say RTV flattens out to about a paper's thickness..... So tell me how would this not affect measurements when we are talking about *thousands* of an inch? Thermal expansion and surface prep should be considered when working with this kind of precision..... am I crazy or are these all factors? It's a 0.005 difference in preload from a stock build to 600TQ build, and I measured a piece of paper at 0.004 in.

I'm not comfortable with leaving my transmission build to guesswork.
 
Dude, it is less than 0.002" from the RTV; the RTV should be still soft and tacky; not set-up when you install the case section. You want it to fill the surface imperfections, not have it harden and install the case section to make a thick gasket.

You are over-thinking this.

Use 0.062" and 3/32" diameter solder in 1/2" - 3/4" length pieces evenly spaced on heavily lubricated (I prefer simple petroleum jelly on the bearings and races and the case section bearing race bores) bearing races making sure to not place the solder on the stamped part number lettering. Doing the final drive first, then the input shaft, intermediate shaft and center diff (or FWD output shaft) with shims in place for the final drive (front diff endplay for AWD or preload for FWD and then preload output pinion shaft for AWD or output shaft for FWD), always remember this:

The likely solder thickness to use for your tests will be:
0.062" diameter solder for:
*Input Shaft
*Intermediate Shaft
*Front Differential
*FWD Output Shaft

3/32" diameter for:
*Center Differential
*AWD Output Pinion Shaft

I normally have my shop at 62*-68* as well as the parts and the solder and lubricants. Install the shafts/diffs you are doing the tests on with the bearing races on the bearings and solder on the races. Slip on the case section(s) and use a pick tool to align the solder pieces so they do not slip out or jam in the case.

Tighten your case fasteners evenly by hand, then torque to 29 ft/lbs evenly. Let sit for like 10 minutes if you like, then disassemble and measure your solder thicknesses. If you have one of the two pieces either damaged or severely different in thickness (i.e. like 0.010" variation, half-crushed, etc.), then do it over until you get accurate measurements that are repeatable. Your ranges should look around the numbers below based on NEW bearings used with undamaged case sections (AWD as an example below):
Input Shaft: 0.035"-0.045" (shim range available is 0.017"-0.057")
Intermediate Shaft: 0.035"-0.045" (shim range available is 0.031" -0.056")
Center Differential: 0.040"-0.060" (shim range available is 0.044" -0.070")
Front Differential: 0.030"-0.045" (shim range available is 0.017"-0.057")
Output Pinion Shaft: 0.045"-0.060" (shim range available is 0.042" -0.075")


This isn't rocket science, but cleanliness, proper setup, and consistency with accurate measurement tools makes this job go significantly easier.


As well, yes, thermal expansion plays a major role in the requirement for accurate shimming. Taking a case and gearset from ambient to normal operating temperature (~120*-140*F) up to high temperatures (180*-210*F) to exceedance temperatures (>220*F) has a large expansion range of the case as well as the heat cycling or overheating of the gearset, components and breakdown of the gear oil. The transmission case can easily expand and contract 0.005" or more while the gearset is slightly less.
 
Last edited:
Dude, it is less than 0.002" from the RTV; the RTV should be still soft and tacky; not set-up when you install the case section. You want it to fill the surface imperfections, not have it harden and install the case section to make a thick gasket.

You are over-thinking this.

Thanks Tim, I love to over think things!
 
Guys, any help filling in the blanks here?

Intermediate shaft - the closest I could get was 0.056 (0.002 under-sized) or I could double the smallest available shim for 0.062 (0.002 over).
Center differential - it's not even close to the available shims
Output Shaft - closest is 0.075 (0.005 under) or 2 x 0.042 = 0.084 (0.002 over)

What do you think?

You must be logged in to view this image or video.
[DOUBLEPOST=1416185108][/DOUBLEPOST]Better measurement on the Center Diff. Bottom race was not compressed all the way. Still outside available shim thickness..

You must be logged in to view this image or video.


Also Attached the reference sheet I'm using to find my shim part numbers.
 

Attachments

  • W5M33 Shim PN.xlsx
    14.3 KB · Views: 289
Last edited:
Are these measurements with new or used bearings?? Also, that is seeming off for your center diff measurements still. Your input shaft bearing numbers look a bit high but can be worked with, the intermediate shaft numbers look like worn bearings, the output pinion shaft bearings look like worn bearings.

