Bearing Frustration

I'm working on a project to build a little mechanism. It's something that I want to hand out at seminars, so I want to keep the cost down. But it needs to be able to hold together while one person uses it intermittently, possibly for years. The current design concept involves a sheet-metal base (with a stiffener in the right place), and a 4mm shaft that supports the working elements.

My specific frustration of the moment is bearings for the shaft. I'd really like something that I could plug into holes in the sheet metal that would support the shaft.

But pricing is nuts. I can go into a hobby shop and buy a package with half a dozen ball bearing assemblies in it for a toy helicopter, and pay $9.99. Or I can go to McMaster and get an instrumentation-quality bearing in the size (4mm ID) I need and pay $12 or so _per bearing_. I'm not even sure if ball bearings are the right choice given the fact I'm putting them into sheet -- a nylon or teflon bushing may be just right if the friction is reasonably low.

I've spent way too much time already looking around on the web. The only thing that I've found that's close is some panel bushings from McMaster -- I may end up with these, at least in the short term, or with hand-machined pieces of teflon just to prove the concept.

Where can I find cheap bearings? I don't need aerospace quality, I don't need copy machine quality. All I need is something that'll reduce friction more than stuffing a steel shaft through an aluminum hole, and that will stand up to light use for an indefinite period of time.

I'm assuming I need to look for something imported (I'm in the USA), and if this idea flies I'll be buying parts several dozen at a time.

Does anyone have any names of companies or distributors that may have these?

Thanks.

Reply to
Tim Wescott
Loading thread data ...

I'd start here...

formatting link
Depending on the material and bearing style you choose, you should be able to find something there for less than a dollar or two. Igus is very accomodating, even on small orders.

e.g.,

formatting link
Ned Simmons

Reply to
Ned Simmons

Tim Wescott wrote in news:YIudnW7xCrtA-sXbnZ2dnUVZ snipped-for-privacy@web-ster.com:

National Precision Bearings, IIRC is the name of the company I use. 4mm ID, 8mm OD, flanged ball bearing, seems like they were $4 or so each, double shielded.

Reply to
Anthony

I'd agree with that, but if you want more help from the NG you need to give us an idea of the loads. How stiff does it have to be, does it need to position the shaft with great accuracy? Is it feasible to make the plate(s) supporting the shaft from something like Tufnol (at least, that's what we call it in the UK, it's fabric reinforced phenolic resin, a bit expensive but it can be machined accurately and drilled and tapped.

Reply to
Newshound

For a lot of small bearings, particularly if they don't need high speed use, I go to

formatting link
They handle a lot of different small bearings, not just the stuff you'd expect in skates and such.

If you know the standard number, you can find bearings there for pretty cheap prices. As an example, the beaing for my Milwaukee angle drills is about $14.00 at my local supplier, but at skatebearings, I can get them for less than a buck each...

If this is really an intermittent, slow speed use, unless slop is an issue, I'd probably use ordinary plastic bushings if a punched hole wouldn't do the job...

Thanks

--Rick

Tim Wescott wrote:

Reply to
Rick Frazier

Actually I think Igus is going to be ideal for what I want. It'll be intermittent use, so the plastic will do quite well. In fact, I think that their clip style bearings are what McMaster carries.

I'm aiming to make this as inexpensively as I can while still ending up with something that will last for a few years. So I'm planning on bending the base plate up out of aluminum sheet with the bearings toward the top. I'll probably have to put a spreader bar right under the shaft to maintain some semblance of parallelism on the uprights, but that shouldn't raise the price too much.

Reply to
Tim Wescott

I've been asked to describe what I'm doing. You may get a kick out of it, so here it is:

I'm planning on starting up a set of control systems seminars for embedded software engineers. There is a huge expansion going on right now in mechatronics, which is creating a crying need for software people who can close control loops.

Unfortunately there aren't a lot of such people, even though there are a lot of people who are good with embedded programming. In my opinion many of these folks are more than clever enough to learn about control theory, and to apply it usefully.

Control systems theory gets pretty intensely mathematical, yet it needs to be grounded in real life. So I want to send people home with some sort of a trainer that they can set up at home or work and keep themselves fresh on the principals. My goals for the trainers are that they be:

  • inexpensive, with a target COGS (or more appropriately COGGA -- Cost of Goods Given Away) in the to 0 range. * Fun. * Not too hard and not too easy. * Visually appealing. * If possible, with many ramifications, so you can do simple algorithms to get basic performance, then more theoretically challenging (and complex) algorithms later for better performance.

One of the things that I think I can do to fit the above criteria is a propeller on a stick. You mount the propeller on a motor on the stick (light aluminum tube), you counterweight and pivot the stick so the thrust of the propeller will move it easily, you monitor the stick's position, then you try to program the microprocessor to control the stick's position using the motor and propeller.

I'm thinking of making this with a base that's bent out of aluminum sheet. The stick will be mounted to a shaft which rides on the bearings in question (thanks all of you for your suggestions). There'll be a spreader bar underneath the shaft to keep the bearings parallel and to strengthen the base, and the counterweight will hook around the spreader bar.

I think I'll mount the shaft to the stick with the counterweight -- I'll drill a hole in the counterweight for the shaft, attach it with a setscrew or glue (I'd like something cheaper than a setscrew, but less permanent than glue, actually).

