Dynamic balancing pulleys

[Steve WithnellParticipant @stevewithnell34426]

[Steve WithnellParticipant @stevewithnell34426]

I have a 6 inch cast iron pulley (it's a single poly-v type belt) which is a bit out of balance (1500rpm is not good) and it would be good to understand how to dynamic balance the pulley whilst it is on the lathe.

I did see an article a while back using an oscilloscope and a loudspeaker as a generator – a piece of fishing line was fixed to the speaker cone and the other end to the lathe, so that vibration in the lathe caused the speaker to generate a voltage, which could be seen on the 'scope.

This sounds like an interesting little electronics project.

Any one done this?

Steve

[Neil WyattModerator @neilwyatt]

Hi Steve,

My problem is my lathe is quite well balanced…

I just tried this. with a string you could pluck it and see a nice sine wave on the scope… but no vibration got through

With a cheap speaker held to the headstock by its magnet and a 4 oz bit of cast iron on top of the cone, to create a sort of inertial detector, with the probe on X1 I could just get a signal of about 10mV. I then put an out of balance lump in the chuck and wound up to about 250 rpm. With the lathe vibrating significantly, the signal only doubled to about 20mV, but was much clearer and less noisy.

I found the best place for the 'inertia sensor' (lump of CI) was partly balanced on the rim of the speaker and the other end touching the cone, so it didn't just force the cone to one end of its travel. A weight glued to the centre of the cone and the speaker held horizontal might work better.

Obviously the frequency is pretty low, with the lathe running at about 600 rpm, the signal is only at 10 Hz. Make sure the scope is set for DC!

Have a play and see what you get.

Neil

[frank brownParticipant @frankbrown22225]

As noted by Neil, a loudspeaker is a low output device. I would try a magnet laying on the headstock and use the coil of a mains relay as your pick up device, lot more turns , lots more voltage? Also overcomes any resonances of the string.

1500 RPM sounds like a the same speed of a 4 pole motor speed, so the problem could be anywhere in the drive chain (belt).

Have you statically checked the balance?, mount the pulley on a shaft and put the shaft across two knife edges. It should not roll and be able to be parked at any random position. Problem could be a stiff section of your belt?

Wish my lathe would run that fast, vibrations or not.

Frank

[Steve WithnellParticipant @stevewithnell34426]

Looks like a bit of trial and error to get a decent voltage generated – I wonder if a crystal microphone insert would do the trick, should be a lot better then 10mV. What I thought was, a hall effect pickup could provide a marker at one point on the pulley (ie mixed with (added to) the vibration signal), so it would be possible to work out where the sinewave peak was in relation to a point on the pulley itself.

Sound like first job is to generate a decent voltage amplitude.

I know the pulley is out of balance, the vibration only arose after I fitted it (one of those amateur mods ), but I didn't do a static balance first. The pulley is about one inch thick CI, so it does have a lot of mass, which I want to reduce as part of the balancing.

The lathe as it is currently configured has a variable speed DC motor, with a top speed of 2500rpm. I did set it up for 3000rpm originally then backed it off to 2500rpm. That's with a 5 inch chuck as standard. The machine ships with the controller set to 2000rpm as a maximum.

Steve

[Les Jones 1Participant @lesjones1]

Hi Steve,
I would have thought that static balancing would be enough for a thin object like your single belt pulley. I think to make a dynamic balancer you would need two sensors to be able to decide how much weight to add or remove from each side of the pulley. (One sensor either side of the pulley.) I think putting the pulley on a shaft resting on two horizontal knife edges will be the easiest way to balance it. Here is a link to a balancer for a very demanding application. (A turbine rotor.) I think this will still only do static balancing. I've just done a bit more Googling and found this information on dynamic balancing.

Les.

Edited By Les Jones 1 on 03/07/2014 20:49:46

[Tony RayParticipant @tonyray65007]

Steve,

There is an excellent article by J. Hugel, Google his name and Vibrascope. He uses an acellerometrer device and a true RMS meter. You can download it from Sribd. What about using a steel guitar string and a pick up ?

Tony

[Clive FosterParticipant @clivefoster55965]

Trying to do the job in the lathe will just generate more noise.

