It has been suggested that I share my recent experience of rebuilding
a turbine pump that I got from Terry Smythe. I understand that this
type of pump is found in Estey "Welte Original Player" instruments made
after WW1 by Welte-Mignon, Inc. This is an excellent pump but, like
most old player parts, is likely to be in need of more than just
routine maintenance by now, even if properly maintained.
The tag on the wooden housing says "Motor Player Corporation". It is
a five-stage stacked centrifugal blower with curved vanes much like a
modern vacuum cleaner blower but it uses heavy cast fans and does not
have the complex housing volutes or whatever they are. It is much less
efficient than a modern pump but much quieter.
The following addresses the installation of new bearings. Of course,
every other part should be cleaned and checked for wear. And be sure
to mark all parts for orientation and document everything you take
apart. There are too many possible ways to assemble it to leave it to
chance. I do not know if the entire unit was dynamically balanced or
if only the individual fans were balanced but it is best to keep them
and their outer housing sections in the original order.
The original bearings were two part and were intended to be lubricated
with oil on an occasional basis. Since the pump runs at a fairly high
speed and since there is no filter to keep crud out of the fan blades,
it is unlikely that any of these pumps could survive without some
damage to the bearings. They are very carefully balanced and heavy
enough to withstand imbalance caused by dirt but the bearing will
suffer and minor damage to high speed bearings becomes major damage
very quickly. After that, damage to the fans will result if the pump
is allowed to run and grind itself to bits.
The original bearings are not available from any source I could locate
so I decided to fit standard ones. The bearing on the commutator side
is less critical than the other bearing since it does not take as much
load from fan imbalance. It is also a standard size and can be
replaced by a number 628 high speed bearing.
I used a stainless steel shielded bearing as its speed rating was
higher than the standard sealed one. I doubt that sealing is of much
use as there is little air over the bearing if any and the shield is
adequate to keep out large crud. The original felt seals can be left
in place if you can work a new spacer/sealing bushing into the retrofit
but I did not bother with the rear bearing.
The fan end bearing seal is easier to leave intact as the sealing
surface is a pressed on bushing which can stay in place. However, the
original bearing size is no longer available and a change in ID or OD
is necessary. I chose to change the OD because it seemed easier and
less risky to use a sleeve on the bearing OD rather than to try to get
a loose press fit on the smaller ID, with my less than perfect machine
All that was necessary was to bore the bearing housing to a an even
diameter and to assure alignment with the axis of the motor shaft very
accurately as little misalignment of bearings should be allowed
especially with the fancy one I chose to use.
It was certainly scary cutting into the old metal with the milling
machine and I made sure that the hold down were secure and proceeded
very slowly. Once the reference to the original bearing location is
lost, it is not easy to find it again without some rather elaborate
alignment techniques. So use an edge locator (or a wiggler, if you
like them) before doing any cutting.
If the original bore is perfect you can just make a sleeve but it will
need to be rather thin walled and I preferred to make a thicker one.
Then it is just a matter of making the sleeve and slipping it all
together, almost. You also have to fit shims and a replacement spacer
to make sure that the fans are held tightly together and in the correct
position in the stationary castings and also not pre-load the bearings
Make sure that the bearing are well seated before you tighten up the
locking nuts on either end or you may damage them or find that the
clearances change when the motor operates. Tighten the nuts very
carefully and check for free rotation continuously until they are
seated. If any binding occurs, stop immediately and determine where
the shimming needs to be changed.
The rotating assembly is held in place while tightening by inserting a
brass drift into one of the plugged holes (after removing the plug of
course) in the outer housing and preventing rotation of the fans. Make
sure to notice that one of the shaft ends is left hand threaded to
encourage tightening (as opposed to loosening) while the motor rotates.
My pump had double locking nuts on the right hand threaded end and
a single left handed one on the other end. I did not use Loctite on
the nuts because they are brass and too easily damaged by careless
loosening in the future. I did lap them on 600 grit paper to make
sure there were no high spots which would prevent even tightening.
If you can find them, it is a good idea to have shaft and housing shims
available in a variety of sizes before you begin. Due to my remote
location and inability to find a mail order source for housing shims,
I simply made all adjustments with shaft spacers and shims. If you do
this you will have to be very careful when tightening the screws that
hold the fan side housing to the rest of the pump and work progressively
to find the right spacing. Otherwise you can just space the bearings
on the shaft so as to lock the fans together, leaving float in the
bearings and adjust this out with the housing shims last.
When you have the bearings adjusted properly, there should be virtually
no end float of the fan/armature assembly and there should be no
scraping of any moving parts at low or high speed. The fan armature
assembly should turn freely with the lightest touch and a very low
voltage applied for testing should cause it to rotate smoothly. I use
a Variac variable transformer to test at low speed before letting it
run at high speed, possibly causing damage. If there is binding but
no scraping sound it is most likely that the bearing are too heavily
pre-loaded or misaligned. If there is scraping it is more likely to
be that the fans are not centered in the stationary housing or they
Since the same nut is used on the fan end to secure and locate the
bearing, as well as to hold the fans together, it is essential to have
the spacers right. Before starting the pump, feel for wobble in the
fan that is last on the stack and do not start the motor if there is
any. This means that the shaft spacer between the fan and the bearing
is too short and the locking nut is binding on the shaft collar before
locking the fans. If you are not using housing shims you will have to
fiddle with the shaft spacers on both ends to get the fans to pack
properly and to align correctly too.
Do not run the motor at high speed if there is any problem at low
speed. This is not a modern type of centrifugal pump with light metal
fans that will balance itself at higher speeds. It will only get worse
and worse in time if there is any misalignment or imbalance.
The bearing I chose to use for the fan side is a number 629 high speed
ceramic bearing. It was very expensive (almost $40) compared to a
lesser bearing but I hope it was the best choice. It required a sleeve
since the OD is 2 mm less than the original bearing. It is not
recommended to use bearings of different width as problems will arise
in fitting the locking nuts.
Anyone who has read this far might have some interest in these pumps
so here is a question. What would be considered to be a safe working
temperature for these motors, and a safe duty cycle? They have double
cotton insulated armature and field windings. I will be looking in my
old electrical books to try to find and answer but maybe someone knows
some practical figures to consider.