Poly-vee problem - S7 mod

I converted my S7 to poly-vee belt primary and secondary drives back in 2012, since when I've had a problem that is sort of resolved, but not entirely.

To cut the grooves in the countershaft and spindle pulleys the otherwise finished pulleys were mounted on mandrels, and the grooves cut with a professionally ground HSS form tool, to ensure correct form and concentricity. The grooves run true to better than 0.0015" radial and axial run-out.

When running on the 2105rpm speed setting (and only on the 2105 range) , the headstock develops a clattering noise, positively identified as being caused by the spindle pulley moving laterally on the spindle and bumping into the back gear.

If I assemble the headstock without the back gear in place (so that the pulley has no restriction to its lateral movement) and run just the pulley, it shuffles around by something like 0.020", perhaps a bit more.

The shuffling gets worse as belt tension is increased, only disappears when the tension is so low as to be all but useless.

Now for the strange bit - it does this once per BELT revolution, which seems to suggest that the pulleys are ok, and indeed they seem to be - see above.

I've tried 3 different belts - Pixar (came with the Hemingway kit), Optibelt and GoodYear. All do the same thing.

I tried biasing the pulley shuffle towards the tail end of the spindle (to stop it hitting the bull wheel) by shifting the countershaft pulley leftwards, to try to hold the spindle pulley against the angular contact bearing assembly. But to stop the rattling the pulley has to be so far offset as to cause the belt to protest.

I discussed the shuffling problem with someone who is expert in poly-vee belts, he had not come across this before. He suggested it might be worth cutting a belt down to 4 ribs, as 6 ribs (as designed) was possibly over-belting the machine.

I did that, it might have made a small difference to the shuffling motion, but nothing significant. However, it did make a difference when trying to bias the movement - the countershaft pully only needed to be shifted left by less than half a groove width to stop the spindle pulley hitting the bullwheel, and the belt seems happy enough with that.

So that's how the machine's running now, no rattling and working ok. But I don't like things that I can't understand, and the solution is a bit of a bodge. I'd rather have it right, full stop!

So, can anyone shed some light onto this strange phenomena?

Any tendency to sideways movement would suggest to me that the axis of the two spindles are not exactly parallel. The effect would be greater with a wider belt.

In one sense a wide polyvee belt is related to a plain flat belt but flat belts invariably run on crowned pulleys.and centre themselves automatically. A Polyvee is stuck where you put it so its needs very accurately made and aligned pulleys to ensure all the individual ribs carry the same load.

It would not surprise me to learn that the fabric backbone of a Polyvee has some inherent bias depending on the warp and weave angles of the fibres. I would think any bias effect would be undetectable in normal setups.

Ian P

It would only take a few thou out of parallel on the jack shaft to cause a problem, could there be a little wear in a bearing at that end of the shaft.

Ian S C

Sounds to me that he pulley bore may not be dead perpendicular to the grooves.

Neil

.

Unlikely, Neil; given that Cabeng wrote:

Now for the strange bit - it does this once per BELT revolution, which seems to suggest that the pulleys are ok, and indeed they seem to be - see above.

MichaelG.

Hi Cabeng - I see our interests coincide again.

I have noticed the same problem with the Z section belt on my Super 7 Plus. Having got it otherwise running like a sewing machine, I wondered why there was a rattle on just the top two highest speeds. A squirt of oil into the headstock pulley bearing confirmed the source was endfloat. The oil stopped the noise for a few minutes until the surplus oil worked its way out. Running the lathe at slow speed revealed that the belt had a sideways lump, for want of a better description. While it is an annoyance, I decided that it was not a 'problem' as such and that I should live with it until the belt is renewed, probably for a Brammer type that has had some positive feedback on this forum.

I suspect that any belt, whether vee- or poly- is going to be imperfect in relation to the 5 thou or so endfloat recommended by Myford. If it is really unbearable then I wonder if there is another approach and that is to stop the metal-to-metal contact causing the rattle. Myford used fibre washers on the ML7 countershaft and that didn't rattle. While I wouldn't choose fibre because it seems to release fibrous material that makes the oil run black, a very thin washer machined out of a slug of PTFE or acetal might do the trick. One washer on each side of the pulley.