The front diff shims that can also be used are:
MN168165
MD710455

If you intend on reusing your bearings for all locations, then I would do something like:
Input shaft: MD710462 or MN168155 with MN168156 stacked (0.060" = 0.007" preload), or replace bearings and remeasure.
Intermediate shaft:MD712331 (0.006" preload), or MD720948 x2 stacked (0.063" total = 0.013" preload), or replace bearings and re-measure.
Center diff: Remeasure, your numbers are off. Confirm you don't have the bearing races in the wrong spot (i.e. cover race in lower case housing and lower housing race in bearing plate); the thinner bearing race goes in the bearing plate, the thicker race goes in the lower case housing.
Output pinion shaft: MD746510 x2 stacked (0.084" total = 0.007" preload), or replace bearings and re-measure.
 
Are these measurements with new or used bearings?? Also, that is seeming off for your center diff measurements still. Your input shaft bearing numbers look a bit high but can be worked with, the intermediate shaft numbers look like worn bearings, the output pinion shaft bearings look like worn bearings.

The front diff shims that can also be used are:
MN168165
MD710455

If you intend on reusing your bearings for all locations, then I would do something like:
Input shaft: MD710462 or MN168155 with MN168156 stacked (0.060" = 0.007" preload), or replace bearings and remeasure.
Intermediate shaft:MD712331 (0.006" preload), or MD720948 x2 stacked (0.063" total = 0.013" preload), or replace bearings and re-measure.
Center diff: Remeasure, your numbers are off. Confirm you don't have the bearing races in the wrong spot (i.e. cover race in lower case housing and lower housing race in bearing plate); the thinner bearing race goes in the bearing plate, the thicker race goes in the lower case housing.
Output pinion shaft: MD746510 x2 stacked (0.084" total = 0.007" preload), or replace bearings and re-measure.

I'll re-measure, but fairly sure I had it all the way flush with the lower case. The races are definitely in correct position, I put it in right as it came. I do have the 4-spider diff from you Tim, so this could be adding thickness?
As for the other bearings, most of them look brand new, the trans was rebuilt by previous owner 7k miles ago. The transmission didn't even have a bad failure or anything, just the center differential gave out under the power. Hence the 4-spider. It still shifted perfect with no grinds.

Any other tecnique to getting the lower case race seated in all the way would be very useful to me at this point. Current method is using an older race to compress new race (w/ assembly lube), hammering around edges.


Thanks for all the help so far!!
 
Got the Harbor Freight equivalent, so I could take it home and use immediately, this was the nicer set they had there came with a carrying case, color coded drivers, & higher build quality / materials.

You must be logged in to view this image or video.


Also got this "3-Jaw Pilot Bearing Puller" but it seems I need a 2-Jaw version :(

You must be logged in to view this image or video.


Also, re-measuring differential shim now, will report back. The Race *definitely* not seated all the way. New results were as expected with new bearings.

You must be logged in to view this image or video.
 
Last edited:
I realize this is a long dead thread and Tim (professional) already assured us that RTV thickness should be negligible, but thought it would be worth sharing my (amateur) measurements since I spent the time to take them. In my case the RTV seems less than 0.001"

Trans is in my basement at 68-70 F, using a (low/mid-range?) Fowler micrometer, averaging 2 measurements of each 1/2" solder piece (.062" rosin core). Both times I let it sit for a day after torqueing before removing to measure.

Initial is I took the trans apart, removed shims, and measured gaps. I then cleaned the RTV, cleaned/chased threads, and touched up the differential-side alignment pin as I noticed it had surface rust (hoping that is the reason for the significant change in that measurement, the normally blind receiving hole broke through the other side so it was exposed).

[measured gap as disassembled] - [measured gap after cleaning]
InputIntermediateOutputFront Differential
RTV Change-0.000650.000450.000150.002975

You must be logged in to view this image or video.
 
Support Vendors who Support the DSM Community
Boosted Fabrication ECM Tuning ExtremePSI Fuel Injector Clinic Innovation Products Jacks Transmissions JNZ Tuning Kiggly Racing Morrison Fabrications MyMitsubishiStore.com RixRacing RockAuto RTM Racing STM Tuned

Latest posts

Build Thread Updates

Vendor Updates

Latest Classifieds

Back
Top