The importance of measuring bell housing runout.

-

Kern Dog

Build your car to handle.
FABO Gold Member
Joined
May 23, 2010
Messages
9,759
Reaction score
28,097
Location
Granite Bay CA
The topic of bell housing runout has gained a lot of attention the last few years with the rise in popularity of the aftermarket overdrive manual transmission kits.
Years ago, I never heard of it, never heard of anyone even talking of it. I’d hear guys talk of clutch chatter, the transmission popping out of gear, hard shifting, noise and stuff but did not ever hear that any of those issues being blamed on the runout.
For those that haven’t heard, I’m referring to the relationship between the centerline of the crankshaft and the centerline of the transmission input shaft. When our engines and manual transmissions were first mated together, the story is that the factory put the bell on the engine and line bored the bell housing to the centerline of the crank. Any mis-alignment that may have existed was now corrected and would be eliminated….until you tried using a different bell housing.
I don’t have a lot of experience with Mopar manual transmissions but the ones I’ve had didn’t have any problems despite parts swapping without any regard for runout.
Now though, the companies offering these Tremec and other kits require proof of the measured runout to validate the warranty.
How close do you have to be?
Can you believe no more than .005” off center total.
Whuuut?
I’m a Carpenter…. All my life I dealt with framing houses and schools where if we were within 1/8 of an inch, we were good. (Kidding)
.005 now?? A human hair is wider than that, isn’t it?
I know a guy that didn’t measure his runout at all and his Tremec was fine. He beat on that car then sold it because he wanted to build a house and move out of the teepee he was in.

7508B639-F540-44AC-A889-1ED8D9535FB0.gif



Yeah…..a teepee.

Tee pee ME.JPG



Back on point:
What is it about these new transmissions that the vendors require such precise alignment?
.005 is the limit for the SST kit. If you’re at .006 they won’t honor the warranty??

I did the dial indicator measuring of the runout twice in my car…. The first time WAS my first time ever. The second time was after I pulled and rebuilt the engine. It was a LOT easier with the engine out of the car.
Now I’m helping a FBBO member with his car and the initial number we got is over .080. That is.040 off center, more than a standard spark plug gap.
Correcting that could be difficult.
There are companies that make offset dowels to shift the bell housing around to align the bell to the crank but the available sizes are limited. Nobody makes a .040 dowel from what I’ve seen. SST suggests scraping any and all paint and crud from the mating surfaces of the engine block and bell. I’d be surprised to learn that it helps but it is a fast and no cost effort to try.
In the meantime….I’d be interested in opinions and comments on those that have dealt with unanswered transmission issues that may have been related to a misalignment.
Finally… how is it that automatics don’t have this issue?
Maybe they do.
It could explain cracked flex plates or leaky input seals?
 
Don't misunderstand me....I do know the importance of precise measurements. I'm just curious as to why it seems that the new overdrives are so sensitive to runout.
My 440 block only needed .014 offset dowels to bring the bell into range.
The car I'm helping with now is over .040 off center. I wonder what the chances are of the bell itself being to blame. I wish I had another bell nearby to check. I'm going to put the word out to some local friends. I'd like to borrow one to do a test fit and measure.

The vendor requires documentation of measuring the runout to no more than .005 or less to warranty the transmission. When you spend $7000 on one of these kits, you're inclined to do what it takes to get it right.
I have one friend that set aside a bellhousing for me to use. I wouldn't mind another though. If the other bells also have excessive runout in the same direction, it seems that would indicate an indexing issue with the crank not being centered or the mounting surface of the block being drilled out of spec.
Sometimes weird stuff happens and you have to find a way around it.
To the topic of automatic transmissions...
I noticed during my 5 speed swap when I pulled the flex plate off of my 440/493, it had 2 cracked ears.

SST 162.JPG

SST 163-2.JPG



I had a buddy that cracked the hubs on 2 different flexplates within a couple of months on this car:

MME.JPG



It was just a mild 360.

What about you? Have you accidently found broken parts like this and didn't know what caused it?
What if all along it was excessive runout?
 
The original dowels in my 440/495 during my Tremec installation:


SST 372.JPG


SST 373.JPG


SST 374.JPG


Those look really short to me. I wonder if the holes in bellhousings are tapered at all at the edges. If so, these dowels are almost useless!

SST 379.JPG


You're looking at the ends that were in the block. I welded lug nuts to the face side to crank them out.

The new ones were much longer:

SST 360.JPG


Look how much they stick out past the bell:

601 I.JPG

601 J.JPG
 
The topic of bell housing runout has gained a lot of attention the last few years with the rise in popularity of the aftermarket overdrive manual transmission kits.
What is it about these new transmissions that the vendors require such precise alignment?
.005 is the limit for the SST kit. If you’re at .006 they won’t honor the warranty??