By the way, I would be interested to know where you were able to obtain NSK 7007A bearings in the UK. I couldn't find any but bought mine while passing through Hong Kong. I was offered two types, both marked NSK 7007A. One had a pressed steel cage like the Myford originals and the other, for the equivalent of a pound or so more, had a beautifully made machined brass cage. I chose the latter and went to the trouble of machining a thin wall tube to press the outer races in without subjecting the balls to any damage. They are totally silent in operation, unlike the steel caged bearings in electric motors.

Andrew M

Andrew: They were indeed difficult to find! As I recall, the problem was the 30 degree contact angle, which seemed to be out of fashion amongst the major manufacturers. 7007B (40 degrees) and 7007C (15 degrees) were more readily available, but I wanted to stick to the Myford original design, and didn't want to use bearings by some unknown manufacturer. I got the bearings from this company:

http://www.compassbearings.co.uk/

They're not far from me, and say they specialise in finding obsolete bearings! Well, they did find them:

but they were a) much higher precision than I wanted/needed, and b) bloody expensive - but I did get a very good discount, so not too bad! I too made a bearing puller to install them in the headstock.

Mine were about £13 each, last year. Perhaps I should import a few?

Andrew M

Hi Caberg, I have a S7 but have never had occasion to dismantle the headstock, however having read your detailed explanation of the problem one thought comes to mind, maybe good maybe bad, maybe unworkable !

Have you considered fitting Belmont spring washers either side of pulley ? Some machining would be required I'm sure and some care to ensure a suitable end pressure. It may also be necessary to incorporate an additional, maybe hardened washer at the bull gear to prevent wear when using back gear ? All hypothetical until a proper study is done.

Regards John

> By the way, I would be interested to know where you were able to obtain NSK 7007A bearings in the UK.

Take a look at this current thread

I'm sure Ketan 'Mr Bearing' Swali wcan get you what you need at a fair price.

Neil

For what it is worth; since you connect the 2 pullies with a drive belt you also have to add the inaccuracies of the two together to get the maximum inaccuracy of the system. Depending on the ratio of the system that maximum occurs only ones in X revolutions of the drivebelt. At high speed it becomes in resonance. Niko.

The fabric of the belt makeup must have a start and end of wrap, on A belts I have found the end often corresponds with the "slap" of the once per rev syndrome. A belts are cut from a tube and the edges will always have fabric strands on the running surfaces. Another clue?

I looked at Speelwerk's scenario today, unfortunately it doesn't fit the once per rev feature.With the pulley diamters and spacing I have, any maximum (or minimum, or anything else) would only repeat once every 2.23 belt revolutions. Also the run-outs on the pulleys are well within the limits specified by Conti-Tech in their design manual.

Ian P & KWIL: the circumferential strand is indeed spiral, so I tried two different 6-rib belts two ways round each - no difference. Also did the same for the cut-down 4-rib belt, again no difference. Cutting it down to 4-ribs ensured that it wasn't contacting the sides of the pulleys, so any protruding matter was running in fresh air.

Andrew M & John F: I tried both suggestions in one hit - made a PTFE washer 0.020" thick, bored slightly undersize on the i.d., so it set itself up as a 'PTFE-Belville' washer when installed. It has stopped the shuffling, which only needed a very small force to restrict it - the pressure exerted by a DTI probe would stop it, even the stylus of a Verdict indicator reduced the shuffling.

So thanks for those suggestions, it's cured and the machine is now very quiet. But I still don't know why the pulley wants to move about!

I've changed belts 6 times, and then stripped the spindle, made the washer, and re-assembled the spindle, all in one afternoon. If there were an international competition for belt changing, I'd be at Olympic standard, I can do a belt change in well under 30 minutes now!

I would add one other idea, when chasing the "slap" I found that any minute tightening of the belts due to pulley run out or belt width variation caused the shafts to pull inwards towards each other. How you might ask, the two motor platform suspension strips (otherwise known as the tension adjustment arms) had a slight clearance on the upper round rod through the casting, and so this can knock. Ends modified to have a boss and set screw, so those do not move unless required by belt adjustment. One less point to worry about.