Hot Rod Forum banner
1 - 11 of 11 Posts

·
Registered
Joined
·
111 Posts
Discussion Starter · #1 ·
I have dial indicated my bellhousing and here are the numbers I got. I started at the bottom at 0 and rotated counter-clockwise (clockwise from the front).
Handwriting Rectangle Font Parallel Wood


I ran the measurement 3 times and they come back consistently as shown above. So it appears that I have a 6 to 12 discrepancy of 0.008 in and a 3 to 9 discrepancy of -0.028 (FYI - I also ran it with 0 at 9 o'clock and I did get 0.028 at 3 o'clock).

Tell me see if I understand these numbers correctly. The bell needs to come down 0.004 in AND it needs to move left 0.014 in. Is this correct? In other words, here is what I am dealing with, right?
Handwriting Rectangle Font Wood Parallel


So... assuming this is correct, here is were I am struggling. I am not sure I understand how eccentric dowel pins can move the bell the way I want. Will I need 2 pins with different eccentricity?
 

·
Race it, Don't rice it!
Joined
·
8,277 Posts
Is the inner circle or the outer circle the bell?

I'm guessing the outer circle is the bell and if so your numbers are correct.
Move the Bell toward the 3 o'clock .014 Then move the the Bell toward the 12 o'clcok .004

The dowels are just on a eccentric lobe so you can turn them and move the bell around.
Hot tip.....Work with one axis at a time for simplicity. Shoot for +/- .005 tolerance.
 

·
Race it, Don't rice it!
Joined
·
8,277 Posts
Is the inner circle or the outer circle the bell?

I'm guessing the outer circle is the bell and if so your numbers are correct.
Move the Bell toward the 3 o'clock .014 Then move the the Bell toward the 12 o'clcok .004

The dowels are just on a eccentric lobe so you can turn them and move the bell around.
Hot tip.....Work with one axis at a time for simplicity. Shoot for +/- .005 tolerance.
 

·
More for Less Racer
Joined
·
20,450 Posts
If those numbers are the hole in the bell is off over towards the 3 o'clock position by.028", and up towards the 12 o'clock position by .008" then that would indicate the furthest off centerline position is somewhere over around the 2:30 o'clock position, just short of 3 o'clock.
if you were to take sweep reading with the indicator over along that section you would find the point of furthest out on the arc relative to the true centerline, and will probably be something like .030" and located about 1/2" up the arc from the 3 position.
Somewhere between .014" and .016" offset.

Tell me see if I understand these numbers correctly. The bell needs to come down 0.004 in AND it needs to move left 0.014 in. Is this correct? In other words, here is what I am dealing with, right?
Yep, you have it right. Bell needs to move in the direction directly opposite the hole offset.
This is where you use the dowels adjustability, to clock each dowel so the offset moves in the direction you need.
i would do the sweep test to find the max runout point so that you can visually see what angle you are going to need to clock your dowels.

You likely can't get an exact match for the offset you need, but I know .014" is a commonly available number, clocked right it should get you within .002"
 

·
Registered
Joined
·
111 Posts
Discussion Starter · #5 ·
If those numbers are the hole in the bell is off over towards the 3 o'clock position by.028", and up towards the 12 o'clock position by .008" then that would indicate the furthest off centerline position is somewhere over around the 2:30 o'clock position, just short of 3 o'clock.
if you were to take sweep reading with the indicator over along that section you would find the point of furthest out on the arc relative to the true centerline, and will probably be something like .030" and located about 1/2" up the arc from the 3 position.
Somewhere between .014" and .016" offset.


Yep, you have it right. Bell needs to move in the direction directly opposite the hole offset.
This is where you use the dowels adjustability, to clock each dowel so the offset moves in the direction you need.
i would do the sweep test to find the max runout point so that you can visually see what angle you are going to need to clock your dowels.

You likely can't get an exact match for the offset you need, but I know .014" is a commonly available number, clocked right it should get you within .002"
Yeah I got the Moroso 37936 0.014" Bellhousing Dowel Pins. Going to be interesting because apparently these pins don't have the offset spot marked. So I guess I will just need to put them in and then start rotating them and checking the runout to see where they need to settle.
 

·
More for Less Racer
Joined
·
20,450 Posts
Using your dial indicator roll the offset pin in a v-block or drill press and find the peak offset side. Then mark it.