Here is a screen shot of a YouTube video showing the tapered roller bearing used on the input shaft of the Tremec transmissions.

01 r o t.png



That explains why the runout has to be so small...The bearing can't tolerate much misalignment.
 
This is a screen shot of a Youtube video made by the company the Tremec kit came from.

1697012326359.png


You all know the phrase..."A picture is worth a thousand words" ?
Well, a video takes that concept a lot further. I had an eye opening moment while watching that video.

To recap, I mentioned first reading about measuring runout in a Mopar Action magazine where Rick Ehrenberg said "Spend the 20 minutes on this and get it right".
It pissed me off because it took me several hours to do this. 20 minutes? What the hell, man?
Watching this video, a couple of simple moves they did suddenly made total sense and it made me see how it is possible to do this with precision in far less time than before.
First, they aimed to find a zero point, then the point which was furthest away, the highest number.
We did that. The number was .081 or .082. THEN they RE-ZEROED the gauge at that highest point.
Now, that new zero point is the direction that the bell has to move to.
Yeah. THAT alone made my eyes perk up. It is so simple but makes sense.
In the case of this car, the zero point was still at 6:00 but the high number is at what was it....12:00 or 2:00? Regardless, THAT will be the new zero. THAT is where the bell has to shift toward to close that number.
Secondly, they installed their dowels with the offsets aimed to where they need to move the bell.
Again, so simple, I'm pissed that I didn't think of it. With my car, I think I did that but can't be sure.
The SST guys didn't test the runout at 4 points, 12 points or anything like that. They tested it at 2 points.....the highest number and lowest. Simple.
I feel like a kid with a new toy....I want to test what I've just learned.
 
I think most people don't realize how important it is to align the bellhousing on a car. It takes time but is worth the effort. Not a (Noticeable) horsepower gain but can save big headaches down the road.
 
yeah thats the best video it makes it all make sense...

the hardest part of the process is getting the indicator to stay put and not fall off. look at the complex arm they have that allows the magnetic base ro be way up on the clutch surface with an articulated jointed support i.e not the stand and arm you get at the local tool shop unless you bend the credit card somewhat

the reason for the focus these days is the fact that the clerances tollerances and as you say type of bearings is now different clearances are tighter cone and cup bearings used instead of bushes or ball bearing based bearings

we are talking manual transmissions that work best on auto trans fluid not viscous gear oil

and most of the have the gear train loaded with near zero end play or some pre load for cone and cup
previously everything was just shimmed in the case and ran on bushes or loose rollers

gears are made from harder materials shafts have less flex,

these days nothing can move in any way other than rotation, unless you have enough torque to stretch the case. no play to cater for a wonky bell.

basically your are manually installing a transmission that has its roots deeply sunk in a design philosophy where the trans the bell and the motor were machined by robots and the lot was supposed to be installed by robots

we are ramming a precision cut diamond into the back end of something that is closer to a steam train than a Prius


Dave
 
Back in the 80's my buddy (a Mopar guru) Dialed in my bellhousing on my 4 spd conversion. It was off and he installed offset dowels. He had dowels because he had ordered some when he put a Doug Nash 5 speed in his 70 Duster.
 
Remember that while you are working on aligning the center-line of the crankshaft to the centerline of the transmission by using dowels, you must also insure the bell-housing sits squarely on the back of the block.

This is to verify the transmission centerline sits both square and plumb in relation to the crank centerline.

I've had to use shims between the block and bellhousing face to get the transmission to sit square and plump to the crank centerline. This is in addition to centering transmission input to the mounting hole.

Shims.JPG
 
When I switched my Dart over to 4spd 22 years ago, I picked up a used Lakewood blowproof bellhousing and all the parts and pieces to put it together. Well it was the worst shifting 833 I ever drove. Long story short found the same thing you found Kern Dog, but my was around .040 to .050 out if I remember right( way to much for offset dowel pins). At the time I found an article about this problem with Lakewood bells (magazine or a book don't remember) the article said to drill the dowel pin holes out over size in the bell, install longer dowel pins, find large flat washers with the center holes that fit snug on the dowel pins (I found two 7/16 washers that fit snug on the 1/2 inch dowels). Bolt the bell on snug enough so you can move it with a few taps of a mallet, tap it around till you get the runout to less than .005, place the washers on the dowel pins and weld them to the bellhousing. That 833 shifted a hell of a lot better after that.
 
I switched the Lakewood bell out to a Quicktime bell when they came out with an SFI rated bell. That bell was with in the .005 run out, but I had to do like jbc426 said and shim the top of the bell out .010 to square it to the back of the block. Pays to always check all driveline parts.
 
I am convinced this was a problem right from OEM on my 70 RR 440-6. It had one of those hated scalloped fake 11" clutches (10.9 or whatever you want to call them) and it never released that clean, and wore pilot bushings like nobody's business.

