Quieting a Noisey Bench Top Lathe. How ??

[Lloyd-ss]

PROGRESS
I am slowly nibbling away at the noise problem and am learning quite a bit from the forum members and my own experimentation.
What I have learned:
I have isolated the spindle bearings and they definitely are noisey and I have replacements on order.
The motor is noisey too. I set up a fractional HP motor to just drive the spindle. It eliminated the motor noise, but not the spindle noise.
The idler stepped pulley has noisey bearings. If I am careful with the belt adjustments, I can minimize that noise. That is a major improvement.
Proper shimming and backlash adjustment in the feed change gears can make a big difference also.

So, progress. After I get the new spindle bearings installed and see what they do, I will attack it again.
Lloyd

edit..... The replacements are on order from Dublin Ireland, and should arrive in less than a week. They are a class P5, one quality class better than the original P6 bearings.

 

[Steamchick]

Hi Lloyd,
On "noisy belts": Some curious things to note:
What many do not realise with belts, is that different designs generate noise a different ways.
I'll comment on different designs - and some of the odd bits of design I recall about them.
Vee-belts, wedge belts, Poly-vee belts In principle these all perform dynamically the same way, more or less. These rely upon the load tension to make the Vee-form wedge itself into the pulley. The initial tension is NOT the driving load tension, and should not be over tensioned as this will cause overloading of bearings, a common fault with re-setting by "un-trained/un-knowing" mechanics. But equally, initial load take-up can generate slippage and unwanted wear of belts when the initial tension is too low. - But rarely a fault seen in service. Most of the wear occurs during service (loaded condition) anyway.
A point to AVOID: Any Vee-configuration belt that is worn to the condition where the "point" (or flat) of the vee shape in the inside contacts a pulley - bottoming in the Vee-groove. This means the belt will SLIP and be noisy because the Vee cannot generate the driving forces that the drive should be transmitting. Vee-belts are the simplest form, Wedge belts just a wider version for higher loads, optimised by the belt drive manufacturers, and poly-Vee belts offer much better cooling (due to their large flattened shape)and a more compact (and durable?) drive.
Link belts are an easy-to-fit "temporary" belt (On hobby stuff "temporary" can be a lifetime!). The links weaken the belt considerably compared to a single loop Vee-belt. and generate a lot of heat - but in the case (comment #39) by Ingnator the damping of noise by friction - link to link - is probably the reason the belt installation is quieter than the previous installation.
Vee-belts do generate noise, caused by the constant engaging and disengaging of the vee-of the belt in the pulley groove. The friction of pulling the belt into the groove - and then back out again - causes a rubbing (like a rosined bow on a violin string) but at super high frequencies that we normally cannot hear. This rubbing is the major (natural) source of heating the belt, The work of bending and straightening the belt is tiny by comparison.
OK. To "toothed" belts, in their various forms. I say various forms, as some teeth are trapezium shaped, some semi=circular and some "hyperbololic" - all in an attempt to reduce local stresses and increase LIFE. But they all generate noise due to the engagement of each tooth displacing the air in the pulley between teeth, and then when the tooth comes out the air rushes in again from each end, causing a shock-wave (tiny, but it exists). This can set-up an audible whistle, related to the frequency of tooth engagement. (rpm x number of teeth, per pulley). But this also can seem quiet at higher speeds if it becomes ultrasonic. These belts rely upon a correctly set tension - so you do not overload the belt and pulley bearings - or under-load the belt so it can pull teeth out of engagement with the pulley (worst case) and the belt slips. Usual failures (as the teeth are the most highly stressed points) are breakages where the tooth to belt leading corner fails and the crack/tear propagates across the thickness of the belt overstressing the tension fibres one after the other to cause them to fail, and the belt snaps. ANY cracks in the polymer and the belt is going to fail SOON. Manufacturers' recommended service (replacement) intervals are based upon thousands of belt tests and failures, so should be adhered to and NOT exceeded. - But on a Hobby lathe this service running time may never be reached, so be aware that the polymers break-down naturally, so will deteriorate and fail with a combination of TIME and Heat. After 15 years, your car cam-drive belt will be likely to fail because it is old enough it will have "died" - or be close to dying - so change it. On your Harley Davidson 2-wheeled tractor, this may be as low as 3-years... The belt's teeth wear on their leading edge, as they rub while engaging and disengaging the pulleys, and the core fibres fatigue and fail after a long life due to the bending and reverse bending around pulleys and tensioner idlers. But the (rubber, etc.) polymer ageing is a major cause of failure in many applications.
Note: All belts load pulley bearings to one side, so are horrible from the perspective of the bearing designer/manufacturer, so likely to be "huge" compared to something like a gearbox bearing that has much less side loading than "belt tension". The pre-tension on all belts must be enough that at full dynamic tension on the driving side, the slack side still has some residual tension to control the belt engagement security. - Unlike a chain that will function reliably with a slack return run.
Hope this is of some interest - tell me if otherwise, or you disagree with my screed? Service experience is often different to "office" experience!
K2