You have to both of them pretty close to the same or the bellhousing won't go over then, because their centerline distance won't be the same if they aren't both angled the same.
 

·
Registered
Joined
·
341 Posts
Don't remember who said what in the above comments, but some have it backwards and some is over complicating this very simple process.

Do it this way !
First, measure side to side inside the housing hole and write down the measurement.
Second , measure vertically inside the housing hole and write down the measurement.

Then measure the diameter of the input shaft retainer plate. This is the trans part that sticks into the bellhousing and will do the aligning. Measure it both horizontally and vertically and write it down. It will most likely be the same both ways.

Now compare the vertical readings to each other.
Also compare the horizontal readings.

This will tell you how much clearance you have in each direction. The ONLY PURPOSE of doing this is to check and see if the bore in the bellhousing is round and how much clearance you have BEFORE you even attempt to align the bellhousing.

The above should be considered useful knowledge so that you know how much the trans can move even if the bellhousing was perfectly centered with the crank.
.................................................................................................................................................................................................
NOW FOR ALIGNING THE HOUSING........................... WHICH IS WHAT YOU WANT TO KNOW
.................................................................................................................................................................................................

I CAN"T SAY THIS STRONGLY ENOUGH, SO ITS UP TO YOU TO DO IT OR NOT

WHEN YOU CENTER A PART, YOU DO NOT WORRY ABOUT ALL FOUR DIRECTIONS AT THE SAME TIME !

You can notice what the dial indicator reads at each 90 degree location, but you want to work on ONLY TWO readings 180 degrees apart. In machinst terms this is called the "X" axis and the "Y" axis.
Font Parallel Circle Design Symmetry


So basically your initial measurement showed you were actually off .004 on each axis, which sounds very good.
Now , check your measurements a second time. The reason for this is that sometimes you (me,everyone) does not get an accurate reading and its best to double check yourself. This can be caused by the set up not being rigid enough and flexing slightly, a small piece of dirt somewhere, or the friction of the indicator against the housing as you rotate it.
Keep double checking till you get a consistant reading. We all have brain farts from time to time.

Once you have readings you believe are correct, insert your eccentric dowel pins and rotate them to try to get the best condition. In your particular case, if the .004 is correct, you do not need the eccentric pins as you are within .004 of the centerline and thats very good.

The reason you work one axis at a time is that if you make changes, they usually affect both axis and then the changes get confusing. When a machinist centers a part under a spindle on a milling machine, they do this same thing. They check the "X" axis and then turn the handle to move the table the part is clamped to. When they get it centered on the "X" axis, they stop. Then they check the "Y" axis and crank another handle till the indicator has the same readings on the "Y" axis. The "X" may read .002 in both X directions. The indicator may read .002 in both directions on the Y axis as well...........BUT it could read .003 on the Y axis. The hole is not round. The important thing is that on each individual axis......YOU get the same reading on THAT axis.
 

·
Registered
Joined
·
111 Posts
Discussion Starter · #8 ·
Don't remember who said what in the above comments, but some have it backwards and some is over complicating this very simple process.

Do it this way !
First, measure side to side inside the housing hole and write down the measurement.
Second , measure vertically inside the housing hole and write down the measurement.

Then measure the diameter of the input shaft retainer plate. This is the trans part that sticks into the bellhousing and will do the aligning. Measure it both horizontally and vertically and write it down. It will most likely be the same both ways.

Now compare the vertical readings to each other.
Also compare the horizontal readings.

This will tell you how much clearance you have in each direction. The ONLY PURPOSE of doing this is to check and see if the bore in the bellhousing is round and how much clearance you have BEFORE you even attempt to align the bellhousing.

The above should be considered useful knowledge so that you know how much the trans can move even if the bellhousing was perfectly centered with the crank.
.................................................................................................................................................................................................
NOW FOR ALIGNING THE HOUSING........................... WHICH IS WHAT YOU WANT TO KNOW
.................................................................................................................................................................................................

I CAN"T SAY THIS STRONGLY ENOUGH, SO ITS UP TO YOU TO DO IT OR NOT

WHEN YOU CENTER A PART, YOU DO NOT WORRY ABOUT ALL FOUR DIRECTIONS AT THE SAME TIME !

