What's new
What's new

FP2 Spindle Assembly (lots of photos)

AlfaGTA

Diamond
Joined
Dec 13, 2002
Location
Benicia California USA
Got everything cleaned up ready to assemble.
In my book this is the single most important part of a spindle overhaul.
You have to get everything spotless. Remove any traces of the old grease and remove any dirt or oil.
My technique goes something like this....
First off everything goes to the solvent tank and i sue a soft brush and solvent to remove all the big chunks.
Then i rinse in a tub of Simple Green. This get any solvent off and also can remove any remaining grease of dirt.
Then i give the part a HOT water rinse and low pressure air dry.
Parts are set in a sheet pan laid out on a lint free shop towel (the blue paper industrial ones made by Scott i think)
I then cover the parts with another towel and set aside.

Here is a close up of one of the threst bearings. These are precision bearings and the type used here has only one ground concave ball track. The opposite race is flat.
This is done to prevent the thrust bearing form being a factor in the spindles rotation...the intent here is to allow all axial location to be dome by the needle bearings.

FP2Spindle30.JPG


Flat race with contact mark on surface...keep the thrust washer facing the way it was on dis assembly.

FP2Spindle31.JPG


Shot of the thrust balls with retainer. The retainer is made with an undercut that is a trap for old grease and dirt...be sure to clean every nook and cranny....
Balls should look bright as polished chrome....Races shouls like wise be bright and free from any pits or surface blemishes in the running areas.
FP2Spindle33.JPG


FP2Spindle32.JPG


I use an air engraver to mark the outside edges of the races and ball retainers so i can keep track of where everything came from. I keep a note sheet showing where the numbers fit in the assembly.
On the needle roller cages i mark them as to "TOP" the way the cage was fitted on the spindle...don't want to mix anything up.

I have rough cleaned all the parts, the spindle and the quill both inside and out....
Now to get serious.
Set aside a clean working area. A place free from drafts and open windows fans or air conditioning...you want still air a clean well lighted area away from chips and distractions.
I like to put down a clean sheet of kraft paper on the bench top, layout the cooking sheet with the parts and have the tools i will need on the bench.
The first assembly step for me is to check the spindle for wear.
I use aresol can "Carburetor" cleaner to wash every part just before assembly, followed by "Brake" cleaner spray and a dry low pressure air or lens dust spray.....
The reason for the two sprays is that i believe the carb cleaner cleans best, and has the best spray off power....but tends to eave a film on the parts...the Brake Clean seems to dry off clean....
Note: all spray off of parts is done well away form the work area...
I assemble the spindle DRY first..no lube on anything. I use dry industrial paper towels to wipe both the inner and outer races of spindle and quill.
The object here is to test the size and wear on the needle rollers, and for that things must be dry.

Here is the spindle with both needle roller cages fitted along with the lower thrust bearing. The thrust bearings are needed ot properly locate the spindle relative to the quill so that the rollers will be tested where they will run.

FP2Spindle34.JPG


The quill is then just dropped over the spindle , then the upper thrust bearing installed along with the spacer washer and top adjusting nut. The nut is tightened up to leave everything just loose with some end play between the quill and spindle.

More to follow...
Cheers Ross
 
Shot of the quill and dry spindle ready for a test assembly.......

FP2Spindle35.JPG


FP2Spindle36.JPG


Assembled quill with all needles and thrust bearings fitted...this style spindle is a bit different from some i have seen. some have both thrust bearings above the needle bearings. This one seems perhaps a bit better as it gives a wider spacing between the needle rollers cages and might result in a stiffer package...
The process again is to lower the quill over the needle rollers and lower thrust baring..it jsut slips together with very little force if everything is correct.
Then you load in the upper thrust bearing being careful to fit everything in the correct order. (the thrust bearings are directional)

With the thrust (upper ) in place the spacer wahser is fitted and then the top adjusting nut. Leave the adjusting nut just loose, so that you can feel end float in the spindle.

FP2Spindle37.JPG


FP2Spindle38.JPG


Once the spindle is assembled the next step is to measure the needle clearance.
This is a dynamic test...ie you are looking for the relative amount of movement between the spindle nose and the quill.

I use my 6" kurt vise to gently hold the quill...almost zero force here.
Easy to squeeze the quill and make the readings not meaningful.
I put a tenths reading indicator against the quill on the moving jaw side and snug until i just touch the quill without seeing any indicator movement....
Then i move the indicator to the spindle nose and using hand pressure push the spindle from side to side....