 

[Steamchick]

I have changed many motorcycle crankshaft bearings, electric motor bearings, etc. and am aware that a common comment is that " they designed this with bearings that were of a tighter tolerance than necessary to make the new ones quiet, but they always fail so a looser grade of bearing is better".
There goes the uninitiated and unknowing...
for "rolling element" bearings, whether ball roller, tapered roller, etc. the significant factors for designing an application are COST (finer toleranced parts are more expensive to make), and Temperature, always assuming that the lubrication and cleanliness is adequate for the designed life of the bearing.
Simply: The rolling element (ball, roller, etc.) heats up and keeps cool differently from the outer or core races because these are only in point contact with material to conduct heat away. The inner and outer raves conduct heat to the adjacent housing material very effectively, due to their intimate contact. So a bearing that is very well lubricated by oil keeps mostly close to the oil temperature, and therefore does not need a large "clearance" to work effectively for the range of temperature and frequency of temperature cycling of the application. But a greased bearing does not have the heat transfer to the lubricant to control the rolling element temperature as well, so needs a wider clearance to accommodate warm-up (rapid for the rolling element, slower for the housing and races).
The salesman always stocks "higher use" larger clearance ranges of bearings, as everything can use them, but finer clearance grades of bearings are usually not "off the shelf" so need special ordering and are higher cost, longer delivery, etc. so the salesperson is induced to tell any technical sounding excuse to sell what he has now, rather than take an order for what is best (original design) for the application. Just human nature & "common business practice".
If the application was designed for a finer grade of bearing, this will last longer than a "pre-worn" wider clearance range... Unless in an application (such as "tuned for racing") where bearing temperatures are not as he designer planned.
Incidentally, poor lubrication always shortens bearing life, but rust, from atmospheric moisture and "dry bearing tracks from "dried-out" old plant" is much worse. And all are noisy as a result!
But even "the best" bearings have a finite life, because the steel surface of rolling elements and tracks "fatigues" due the the surface compression every time a rolling element goes past. Eventually, the hard surface breaks away from the sub-surface and microscopic flakes (micro-dust-sized particles) leave the rolling contact surfaces, so clearances increase as the bearing wears. This "hard particle" dust also increases the wear rate as it is ground into other contact zones... So bearings have a finite working life. (At least 1 week beyond the Guarantee expiry?).
I was advised by a rolling bearing manufacturer's engineer (professionally) that using such low contact friction materials as Molybdenum grease, can cause early life failures because "new, tighter" bearings can slide rather than roll due to the low friction properties of the moly... causing adverse uneven wear and early life failures. Lithium grease does not cause those failures, because the rolling elements can "grip" the surfaces and roll instead of sliding. But re=packing bearings with Molybdenum grease as per service intervals will likely increase the surface durability and service life, by reducing friction after bearings have "settled-in"....
K2

 

[Lloyd-ss]

Hi Lloyd,
On "noisy belts": Some curious things to note:
................................................. edit for emphasis by lloyd
Hope this is of some interest - tell me if otherwise, or you disagree with my screed? Service experience is often different to "office" experience!
K2

Steam,
I always appreciate your responses, and often wonder if these just "flow" out of you. I personally know from other forums I have participated in how much time and effort goes into composing and typing such detailed, and coherent, responses. And if testing and graphics are involved the time seems to go up exponentially. But such a response, in my opinion, is always a pleasure to read. And the fact that the writer (you) took the time to personally share knowledge and experience only adds to the intimacy of the sharing. So yes, there is always something to be learned, even if we don't personally acknowledge that fact to the writer. For some, the writing might be prolific and not difficult; but for others, composing a post is a major struggle. But ALL are appreciated and worthy of our attention.
That is as much a personal philosophy about learning as anything else.