I can get some potentiometers very cheaply that will accept a 4mm shaft with appropriate flats, so I'll put that on the board with my microprocessor and mount it to one of the uprights. Some paint or anodizing, a tasteful silk-screened logo, some rubber feet, a power supply and a JTAG debugger for the microprocessor will round out the physical assembly.

After that it's just software.

Reply to
Tim Wescott

IF I read your description correctly, what your trying to do is this... take a pyramid can be 3 or 4 sided with a bearing in the top and one in the baseplate. through these 2 bearings is a vertical shaft. on top of this shaft, like a letter "T" is another rod/shaft with a motor and propellor( horizontal axis) on one end at rt angles to the top of the "T" and a counterweight on the other. So if for example you feed power to the motor via a pair of slip rings, the propellor will spin and rotate the "T" round and around . the speed will depend on motor rpm/prop thrust . balanced by friction losses. However if you want to add some positioning potentiometers to the vertical shaft you will need some stops on this shaft so it cant go beyond the potentiomenter limits. Ie it wont be able to go round and around Also there will be inertia effects of the movement of the motor/counterweight mass. as well as prop and motor rotational inertia. this will give you over compensation problemsfor the micro controller.

One solution to this is not to use a motor and propeller as the rotational force. Consider using a simple back and forth solenoid that works like a clock escapement ( rachet and paul) that turns a small gear wheel meshing with a larger gear like a clock face so you could call say for a hand to position to 9 oclock and it would click its way to this position then stop. As you will need it to go the other way ie anti clockwise your need another solenoid working on the other side of the small gear wheel turning it the other way. Hope I followed your idea in principle. Ted Frater Dorset UK.

Reply to
Ted Frater

SMF74ZZ 4mm ID 7mm OD 2.5mm thick 2xShielded. I would pay £2 for one-offs. In your location I would probably find a bearing house and mention 10-off or

100-off.

regards Mark Rand RTFM

Reply to
Mark Rand

Propeller on a stick? Propeller? Look at Parallax for their Propeller! One of the most stunning microcontrollers ever.

But that OT here ...

Nick

Reply to
Nick Mueller

Stainless steel pop rivet in the hole and drill and reem it to the size of the shaft +.001 to .005.

Or,

Turn brass bushings and stake or glue them in place.

Reply to
Roger Shoaf

Pin it. Drill for a press-fit pin - press in, press out. Cheap, not permanent (barring massive corrosion)

Reply to
Ecnerwal

More or less. Actually, I'm planning on a horizontal shaft, so the sheet metal piece is bent in a 'U' with the bearings toward the top, and clearance for the counterweight below.

Right. I think I need a brace in there to keep the bearing supports parallel; this brace will also serve as a stop.

Exactly. This is the sort of compensation problem that I want to present to my students. Not that you want to go out of your way to make it difficult in real life, it's just that real life often presents you with these difficult compensation problems.

Your solenoid suggestion would be good for a suggestion of lower-tech alternatives.

Reply to
Tim Wescott

Of course. I prefer to think of it as "inexpensive", as in "my brother's VW bug is inexpensive" vs. "my Vega is cheap".

Reply to
Tim Wescott

Im not into electronics- im a mech eng, but am interested in this sort of problem. Now theres already a fully working positioning system in you computer mouse. The early IBM aka pre 1990 had a contact system, but later ones had infra red? slots in a wheel means of counting howmany times the slots exposed the sensor to the emitter. por mm's movement on the mouse mat. In 2 dimensions. but 4 ways. up, down, side to side you only need it in 1 dimension and in 2 directions. Then you wouldnt need a potentiomenter and its limitations on rotary movement. you could then go round and around several times before coming to a predetermined stop position. The innards of a computer mouse could be cannibalised to do what you want. Getting back to your original question,as to bearings . If you look at your 20 yr old black and Decker electric hand drill it has sintered brass self aligning spherical bearings held in the plastic casing. This would be your best type of bearing . Held in place with a simple domed piece of ali. your U shaped support would also be domed to hold the other bearing side. can only cost a few cents. Could set it all up in a couple of hours. Cos I /design/make this sort of small thing for a living.

Reply to
Ted Frater

So where do I find the bearings, cheap?

I had considered cannibalizing parts from mice, but an encoder like that requires an index pulse, and as I'm going to be making multiple copies I'm willing to spend a few cents more to buy parts that I can just plop into the finished product.

Reply to
Tim Wescott

Well as im in the UK I can only say what Id do here. Go to google USA and type in sintered bronze bearings. Ive found some 10 links even from here. Good luck.

Reply to
Ted Frater

What about using a piece of extruded square U channel instead of the bent aluminum? I don't have a good idea of the size you require, but chopping up sections of U-channel will be a whole lot easier than trying to fold & brace a piece of sheet metal. It will also be easier to accurately drill the bearing holes.

Doug White

Reply to
Doug White

snip

Your class sounds like fun.

You might be able to use the bushing on the pot instead of a nice bushed, machined base. Mount the pot to the vertical leg of a sheetmetal angle bracket with a long enough horizontal leg to serve as your spreader bar.

Something that I did not see mentioned was an adjustable friction source. Friction makes loop closure even more fun, especially nonlinear friction.

Regards, Bob

Reply to
BobH

skate board bearings sell on ebay and other venues for less than a dollar each, but since you want to teach as well, I'd just use plastic bushings -

Reply to
William Noble

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.