What's new
What's new

How to measure a cam

Machinist_max

Aluminum
Joined
Dec 15, 2016
So I have a cam that I need to copy so that I can produce it on my cnc.
This cam is part of a timing device that lower and raises a table on a cardboard cutting and printing machine.
There is two cams that have follower "fingers" that ride on the outside.
It has a flat spot on it along with two different arcs. [ what we think is two arcs but maybe not]

So my question is... is there a easy way to measure the outside to get the arcs ["size and locations where they meet including the flat spot] that I can draw it up in cad.

I do have a brown and sharpe manual cmm machine with the reflex control. I have only had it for a couple months so still new to that machine.

The cardboard machine company who made it originally went out of business many years ago and the ordered replacement [ one in the pictures that Im coping] is $$$$$, hence why we are trying reproduce it.
The cams life span is around 10 years.

Tolerance spec-ed is +/- .003

cam 2.jpgcam 3.jpgcam 1.jpgus corr.jpg
 
Seems that you have the tool for the job right there, the CMM. If you're not comfortable using it for this job get a B&S applications engineer to visit and go over the process with you, you'll get the part data and an education at the same time.

Note that it looks like the OD of the cam was induction heat treated, so factor that into making new ones.
 
So I have a cam that I need to copy so that I can produce it on my cnc.
This cam is part of a timing device that lower and raises a table on a cardboard cutting and printing machine.
There is two cams that have follower "fingers" that ride on the outside.
It has a flat spot on it along with two different arcs. [ what we think is two arcs but maybe not]

So my question is... is there a easy way to measure the outside to get the arcs ["size and locations where they meet including the flat spot] that I can draw it up in cad.

I do have a brown and sharpe manual cmm machine with the reflex control. I have only had it for a couple months so still new to that machine.

The cardboard machine company who made it originally went out of business many years ago and the ordered replacement [ one in the pictures that Im coping] is $$$$$, hence why we are trying reproduce it.
The cams life span is around 10 years.

Tolerance spec-ed is +/- .003

View attachment 197054View attachment 197053View attachment 197055View attachment 197056

Looks like you need to make a flanged shaft to bolt the cam on. put a degree wheel on the shaft and a dial indicator on the cam. Pick a bolt hole to call "0", set a pointer to 0 on the degree wheel, probably check and note the dial indicator reading about every 5 degrees or so and there you go.
 
That is what I was thinking, use the cmm to set that center hole as my 0,0 and basically pie cut it into x number of points that I can get a distance from the origin. In cad just connect the points in essence.

does anyone see a issue with through hardening the entire part not just the "following surface". Over the ten years they get out of the cam they literally eat through that entire hardened area.
 
If you have any CAD capability you should be able to spline a sufficient number of points into a usable cam profile, at least 20, more in areas of transition. More is better. Does the CMM offer any software for point output? Do you have the manual for the machine?

No big issues with through hardening the part as long as you don't have further machining on it. If you're wearing off that much material, what does the mating part look like? I'd make extras and change them before they get thoroughly worn out.
 
I can't be sure but that looks like a harmonic curve. Ask a B&S or Tornos type screw machine shop where they get their cams made. You may strike lucky and find a cam maker with a CNC cam mill,harmonic movements a doddle. If you do find one ask him if he will need lever arcs.

Cam maker on every street corner on this side of the pond!
 
So my question is... is there a easy way to measure the outside to get the arcs ["size and locations where they meet including the flat spot] that I can draw it up in cad.
You could do it yourself but if you want to save a little time, just send it to

Megacycle Engineeering
San Rafael, CA
(415) 472-3195

Cams are their business so they are all set up to do what you want. If you call them you'll get Barbara, she can be a little gruff but knows her stuff, hang in there.
 
Sending it out to another place to defeats my reasoning to make it.
I make quite a few other parts for this manufacturer and they asked if I could make the cam at a lower cost.

I have no issue with making it, and am greatly against sending something out to another shop to produce something that then I put my name on and a mark up, poor business methods.

But yet I am a small two guy machine shop, so I have come to notice that the businesses I deal with like it that they guy they talk to over the phone and meet is the guy that is going to produce the part they need.