----------
The discussion on bearing accuracy (quality class) is interesting. Part of the answer to proper quality class selection is "how much effort are you willing to put forth to actually take advantage of better bearings?"
In this particular lathe situation the picture in post #29 speaks volumes. A big gap for all sorts of debris to directly enter the bearing, and only an occasional squirt thru the ball oiler is provided to flush the bearing out. Why not just throw a handful of sand into the gap? I will be fabricating a seal to cover that gap when I install the new bearings and will also be much more liberal in my "oiling regimen."
------------
As I get older, the new knowledge I obtain seems more and more valuable, and I wish (foolishly, ha ha) that I had another 70 years to utilize it. Too much introspection, LOL.
Lloyd

 

[Ken I]

LLoyd, Ther's a post on spindle bearings - for grinding :-
Grinding spindle.
See my post #41 and #39
For a lathe spindle you don't normally put to angular contacts at either end because of expansion and contraction - it is more normal to placve them in close opposing proximity to the chuck end - with the outboard end supported by a siding fit ball bearing.
My post #41 shows just such an arrangement with a slinger ring and outward facing oil seals to keep coolant and junk out of the bearings rather than keepig the lube in.
Regards, Ken I

 

[Lloyd-ss]

LLoyd, Ther's a post on spindle bearings - for grinding :-
Grinding spindle.
See my post #41 and #39
For a lathe spindle you don't normally put to angular contacts at either end because of expansion and contraction - it is more normal to placve them in close opposing proximity to the chuck end - with the outboard end supported by a siding fit ball bearing.
My post #41 shows just such an arrangement with a slinger ring and outward facing oil seals to keep coolant and junk out of the bearings rather than keepig the lube in.
Regards, Ken I

Ken,
Thanks for the link and the comments. The grinding spindle thread dives pretty deep, but also gives credence to different design approaches. At the price-point of my lathe I guess the pair of tapered roller bearings, spaced about 6" apart was the chosen method. With the type of work that I do on the lathe, operator skill, and not machine accuracy, is usually the limiting factor, ha ha. The debris seal for the front bearing is number one modification priority me right now.
Lloyd

 

[pete]

It's tough to make blanket statements about spindle bearings that apply to all lathe designs. But in general, what you'll find on a lot of industrial lathes is a pair of opposed angular contacts at the front of the spindle, those usually would be a press fit on the spindle bearing journal. They take most of the radial and axial cutting loads. Different makes use different methods for additional thrust bearings or even bronze thrust washers towards the rear in most cases. Then a lot will use a simpler slip fit ball bearing on the rear spindle journal. On those the outer bearing race is usually a press fit in the head stock casting. For the most part that rear bearing is only there for the radial loads and add the much needed spindle support. On the less expensive machines like most of us are dealing with, you may have a single ball, roller or the angular contact bearing that are (or should be) a press fit on the front of the spindle. At the rear there set up as that light slip fit to allow for spindle expansion or contraction due to temperature changes. Once you understand there has to be a method to allow for those slight spindle changes in length, then its easy to at least check the manufacturer did design and grind the bearing journals to retain the spindle in a fixed position at the front and still have that precision slip fit at the rear. Since I don't own the same machine, then there's no way for me to know what exact method the manufacturer used for your bearing arrangement. So I can't be more specific.

Some may take this as China bashing or whatever, its not, I'm trying to be factual about these machines. The lower cost off shore lathes and mills are obviously built to meet a very specific price point and a whole lot of corners will get cut. What the manufacturer states they use for the class of bearings may or may not ever meet the actual numbers there supposed to. And for very hard to verify or test items, lets just say some companies aren't above fudging the real numbers a bit. So those new spindle bearings may be far better than you think compared to what your replacing. Then there's the unavoidable facts of these machines not being assembled in anything approaching clean room conditions. That paint over spray in the bearings as an example. Even my much more costly Taiwan built Bridgeport clone wasn't completely immune to that. While I didn't find a huge amount, there was still enough grinding dust plus a few stray metal chips that I'm thankful I took the time to disassemble it into its major components and solvent wash all the feed screws, nuts and dovetails. Out of the crate it was pretty good, after that detailed cleaning, feed nut and gib readjustment it was literally multiple times better.

Fwiw and I doubt I would have gone to the same effort on a mini lathe. There could be something in this that would help with adding that seal Lloyd. But adding something just to prevent any debris from entering should be fairly simple.

 

[Nikhil Bhale]

Hi,

I have a similiar lathe as yours. A Sieg C4B. I recently renewed all the bearings for this lathe. I also converted the lathe spindle to oil bath lubrication. Its an relatively easy mod to do. For seals I used generic oil seals that I got from Amazon.
You can use these seals to keep the debris out of the bearings.

Regards
Nikhil

 

[Steamchick]

Hi Lloyd, Re #44: Thanks for your kind words. My simple ethos (From my Dad) is that "we pass these roads so very quickly, that it is all we can do to take some small time to share some thoughts with others that we meet on the way. Thereby learning and teaching at the same time".
- Filippin 'eck! I didn't mean it so sound so philosophical!