Back in the late 1970's I used a high sensitivity turbocharger(?) rotor balancing machine to balance experimental rotating scanner systems. Naturally once the project was finished we took the covers off to see how it worked and why it cost so much. The thing was exceedingly simple. Two sensors, two Vee blocks for the rotor shaft to run on, two variable gain amplifiers and two voltmeters for output. As I recall matters the meters and amplifiers weren't anything special, certainly no messing around with true RMS. Drive was via an O ring from a variable speed motor in the base. Operating speed range was pretty wide topping out at several thousand RPM in our case due to the thin shaft used wth our little 3 and 4 inch diameter scanners. Ultimate maximum was well north of 10,000 rpm. Worrying when you consider that the O ring drive was the only thing keeping the specimen in the V blocks.

Despite the simplicity it worked very well indeed.

Clive

[Tony RayParticipant @tonyray65007]

Steve,

Smartphones have accelerometers and free or cheap apps, Google 'dynamically balance EDF units' it's all there.

Tony

[HopperParticipant @hopper]

Seems like static balancing between centres or on rollers etc should be good enough for 1500 rpm use, shouldn't it?

[Steve WithnellParticipant @stevewithnell34426]

Static balancing would be perfectly good, and in any event, if I do set up some form of dynamic balancer, then I would only be looking to fix the static balance anyway.

Given the lathe is variable speed and a I have a brick wall as a fixed reference, it seems like something to play around with.

Steve

[WALLACEParticipant @wallace]

With all due regard to it NOT getting caught up in things going round, would not an engine stethoscope be simpler ? Or maybe the low-ish
frequency wouldn’t be to easy to detect ??

Thinking aloud !

W.

Edited By WALLACE on 04/07/2014 08:44:01

[WALLACEParticipant @wallace]

Another idea.

How about a laser bolted to the lathe pointing to the ceiling ? Adjust work piece for minimum deflection of the dot?

Maybe spring / mass load the laser so it resonates at a couple of hz, if you use a vfd to ‘tune’ the lathe’s out of balance vibration to be the same, this would effectively ‘magnify’ the wobble on the projected laser line ??

W.

[Clive HartlandParticipant @clivehartland94829]

Hi Steve, initially you should carry out the machining of the pulley before any balancing act you do! At the moment you have a chunk of metal with minimal machining done on it so there will have unbalanced forces if it is rotated. Thinning of the web and profiling the boss will soon remove any of these out of balance factors. I would surmise after machining that there will be no unbalance effect.

Clive

[Les Jones 1Participant @lesjones1]

Hi Wallace,
I like your laser idea. An extension to the idea would be to mount a photo diode on the ceiling so it could be positioned at one end of the line produced by the laser. The signal from this could be amplified and used to trigger the flash of a strobe which would then help to work out where to remove metal from the pulley.

Les.

[Neil WyattModerator @neilwyatt]

Or even an extension of noise cancelling techniques and have a couple of large masses on actuators so you can cancel out any vibration before it happens

Neil

[Ian S CParticipant @iansc]

I would go along with Clive, and you have plenty of metal so that after you thin the web between the hub, and rim, you could bore maybe half a dozen holes in the web area to lighten the pulley.

Does the fact that the motor was set at 2000rpm at the factory not indicate that this might be the speed they wanted the lathe to run up to?

Ian S C

[Steve WithnellParticipant @stevewithnell34426]

Hi Ian,

I'd have to go through the documentation again, but I think the top end speed setting has an impact on bottom end torque, so I think the issue is one of a trade off. (lots of thinks in there, it was a while ago…). The lathe is shipped with a 2:1 reduction between motor and spindle and I introduced an additional 10:1 – they didn't intend that at the factory either! The problem that introduces is that the additional speed reduction interfers with the current overload protection, which is why I suspect the mod isn't done commercially. I just have the belt set so it will slip if there is a jam. Anyway that is why there is a big 6 inch pulley in my Sieg bench lathe.

The pulley does need quite a bit of machining, which as Clive points out might make the problem disappear anyway. How thick should I leave the web?

I've ordered some piezo transducers (aka buzzers) they are as cheap as chips – 10 for £4 so that will give some scope to play around, I also bought 10 laser modules a while back for about the same money, seemed a handy thing to have in the "makings" box.

Steve

[David KavanaghParticipant @davidkavanagh47654]

Hi

I am very interested in building a dynamic balancing machine based on speakers but does anyone know where I can down load the software program from to run it like detailed in "les Jones 1 post this information" any help would be appreciated.

Dave

[Ian S CParticipant @iansc]

You could probably take the web to half the width of the rim, say 10 mm, don't know the size of the belt.

Ian S C

[Author]

Posts