Yes, probably. A/k/a total loss lubrication

Yes it’s normal. Slightly annoying but normal.

On 20 September 2025 at 19:39 SillyOldDuffer Said:

On 20 September 2025 at 18:46 nappa Said:

I thought these cowells lathes where the dogs gonads

It’s all relative.   Taig are moderately outperformed by Sherline, and a Cowells should outperform both, and last for donkey’s years.  But you can spend far more if better is needed!  In decent nick, Schablim’s cheapest lathe costs more second hand than a brand-new Cowells, as do many posh watchmakers lathes.

Whether or not a fancy machine is needed for amateur purposes is moot.  Skilled operators get good results out of very ordinary lathes, provided they take their time.  A well-made lathe in good condition just makes it faster and easier to get results, and smooth controls make driving the thing more pleasurable.   I started with a Chinese mini-lathe.  The problem wasn’t that it didn’t provide the accuracy and precision I needed.  No, the problem was that it was too small.   As is the sainted Myford 7…

Dave

 

 

In reality, the plain bearing works much better than one might imagine. I too was a bit sceptical after first learning about it. But it’s incredibly rigid and accurate compared to similar lathes with ball bearings. Most watchmakers lathes use a similar arrangement, and they of course need to be super accurate.

But the obvious question is, why didn’t Cowells use a felt washer to retain the oil and slow the loss of oil?

If you want high spindle speed the 90cw appears to offer 40 – 4000 rpm. It uses different plain bronze bearings.

When you top up the oil how long does it last? If it is all gone over night it MIGHT be that the bearing needs adjusting. Since the lathe is new this seems most unlikely..

If after some while (six months?) you are using a lot of oil then the bearings may need adjusting. Tighten down until you can feel some resistance when you turn the headstock arbor. Then back off until it spins freely. This adjustment is very sensitive – a tiny movement of the screw can make a big difference..After adjusting run the lathe at top speed for a couple of minutes. If it slows or the bearings get hot back off a little.

On 21 September 2025 at 08:56 cogdobbler Said:

On 20 September 2025 at 19:19 SillyOldDuffer Said:

On 20 September 2025 at 17:44 nappa Said:

thanks guys you would think after spending 5k you would get bearings.

…

There’s method in the madness!  Imperfect bearings cause vibration and spoil the finish whereas it’s easy to make smooth plain bearings.   The main disadvantages are the need for messy total loss oiling, speed limitations and gradually wearing out.

Took about 5 decades for industry to learn how to make perfect spheres, so even very expensive  lathes tended to avoid ball-bearings.  During WW2 the allies and Nazis both imported high-end bearings from Sweden because only SKF made the really good ones needed for high-performance machines.  In the 50’s other makers cracked it.   Quite recently,  the Chinese Prime Minister remarked that China still couldn’t quite make the precision balls needed for ball-point pens, not cheaply anyway!

Today, really good bearings are still seriously expensive.   Not unknown for folk to buy a worn second-hand high-end lathe only to find that a set of specified replacement bearings cost far more than the machine!  Cheaper can be substituted, but the lathe won’t be as good as it should be…

Plain bearings are a good answer for the type of work done by a Cowells.

Dave

 

Tosh. Roller and tapered roller bearings have been commonly used in lathes — professional and hobby grade — since just after WW2. The Luddite  argument that plain bearings somehow run smoother blah blah was well and truly discredited within a couple of years of that. There is absolutely no basis in fact for it. As for expense, many small lathes use the same taper roller bearings used on car axles.  Dime a dozen.

 

As cogdobbler has used the pejoratives ‘tosh’, ‘luddite’, and “blah blah”, I’m tempted to reply in kind.  I’ll restrain myself to noting that his wording is offensive and he’s missed the point.

I don’t mind being corrected on errors of fact, but cogdobbler’s remarks show he hasn’t understood my post.  For example, Cogdobbler says, “Roller and tapered roller bearings have been commonly used in lathes — professional and hobby grade — since just after WW2.“.  Yes but so what?  Cogdobbler ignores that I said  “In the 50’s other makers cracked it.“. Claims I’m talking tosh, and then says exactly the same thing.  Duh!

I’m not a luddite.  I try to present a balanced view, in this example explaining why a lathe maker might legitimately choose to employ plain bearings.  I’m not suggesting that all lathes should be retrofitted.   Nor did I say that a plain bearing is better than a well-made roller!