Really I wanted to share a "thankyou" to you and all the other contributors from whom I have learned a lot. - And hope to do so for many years yet? (Can you shut me up? = Yes, you control the Off Switch!).

K2

 

[Steamchick]

My lathe has an "Interesting" characteristic, one that I expect to lead to mainshaft bearing failure - one day.... when I'll have to change the bearings, so I am seriously trying to learn all the tips here.
What it does...
When cold, and I switch ON at the start of a session, I need to run the lathe at just a couple of hundred rpm for between 1 and 10 minutes (Cold in my Garage may be 15C. / 60F. down to 0C. / 32F.). Before the bearings have warmed, at "Cutting speed" (anything from 300~1500rpm - varies day-by-day) there is an interesting and intermittent "click-click" noise that I assume to be something to do with the bearing Ball-carrier... I hope it is not odd balls "stopping momentarily" or more drastic. But run slowly this does not occur. So I just turn on the lathe at ~200rpm and let it run while I decide what tooling set-up, material, etc. I a going to do. After "warming", all is quiet...
By the time my brain is "in gear" the bearings have warmed (or whatever?) and the lathe runs just fine.
Has anyone any clues as to what this "click" noise really is, and what I should do - if anything?
On Lubrication: The "book" says NOTHING about bearing lubrication, so I assume "sealed for life" (and life can be very short?). I have had the lathe more than 10 years... but rarely do more than a 10 hours a month. Any more advice is welcome.
K2

 

[Lloyd-ss]

Bearing preload

I got the chuck back on last night and was able to do a little work. The first job was working on a 10-32 screw including drilling a .045" hole thru it. All went well.

Then a piece of 1144 stress proof steel and a parting blade on the now-warm machine. Chatter chatter. I touched up the edge on the blade and experimented with tool height. No luck.

I finally figured "bearing preload." I actually ended up tightening the preload by 1/4 turn. The bearing arrangement is 2 tapered roller bearings like a car front axle. That did the trick!
As someone previously suggested, the bearings might need to be tightened to a small percentage of their max axial loading.
And so it seems.

 

[Richard Hed]

My advice, do more cutting. make it 30 hours a moneth.

 

[L98fiero]

Bearing preload

I got the chuck back on last night and was able to do a little work. The first job was working on a 10-32 screw including drilling a .045" hole thru it. All went well.

Then a piece of 1144 stress proof steel and a parting blade on the now-warm machine. Chatter chatter. I touched up the edge on the blade and experimented with tool height. No luck.

I finally figured "bearing preload." I actually ended up tightening the preload by 1/4 turn. The bearing arrangement is 2 tapered roller bearings like a car front axle. That did the trick!
As someone previously suggested, the bearings might need to be tightened to a small percentage of their max axial loading.
And so it seems.

Admittedly they were different but I've rebuilt Bridgeport, my Excello 602 and a 15 horsepower, 8000 rpm spindle, all had angular contact bearings at the spindle nose and a floating angular contact pair at the tail of the spindle. Someone with much more experience at spindle rebuild than I, said the bearings should be tight enough to get quite warm to the touch, but you can still hold your hand on the housing after the spindle has run long enough, at top speed, to stabilize the temperature.

 

[Lloyd-ss]

Admittedly they were different but I've rebuilt Bridgeport, my Excello 602 and a 15 horsepower, 8000 rpm spindle, all had angular contact bearings at the spindle nose and a floating angular contact pair at the tail of the spindle. Someone with much more experience at spindle rebuild than I, said the bearings should be tight enough to get quite warm to the touch, but you can still hold your hand on the housing after the spindle has run long enough, at top speed, to stabilize the temperature.

Good point about checking the machine when it is fully warmed up. 130 deg F is "quite warm" to the touch and 140 deg F is too hot to keep your fingers on. I have heard that same rule of thumb about circuit breakers: when fully loaded they will be quite warm, and that is OK.

I am expecting the new bearings today but will need to fabricate the dust shield and a couple of removal and installation tools before I tackle it. I am getting the feeling that what seems to be a slightly aggressive preload might be where I end up. Seeing I don't have a preload spec, I will just try to creep up on it until it seems right.
Thx, Lloyd

 

[Steamchick]

Hi Lloyd, - to digress a moment: Your comment: "I have heard that same rule of thumb about circuit breakers: when fully loaded they will be quite warm, and that is OK." --- Well, you hit the bull's-eye again. Yes, I did work on Circuit Breakers... (only 4 years in design mind you). And the full load rating was at a temperature of 80degrees C mav, based on the hottest "Ambient" of 55 degrees C = a 25degree rise above ambient. But a circuit breaker in a hot place with 55 degree ambient is almost too hot to touch, and 80C is definitely "ouch" time!
But "my" circuit breakers were carrying 5000Amps at anything from 145Kv to 500kV... so electrically, you would not want to be within 10 m of them! OUCH!