You can notice what the dial indicator reads at each 90 degree location, but you want to work on ONLY TWO readings 180 degrees apart. In machinst terms this is called the "X" axis and the "Y" axis.
View attachment 619508

So basically your initial measurement showed you were actually off .004 on each axis, which sounds very good.
Now , check your measurements a second time. The reason for this is that sometimes you (me,everyone) does not get an accurate reading and its best to double check yourself. This can be caused by the set up not being rigid enough and flexing slightly, a small piece of dirt somewhere, or the friction of the indicator against the housing as you rotate it.
Keep double checking till you get a consistant reading. We all have brain farts from time to time.

Once you have readings you believe are correct, insert your eccentric dowel pins and rotate them to try to get the best condition. In your particular case, if the .004 is correct, you do not need the eccentric pins as you are within .004 of the centerline and thats very good.

The reason you work one axis at a time is that if you make changes, they usually affect both axis and then the changes get confusing. When a machinist centers a part under a spindle on a milling machine, they do this same thing. They check the "X" axis and then turn the handle to move the table the part is clamped to. When they get it centered on the "X" axis, they stop. Then they check the "Y" axis and crank another handle till the indicator has the same readings on the "Y" axis. The "X" may read .002 in both X directions. The indicator may read .002 in both directions on the Y axis as well...........BUT it could read .003 on the Y axis. The hole is not round. The important thing is that on each individual axis......YOU get the same reading on THAT axis.
Well wait! If the 3 oclock is +0.018 and the 9 oclock is -0.010, the actual travel is 0.028 no? +0.018 -0.028 yields -0.010. So the offset should actually be 0.028 the way I understand it. And I need is to move it towards the center 0.014 in. That is also confirmed by zeroing the dial indicator at 9 oclock and then rotating it 180 degrees to 3 oclock, which gives me 0.028. I even tried zeroing it out at 3 oclock and rotating it to 9 - I get -0.072. Again the travel is 0.028 in. I am not sure I am understanding your math here.
 

·
More for Less Racer
Joined
·
20,450 Posts
Well wait! If the 3 oclock is +0.018 and the 9 oclock is -0.010, the actual travel is 0.028 no? +0.018 -0.028 yields -0.010. So the offset should actually be 0.028 the way I understand it. And I need is to move it towards the center 0.014 in. That is also confirmed by zeroing the dial indicator at 9 oclock and then rotating it 180 degrees to 3 oclock, which gives me 0.028. I even tried zeroing it out at 3 oclock and rotating it to 9 - I get -0.072. Again the travel is 0.028 in. I am not sure I am understanding your math here.
You have it right.

Ekimneirbo's X-axis math is incorrect, the indicator is reading total sweep of the needle... the negative symbol in front of the number has to be ignored......and you've proven that fact by zeroing your indicator on the 9 o'clock position and ending up with the full +.028" reading at 3 o'clock.

Your offset is .014" and .004" for the respective axis.

Have you re-indicated to find the exact high and low direction/angle for the offset yet?
 

·
Registered
Joined
·
111 Posts
Discussion Starter · #10 ·
After about 4 hours of twisting the dowels, grinding, cranking the engine, taking down and removing the bell, I got the dial to read 4 zeros (+/- 0.0005). Unless I am crazy, this means it is perfectly centered. I am going to bed now and will check it again in the morning to make sure I didn't dream this up. Thanks for all the help on this one.
 

·
Registered
Joined
·
341 Posts
Well wait! If the 3 oclock is +0.018 and the 9 oclock is -0.010, the actual travel is 0.028 no? +0.018 -0.028 yields -0.010. So the offset should actually be 0.028 the way I understand it. And I need is to move it towards the center 0.014 in. That is also confirmed by zeroing the dial indicator at 9 oclock and then rotating it 180 degrees to 3 oclock, which gives me 0.028. I even tried zeroing it out at 3 oclock and rotating it to 9 - I get -0.072. Again the travel is 0.028 in. I am not sure I am understanding your math here.
Your experience with the dowel pin offset is the same as the bellhousing . However much something is off in total, you only have to move 1/2 that distance to be centered.............because both sides of the hole move. If you move .007, your total change is .014.

Rectangle Slope Font Parallel Triangle


I admire your dedication and perseverence to trying to get it perfect. You should keep asking questions when you don't understand. When it does sink in, you will see how simple it really is.
 
1 - 11 of 11 Posts
Top