I’ll try again:

• Ball bearings first came into prominence about 1870, needed for push-bikes.
• The theory is simple, reduce friction by rolling metal over metal, rather than rubbing.   Benefits include low-start friction, low running friction, and high RPM.   All very desirable, and huge effort went into development,
• Practice proved difficult. If the fit between journal and rollers is even slightly inconsistent, roller bearings brinell, wear rapidly, and vibrate.   Perfect identical spheres being hard to make,  I hope cogdobbler understands why it took industry nearly 60 years to develop a satisfactory way of making them, and why the process remained cloistered until well after WW2.   After it was proved possible, it took the major industrial nations another 15 years to make balls as good as Sweden, and as cheaply.
• Though their disadvantages mean plain bearings have mostly been replaced by rollers, plain still makes sense in lathes.  Easy to make, run smoothly, and provide good longitudinal support.  In good condition a 1925 Adept with a spindle running directly in a bored cast-iron headstock performs with low vibration and shake.

Consider mini-lathes.  Their design is 1960s-ish, and utilises ball-bearings.   Early examples were fitted with too cheap bearings, which soon misbehaved.  Owners had to replace them, often upgrading to taper bearings.   I guess later mini-lathes come with better made bearings, because complaints are fewer.  Point being that merely fitting ball-bearings isn’t “good enough”, lathe bearings also have to be “fit for purpose”.

Plain bearings are a reasonable alternative.  I can think of several reasons for not wanting a Myford, but their having plain bearings and total loss lubrication isn’t one of them!   Likewise, Cowell’s get consistently good reviews.  Other than splashing oil, there doesn’t appear to be strong reason to change the Cowell’s design.

Dave

I’m keeping out of this one, except to say that exceptionally good lathes, and utterly mediocre lathes, have probably been made with any type of bearing that you care to imagine.

MichaelG

Here are some ‘alternative facts’ 🙂

1. One of the main reasons Sweden took the lead in rolling bearing production was because their local iron ore was very pure. The fatigue life of a rolling element bearing is extremely dependent on the purity of the steel. Microscopic particles of dirt dispersed throughout the steel act as crack initiators which cause the raceway to break up under the cycling stresses of rolling contact. Swedens clean ore meant their bearings tended to last longer than others.

Nowadays the micro slag etc can be removed by almost anyone using vacuum degassing, vacuum remelting and various other processes, depending on how much you want to pay. The cost of these is coming down all the time so having pure iron ore at the start is no longer the advantage it was in the early days. The current cleanliness of the steel in some Japanese commercial bearings for £3 today can be better than that of very expensive aerospace bearings costing £5000 of just a few years ago.

2. The world still can’t make perfectly spheres and never will. The needs of industry are always one step ahead of typical production. Sweden might have had a head start in the early days of bearing production but others caught up quite quickly. I ‘think’ in the early 1900s, Hoffmann’s in Chelmsford were considered to be one of the world leaders in ball manufacturing. But the history is not clear on what this ‘lead’ really was. It may have been more to do with production capacity than precision. However I’m not sure whether the precision was actually that important in those early times. Most applications just wanted low friction, no one was too fussy about noise and vibration. What I do know is that a while ago I had the pleasure of stripping a 100 year old SKF bearing. I was shocked at the out-of-round in the balls!

Bringing bearings in from Sweden during WW2 was more about the high demand and lack of capacity in the UK at the time. We also imported bearings from the USA for use in Rotol hubs, because we simply couldn’t make enough.

3. It is a fact that if you want to analyse the smoothness of rolling element bearings, you test them on a spindle that’s supported on plain bearings. Plain bearings can be easily made to run very smoothly because, when correctly made and used within their limitations, the shaft is for all intents and purposes, supported/located in place by a film of oil which prevents any interaction between asperities and machining errors in the structural parts. The shearing of an oil film doesn’t make a noise or cause vibrations. Their main limitations are load and speed. The supporting film of oil is not very strong and will be broken if you apply too much load. Too little speed and you won’t create the film and as a consequence things will rattle around and the opposing surfaces will wear. Too much speed and it will get hot and seize.  These limitations can be coped with if the application is not so demanding, but stuff in a rolling element bearing instead and by and large these challenges melt away. Unfortunately, as I have said before, rolling bearings need 10 times more care and attention to the details around the companion structure. Too often rolling bearings don’t get what they deserve.

4.  Up till the early 2000s nearly all computer hard-drive spindles were supported on rolling bearings. Throughout this products design life, a huge amount of effort was expended by the Japanese bearing industry to constantly improve NNR (non-repetitive runout). This is the variation in runout that changed a tiny bit with each revolution of the inner ring. It is important if your disc tracks are highly compressed and you don’t want to keep losing your place. The main cause of NNR is ball sphericity plus that ball/cage speed is not synchronised with inner ring speed. Eventually demands of the computer industry for ever tighter NNR got too much and a market of hundreds of millions of pieces a year of rolling element bearings disappeared very quickly, to be replaced by plain bearings, first oil, then air lubricated.

Conclusion: What’s best Plain or rolling ?… it’s always just a combination of horses for courses technical, marketing and good old fashion economics.

Gerry