Like you, we we all kids then, no internet, and even though it might have been mentioned in a magazine or two, well, ..............................
 
I am trying to look at yet another issue with our flawed cars, from a humorous angle. I'm not carring enough, give me some more, and I just got the engine and trans back in the '69 ! Probably, 97% of us should just leave our cars in the garage, there are just too many flaws that we don't know about!
I'll check that the next time I have the engine out. I never would have thought about the bell being square to the block. Next time, I'm driving it now.
Let's be safe out there!
 
.004 is the average thickness of a human hair. .080 TIR is alot of miss alignment. It doesn’t surprise me that you have to shim a scatter shield flat to the back of a block. Those shields are stamped and formed in a punch press. Being out of flat will add to the miss alignment so shim it before you indicate it true. To keep it from moving around, shim bolt holes to look like a washer, you know a shim with a hole in it.
 
I laugh at guys selling used scatter shields for big money, more trouble than they are worth. And definitely if the are out of date safety spec.
 
I do think it's important. When I finally start the T5 swap on Vixen, I will measure the runout since I am using a transmission adapter from Gill Welding.

Having said that, I wonder what the factory tolerances were? Not listed tolerances somewhere on paper, but ACTUAL MEASURED tolerances? I mean, if their actual deck heights and combustion chamber sizes were any indication, I'd say those tolerances were pretty wide.

Again, I'm not arguing against doing it right. I just wonder. Did the factory measure runout on every bellhousing that left the factory? Of course not. I guess what I'm saying is, there has to be a pretty wide tolerance there, or else we'd have seen lots of damaged parts from the factory. I don't think there's any way they were all the same or even within the same range when they left the factory.

At least that's my opinion. lol
 
I struggled with this setting up my TKX.

Most of us don't have a snap on magnetic base and dial indicator. I used Harbor Freight since it's something I use every once in a blue moon. Wound up having to remove one of the flywheel bolts to get the magnetic base to stick. Probably rotated the engine around 30 times until I finally got the reading to stay consistent from rotation to rotation.

Bought the offset dowels, installed them the way the video says to install them and it actually made the run out worse. Wound up just guessing and turning them 90* off of where I expected them to be pointing and tested a handful of times and got consistent readings within spec.

I don't think you have to do anything other than write down what your runout is and send it to SST in order to stay in warranty. Maybe my memory isn't serving me correct but I feel like anyone could just lie and write down whatever they wanted. Probably not smart unless you want to up your chances of having issues down the road.
 
Lakewood had a terrible quality control problem. There were three guys in the late 70's/ early 80's involved with 4WD swaps. One put a Chev SB, Jeep, Lakewood, Ford 4 speed. Out of spec by far. another had a stock early 70's Jeep V6, Lakewood, Ford 4 speed. The 4 speed bolt pattern was drilled "twisted" and the transfer case hung down further as if the engine sat rotated in the mounts. The last one I never understood, he sent it back for another, at considerable shipping expense.
 
I am the owner of an old Lakewood, I was in the ballpark, with filler insert for the small bearing retainer and I don't know how many rotatations and 2 different sets of offset dowels. Never thought of the block to bell being correct. Having said this, don't think that I would ever put some kind of tapered shim between the block and the bell if it was off. My brain has turned to jello in all of the planes possible from top to bottom. Not saying that it shouldn't be checked, but have no idea how how I or most people would ever correct it if it was off.
 
Did the factory measure runout on every bellhousing that left the factory? Of course not. I guess what I'm saying is, there has to be a pretty wide tolerance there, or else we'd have seen lots of damaged parts from the factory.
The story that I heard was that the manual trans cars had the bellhousing mounted to the engine, then the bell was line bored to center of the crankshaft.
I did write that in the first post.
 
The story that I heard was that the manual trans cars had the bellhousing mounted to the engine, then the bell was line bored to center of the crankshaft.
I did write that in the first post.
I gotta say, I just don't buy that. Just imagine how many millions of vehicles they'd had to do that on. Stop and think about it. That would have cost them a crapton.
 
however the process to both shim and align a bellhousing is in my factory manual
decsribed in a way that makes you mentally i'll....:) covers both auto and manual
they say it must be done if you swap the bell from the orginal one for a manual trans

1971 VH Valiant Australia factory manual... presume others have it as well. chrylser stock photos rather than car specific. i.e numerous LHD photos in a manual for RHD Cars in the suspension and drive train sections

the first post video is prefect for run out...its a great thing
 
Last edited:
Chrysler thought it was important enough that the bellhousing was line bored with the block when new. Perfect alignment. Anytime machine work is done to an engine, or a bell/engine is changed is when you get misalignment.
 
Notice the bellhousings bolted to some of the engines being assembled, but not all of them.

bellhousing bore allignment-factory.jpg
 
-
Back
Top