Ken

 

[Zeb]

Thoroughly enjoying this thread. I used the hand test on turbine fuel-oil heaters after shutdowns. So many opportunities in so many fields to melt calibrated skin!

I set the bearing preload on my Sherlines the same way. If the heat stabilized on the spindle, I was happy. Something I really like about the belt design is that it runs quiet on the kitchen island while the wife and kids watch movies nearby.

Hollow castings or gearbox covers will amplify sound as well. The inside of covers can get sound dampening material or often the bottoms are open to amplify noise off the floor.

 

[Lloyd-ss]

Thoroughly enjoying this thread. I used the hand test on turbine fuel-oil heaters after shutdowns. So many opportunities in so many fields to melt calibrated skin!

I set the bearing preload on my Sherlines the same way. If the heat stabilized on the spindle, I was happy. Something I really like about the belt design is that it runs quiet on the kitchen island while the wife and kids watch movies nearby.

Hollow castings or gearbox covers will amplify sound as well. The inside of covers can get sound dampening material or often the bottoms are open to amplify noise off the floor.

I am always encouraged to see how many of our responses and advice come not only from our model engine experiences, but from the other lives we have that help support our hobbies. Always something new to learn.

 

[Steamchick]

Hi Lloyd, As you know.
Simply: the laws of physics and engineering are universal. Full-sized or models, the same mathematical rules apply, and the same effects of lots of extraordinary factors like lubrication, heat flow, stress, gas dynamics, friction, etc., often with square, cube or quartic (fourth-power) factors affecting performance, strength, stiffness, durability, fatigue life, etc., of the models as well as the full sized whatever.
So to bring experience of such issues to the forum seems reasonable to me.
As they say "Size matters, but performance is everything, whatever the size!"
K2

 

[TimTaylor]

Lloyd,

I'm late to this thread - lots of good info here in the various posts. Sounds kike you have found some issues and are dealing with them.

One thing you might look at is the vibration signature of the various rotating elements to help ID sources. If you know someone with a vibration analyzer, that would be ideal, but there are alternatives. There are a number of free mechanical vibration apps that can be downloaded to your cell phone - same for audio. You want something that will give you an FFT readout that you can capture, The FFT will show what frequencies the noise/vibration is being generated at. Each frequency can be traced back to a specific source, i.e., bearings (ball spin, inner race, outer race), balance, alignment, belt pass, gear mesh, mechanical looseness, resonance, etc. Likely you have more than one source causing your noise. In a complex machine such as a lathe, vibration energy will also get transmitted to other components, so if you see the same fault signature on multiple components, you want to find the one where it is the strongest.

In belt driven machines, belt misalignment and improper tensioning are common causes of excessive noise and vibration, so if you haven't already, I'd for sure look there.

 

[Lloyd-ss]

Lloyd,

I'm late to this thread - lots of good info here in the various posts. Sounds kike you have found some issues and are dealing with them.

One thing you might look at is the vibration signature of the various rotating elements to help ID sources. If you know someone with a vibration analyzer, that would be ideal, but there are alternatives. There are a number of free mechanical vibration apps that can be downloaded to your cell phone - same for audio. You want something that will give you an FFT readout that you can capture, The FFT will show what frequencies the noise/vibration is being generated at. Each frequency can be traced back to a specific source, i.e., bearings (ball spin, inner race, outer race), balance, alignment, belt pass, gear mesh, mechanical looseness, resonance, etc. Likely you have more than one source causing your noise. In a complex machine such as a lathe, vibration energy will also get transmitted to other components, so if you see the same fault signature on multiple components, you want to find the one where it is the strongest.

In belt driven machines, belt misalignment and improper tensioning are common causes of excessive noise and vibration, so if you haven't already, I'd for sure look there.

Like most of our projects, deciding how deep you want to go down the rabbit hole is a cost/benefit trade-off. I have the new bearings in-hand but need to make remove/install tools before I take the lathe apart. I think the new higher quality class bearings will make a difference, but not what I am hoping for. The motor is definitely a big source of noise, too, and worthy of some effort.
Like many of us I have a tendency to turn solving a problem into a major science project. I want to avoid that this time..... If possible.
Lloyd