FP2Spindle39.JPG


The spec given to me is that the movement at the nose should be around .010-.015 mm (.0003-.0005")

The test is done at different rotational spots on the spindle. Any change of reading would indicate a spindle worn out of round.
Fiurther teh test is done changing the position of the quill...(rotate) so as to sheck if it is oval.....
This assembly measured in at just under .0003" of play ..very nice!

More to follow........
Cheers Ross
 
Additional poto of test on spindle play with bry bearings installed.
A better setup would have been to hse the vertical head housing to hold the quill and i have done this in the past. Here however because of other work going on with the machine i did not have the head housing available. It was fitted to the machine at home.

FP2Spindle41.JPG


Once the clearance/wear test is done if any size correction were to be needed this is the time. Wear and excessive play can be corrected by fitting larger needle rollers. There were markings on these cages that indicated the relative size of the rollers...Deckel tech's have sheets telling the size of the rollers based on these marks so making the correct change can be done.
Believe Don Sentner has a supply of needle rollers and cages in different sizes.
If a roller change is made a re test of the clearance as above should be done...do not make the fit too tight it will cause trouble and potential failure.

After testing it is time to do a final assembly. The spindle is taken apart again and cleaned again, and grease applied to the roller cages/rollers and the thrust balls/cages.

FP2Spindle42.JPG


FP2Spindle43.JPG


FP2Spindle44.JPG


Look at the above photos.. note how little grease there is applied to the cages and rollers/balls. Too much lube here is a bad thing. Excess grease will cause the rollers to skid and overheat in short order. A little of the correct grease goes a long ways.

FP2Spindle45.JPG


I used the Isoflex "Supertel" on this spindle. If it had been toward the upper end of the wear range i would have opted to use a heavier grease...."Isoflex NBU 15"


Once the bearings are greased, time to final fit the quill.

More follows........
Cheers Ross
 
On the last part now....
Assemble the quill to the spindle just as when doing teh "dry" assembly: Fit the lower needle cage, lower thrust bearing then upper needle cage unit. Lower the quill on to the spindle and fit the upper thrust bearing, spacer and adjusting nut.

Run the nut up until almost tight.
I snug the clamp screws just a bit to make the nut slightly stiff to turn.

Clamp the end of the spindle in the vise or other holding fixture and setup your indicator on the end of the quill.

FP2Spindle46.JPG


FP2Spindle47.JPG


The object here is to adjust the nut to have .005mm (.0002")of end float between the quill and spindle.
This takes a little work as the bearings now have grease . Pushing and then pulling on the quill will show the movement but be careful to allow time to settle as you push...I push in one direction then zero the indicator then pull the opposite direction. At each end of stroke i give a little back and forth rotation to make sure the balls are full down in the grease.
Adjust to give .0002" end float...DO NOT OVER TIGHTEN .....these thrust bearings need clearance to live.
Tighten the locking screws and re test to make sure nothing has changed.

The spindle is now ready for install and break in.
You should rotate the spindle by hand and check for ease adn freedom of movement...
ANY glitch or rough feeling on hand turning is a problem and needs to be corrected.
Rough spots in rotation are almost always the result of dirt in the package and you should immediately disassemble and clean everything re apply new grease and re assemble.......

Re fitting to the machine is the reverse of removal...be sure to lube the outside of the quill when applying to the head casting......

Cheers Ross
 
Last edited:
Wow. Two thou end fload on the spindle?

:)

Obviously a (corrected) decimal point error.

For bearings like this, I use a solvent like WD40 for the initial clean as it
removes grease and oil well. Short bursts of ultrasonic in a cleaner helps
to break it up.

Final assembly I like to use pure ethanol as the cleaner.

It's tough getting good numbers with a DTI like that, you need to be very
fiddly getting it right. I have access to an electronic indicator setup, and
it is very very handy.

Excellent photos and a challange to me if I ever decide to do the spindle in
my machine. Thank you for taking the time to document 'the right way'
to do this stuff!

Jim
 
I've gone through the vertical quill in my FP2NC once and the horizontal quill a few times (remember the squeal issue?). When I check clearance, I put the mag base on the quill housing and grab the spindle spline in a vise with copper jaws when using the 10ths indicator. I apply hand force to the the quill housing. Also, I do it with lube in place. I am unwilling to put clean steel surfaces together without any lube, as I think there is too much chance of micro-welding or galling of the surfaces. I don't think the oil films are so strong that hand force cannot push through them. YMMV.