Most of the stuff they need is modifications. New bushings and repairing long shafts with torn up bearing journals. [weld build up and turn them back to diameter]

Do most shops not want to do singular parts, because that is the stuff I love. Its always interesting
 
Sending it out to another place to defeats my reasoning to make it.
I meant measure it. Sorry. They have a computerized cam checker, can just roll the part between centers with a probe and viola, there's your numbers.

But yeah, it was just an option. You can do the same thing, just slower.
 
Sending it out to another place to defeats my reasoning to make it.
I make quite a few other parts for this manufacturer and they asked if I could make the cam at a lower cost.

I have no issue with making it, and am greatly against sending something out to another shop to produce something that then I put my name on and a mark up, poor business methods.

But yet I am a small two guy machine shop, so I have come to notice that the businesses I deal with like it that they guy they talk to over the phone and meet is the guy that is going to produce the part they need.

Most of the stuff they need is modifications. New bushings and repairing long shafts with torn up bearing journals. [weld build up and turn them back to diameter]

Do most shops not want to do singular parts, because that is the stuff I love. Its always interesting

I have been following this thread, as I have a similar problem. This is not so simple. I think you may have underestimated the task. First, most cam profiles are not linear. They rarely are made up of simple arcs. Most cams provide varying rates of follower acceleration and deceleration within their cycle. So, to document a cam shape, you need to plot follower motion per degree (or less) of cam rotation from some index point on the base circle. You may also note that the contact point on the follower likely sweeps across the follower pad surface during the cam's cycle. This means that the type and shape of the follower is also very much part of the plot. I have done this with a simple degree wheel mounted on the cam and a dial indicator on the follower, but the plot is quite crude. I am certain using a CMM is the better plan, but every CMM that I have seen has no facility to accurately rotate the cam. I assume there should be some kind of a dividing head used, but all of mine are way too large to fit on any CMM that I have seen. I hope that someone can enlighten me on how to do this on a CMM.
 
I have been following this thread, as I have a similar problem. This is not so simple. I think you may have underestimated the task. First, most cam profiles are not linear. They rarely are made up of simple arcs. Most cams provide varying rates of follower acceleration and deceleration within their cycle. So, to document a cam shape, you need to plot follower motion per degree (or less) of cam rotation from some index point on the base circle. You may also note that the contact point on the follower likely sweeps across the follower pad surface during the cam's cycle. This means that the type and shape of the follower is also very much part of the plot. I have done this with a simple degree wheel mounted on the cam and a dial indicator on the follower, but the plot is quite crude. I am certain using a CMM is the better plan, but every CMM that I have seen has no facility to accurately rotate the cam. I assume there should be some kind of a dividing head used, but all of mine are way too large to fit on any CMM that I have seen. I hope that someone can enlighten me on how to do this on a CMM.

We shall find out later this week when I start on it.
Ill take pictures through out my process and put them up.
 
If you happen to have a 5-axis machine, or a normal machine, with a good rotary table and a good probe....
Mount the cam on a shaft. Probe. Turn shaft as small amount as possible. Probe again.
Idea here is what Moonlight machine described, except you are getting a machine to do it, and trying to get lots and lots of points.
(As in, no two points are more than a few thou apart.)
 
I have been following this thread, as I have a similar problem. This is not so simple. I think you may have underestimated the task. First, most cam profiles are not linear. They rarely are made up of simple arcs. Most cams provide varying rates of follower acceleration and deceleration within their cycle. So, to document a cam shape, you need to plot follower motion per degree (or less) of cam rotation from some index point on the base circle. You may also note that the contact point on the follower likely sweeps across the follower pad surface during the cam's cycle. This means that the type and shape of the follower is also very much part of the plot. I have done this with a simple degree wheel mounted on the cam and a dial indicator on the follower, but the plot is quite crude. I am certain using a CMM is the better plan, but every CMM that I have seen has no facility to accurately rotate the cam. I assume there should be some kind of a dividing head used, but all of mine are way too large to fit on any CMM that I have seen. I hope that someone can enlighten me on how to do this on a CMM.

I agree. It is most likely not arcs at all, but nurbs. Alot of cnc/cad-cam can program to nurbs, but they will be line segments and not smooth at all. Look up G06.1

G-code - Wikipedia

Also, I know Pro-E (creo) can do this, but you can do an analysis of the best possible arc to arc to line segment in the world and it will have stress risers at the transition points. Only a true nurb will not.
 








 
Back
Top