5/7/2006
The Teres Bearing and wear.
After the platter upgrade in December, 2004, I spun vinyl regularly until this day when I disassembled the Teres bearing to examine for signs of wear. Below is what I've found.
1) 
click on thumbnail to view full size image.
photo 1: After removing the platter from the bearing.
2) 
photo 2: After pulling the bearing out of the housing. Ball did not come with the shaft. Tweezers were used to extract the ball. Then the thrust pad was extracted using a small piece of foam tape and the tweezers. The interior of the bearing housing was cleaned of excess lube prior to the extraction to aid adhesion of the tape to the thrust pad. Note that the lube in use was supplied by Teres. It appears to be a type of automatic transmission fluid. Perhaps Marvel Air-Tool Oil...? Red runny stuff.
3) 
photo 3) Thrust pad and Ball. Red arrows point to visible wear areas on either part. Using a 0-1 inch Starrett No. 436 micrometer, I measured the ball diameter away from the wear spot at .3748 inches. Measuring directly over the wear spot the reading was .3723. That is .0025 inches of wear at the thrust. Keeping in mind that this assembly, pad and ball, is supporting a 27 lb platter spinning mainly at 33 -1/3rd rpm. Once in a rare blue moon, a 45 rpm disc is spun.
4) 
photo 4: Close-up detail of the wear to the ball. As noted above, the flat spot is worn .0025 inches. 180 degrees away from this flat spot is another visual wear area but wear is much less apparent. My take on this is that the large flat spot must have registered while thrusting against the spinning bearing shaft and the smaller opposite end spun in the Delrin. Evidence that the bearing shaft spins yet also slips against the brass ball and the brass ball spins in the Delrin pad. Perhaps, if this wearing action were allowed to increase, there might be something akin to lock-up with shaft and ball; with only the ball spinning against the Delrin pad. What would be the sonic affects of that....?
5) 
photo 5: Close-up detail of the bearing shaft. Using the same micrometer that was used to record its dimensions when new I have taken readings again.
When new the bearing shaft showed .0007 inches of taper, top to bottom, in its diameter measurement. The lower register read slightly smaller in diameter than the upper register. The lower register read .7487 when new. The upper register read .7495 inches in diameter when new. Now, after 18 months of continuous use the readings are unchanged at the lower register but the upper register reads.7492 to .7490 inches in diameter. This is consistent with the visible wear pattern and indicates that the upper portion of the bearing sees more wear than does the lower. This is normal for a belt driven turntable.
It isn't practical at this time to record the bearing housing diameters so the precise amount of wear to the bearing registers within the bronze housing isn't known. When new I did manage to gage the diameters of this part using an "air gage". At that time the bushing diameters recorded were .7499 upper and .7497 lower. Giving a running clearance between shaft and bushing walls at .0005 to .00025 inches. These are very tight clearances yet the bearing turned freely. Now the running clearances will have increased. We know this because the shaft measures .0003 to .0005 inches smaller in diameter at the upper register.
6) 
Looking at the counter-bored pocket the bearing ball fits into.....and looking
at the machined face that takes the vertical thrust against the ball.
The brass ball was discarded in favor of a tungsten carbide ball, which should have much better wear life and should also take the vertical thrust without the possibility of deformation. Hopefully, it won't decide to shatter into a million shards of...well, carbide. The bronze/delrin thrust pad was retained and the bearing was re-assembled using Marvel Air Tool oil as lubricant. The Teres owners guide was consulted for the assembly procedure.
Summary and comments:
I should have taken the time to make a Rockwell hardness test on the SS bearing shaft. This shaft is easily marked by the bronze bearing housing during assembly / disassembly. (see photo #5) This indicates that the material likely hasn't been hardened as part of the manufacturing process. More likely, SS round bar is cnc turned directly into finish configuration and the fine surface finish at the bearing registers is applied by hand, holding emery and or crokus cloth against the shaft while the lathe chuck spins the part. That is what finishes off the SS shaft. Then again, Teres doesn't charge very much for this bearing assembly. At this time they only take $295.00 for a Teres Bearing. Regular maintenance should help to preserve this bearing for many years and, as long as the Teres guys are around, replacement bearings can be had. On the negative, lack of maintenance could result in the need for a new bearing every several years or so.
How about upgrading the process...? Apart from raising the cost, what if:
the shafts were cnc-turned leaving a slight amount of excess material at the bearing registers....then the shafts ( in a lot quantity) were sent offsite to a heat treater for hardening. Upon return from the heat treater the shafts would have their bearing OD's ground on an OD grinder to a very close tolerance, adding more precision and part-to-part consistency to the very slight taper they install. Instead of seeing between .0002 and .0005 inches of wall to wall running clearance we could tighten that up a couple of tenths and see essentially the same running clearance top and bottom.
Also, the bearing ball pocket (in the shaft bottom) could use a bit more attention. How about a precision ground thrust area, (hardened), to handle the vertical loading against the bearing ball...! This should improve the wear life of the bearing. What affect would this have on the sound....? Conjecture suggests that ultra-hard, finely finished thrust pads above and below the hardened bearing ball should result in lower drag and tend to chatter less. I'd think we'd have a quieter running -- longer wearing bearing.