I find that sort of work nerve-wracking. It's probably how a brain surgeon feels.
 
Rich:
Interesting point on the micro welding....
All the above method was taught to me by a German professional Deckel tech...Volker Spitz. Perhaps he used "dry" on the bearings because he was changing size of the rollers and did not wish to grease each trial assembly before he found the correct sized fit..anyhow it would be good test to check the same spindle both ways to see if any difference showed up on testing.....
I have the Horizontal to do yet, perhaps i will check that one both ways.

Jim:
yes the indicator work is fiddly and you have to take lots of readings and hold your mouth just right.
Not living in a very "techy" world here doing old cars, can you explain how an electronic gauge makes this less fiddly to get a size?
Do i need to make an investment?
Corrected the places on the clearance..wrote it wrong first then correct later in the same post....Saw teh correct number and thought i was OK.....OH well it is correct now.
Cheers Ross
 
Less fiddly, because you can increase the range of the indicator substantially.
The federal one I've used, will reliably indicate down to a micron. Of course
it can be ranged up as well.

There are two issues, one is that the DTI you are using reads to about 0.0001
inch, or maybe half of that if it's in good shape. You're really trying to set things
like the end float to two tenths, so that's basically a one part in four. So you
are fighting that somewhat.

Other problem is that DTIs like that have a cosine error, so they don't work well
as absolute measuring devices. Of course you took pains to have the lever at
90 degrees to the axis of motion but it's still a fight.

The electronic indicators often have a sort of oval-shaped tip that is *supposed*
to provide some correction for the cosine error problem.

But I recall setting up some bearings to replace those in an M head, and I also
washed them out before trying the job, and was using that Federal indicator
to get the clearances down below 0.0001 inch or so. I don't think I could be
good enough to do that with a mechanical indicator.
 
nice write-up, although the eyes tend to wander off to the background with all the listas you have :D
 
Ah, the background action lista-porn. This is a sign of the
truly deviant mind! :) :)
Tsk, tsk...such a hodge podge of Lista's however...old style blue (the ones with the handles that aren't flush with the outside), new style blue, new style green...and no labels on any of the handles ! how does he stand it ? ;)
 
I work with this style of roller bearing in other applications.

A few observations I picked up, some from bad experience, that led to the dreaded "mystery bearing failure".
Mystery bearing failure is, when the bearings and cages end up a gobbled up mess, and its not possible to determine what caused it.

Some things to look out for.

Never assume, the bearing is OK just because its new.
Snap the rollers out of the cage, gage each one, and examine under 15X.
After replacing the bearings back into the cage, make sure each roller rotates freely in its slot. Any friction there can cause a roller to skate, skating is the worst, next to having a roller or two that are oversize and or a flawed roller.

In the other applications I do, grease is not recommended at all, that makes this application even more tricky, as grease will tend to cause skating.
A careful bench run in, at slow speed in both directions would seem to be very important in this case.
These bearings tend to rumble for a little time when first started, as they find their home. always seems to be best for a bench run in before use.
 
At work, I deliberately mis-lable all my lista draws when they get labels at all. That
way folks who come in trying to raid one of my labs, get flummoxed. Better yet
is the characteristic slam-slam-slam-slam of the drawers being ransacked.

This brings me like a shot out of my office, to find the offending pillager.
 
As to the cabinet labeling, Jim is dead on the money...done on purpose to regulate the "Dork" factor....
I can tell no matter where i am standing in the shop if someone is rooting about in the tooling and parts cabinets....the drawer sounds are my cue to check out who is messing with my stuff, and what they are looking for.

Yep cabinet a collection or old and newer/Green and Blue...still open and close the same, hold all my stuff in some sort of order....
Been bought piece meal over some time as the place grew....Guess i never thought too much about the need to have them all the same, i was just happy to have any.
Cheers Ross
 
I've been checking the vertical spindle of a Deckel FP3 lately, and thought I could add several useful infos to this thread.

The spindle was suffering from excessive runout when I got the machine. Having no indications about the machine past life, I started to search for the cause of the problem.
The curious thing is that the spindle seemed to have radial play in some directions, while it was rock solid in others.
I was afraid for a while, that it could be bent. After all, who knows ?



I decided to check its straightness.
First problem to solve was to find a way to hold the part without adding any external source of errors such as centers.
For that reason I decided that vees were the solution.
But the rear end of the spindle has two slots that would have disturbed the measure on a vee. So I grabbed the first steel plate I saw, and bored it very precisely to the diameter of the spindle rear end. That way, I could spin the spindle on a full revolution without caring about the slots, and see what was going on.



To my great surprise, all the readings were perfect. Not the slightest run out at both ends nor in the middle of the spindle, even with a Mahr millimess indicator... Hum !

I started to look further and checked every part of the spindle assembly again. Of course, I had already done it when disassembling the spindle and cleaning everything, but the spindle beeing straight, there was necessarily a problem somewhere.

The thrust bearings were perfect looking, and so did the cups and washers.

Only the Spieth lock nut was showing marks of slightly uneven wear.





I decided to investigate in that area, and grabbed a micrometer to check the thickness of the cup located at the top of the assembly...
Gosh ! There was more than 2/100 mm of error on a full rev. Incredible for a part that showed superb marks of fine grinding !







I surfaced both faces of the washer part in one sep up to guarantee perfect flatness.



Checking the lock nut was a bit trickyer, as the locking feature of the design rely on the flex and compression of the part.
I decided the best way to be sure was to measure the part in situ, locked on the spindle itself.

To do so, I put the spindle in a lathe, mounted the locknut, tightened it as evenly as possible, and started to indicate the face of the nut.
And there was some significant error here too !

I surfaced the locknut in place to get rid of the problem.





It was clear then, that the defective cup and locknut were forcing the whole spindle in a kinda pendulum position in the quill housing, taking up all the radial play in one direction. Hence the run out of the spindle and the differences of behaviour when applying efforts in different directions.

The marks I had noticed at the upper part of the spindle confirmed that theory.



After that, it turns out the final checking of the re-assembled spindle showed excessive radial play yet (but the run out part of the problem was cured) so I had to send it to Herr Singer to have new needle bearings fitted anyway. Interesting to note that in my quest for a solution, I asked FPS if they would be willing to sell selected needles, to no avail. Their process implies that one send the the WHOLE head (that is, including the ram) so that they can inspect it and make a repair quote...
From the pictures they sent me, I'd say that the wooden box in wich they ship back the head once overhauled, is worth more than what I paid for my machine :nutter:



Nevertheless, I thought it may be interesting for others to know that :

One must not trust factory parts, even perfect looking ones
Runout problems can have surprising causes and can sometimes be solved (at least when they don't add up to excessive play).
 
Last edited:
Over the past few years there have been some great pics of Deckel related projects and repairs. The pics of the older posts are no longer linked.
Is there a way to link to the pics? Both the photography and the content are excellent and I would like to be able to save them.
Thanks,
Rob Daniel
[email protected]
 
Good work Tien!
And I see that you still use the 125 beside the 150 lathe.

A Deckel is a well build machine, but a lot of parts that have a pretty critical function (like the one's here T has to overhaul ) are made of simple soft steel.
 
A Deckel is a well build machine, but a lot of parts that have a pretty critical function (like the one's here T has to overhaul ) are made of simple soft steel.

The parts I had to machine were obviously not dammaged but rather out of tolerance from the start (read : from the day they were manufactured). They showed absolutely no burrs or dings.
The fact that they were made out of soft steel made the repair easy, but that was clearly not the root of the problem.

I think the bad cup probably gently forced the spindle in a wrong position but the problem was not visible untill some radial play developped, thus allowing measurable runout to appear.

Despite a very clever design, the Spieth locknut doesn't seem a good solution for insuring a dead square face of the nut relatively to the spindle axis in my opinion.


MSR%20Spieth%20Locking%20Nut.jpg


Too much chances of putting the nut slightly out of square when tightening the screws.

I think a good old slitted locknut (with proper tolerances on the threading and perhaps a different slitting pattern from the example below) would have been better. Don't ask me why. That's just the way it "feels" to me.

bague-arret-taraudee-fendue-simple-acier-31_761-pmod1-ganter.jpg
 
You are right, it has al to do with the direction the clamping force is directed to, axial or radial. But when it comes to precision,I think you cannot really relay on thread, because the nominal tolerances are too wide.
 
Nice work T. Thanks for sharing .
Nice photos....

Rob:
As to the old posting and the photos...for my part i used to use "Metal Working .com" they used to provide a dropbox free that hosted the photos.
Well the site has changed and i have no idea where the old pictures went...That is why my stuff at least no longer shows...
too bad really...
Cheers Ross
 








 
Back
Top