I hope i'm putting this in the right place, if not, please let me know...

I have an idea for something i want to build, but it requires that i build a double sided cam of sorts..

i have a lathe with independant 4 jaw, and a manual mill but i havent completly got my head wrapped around how to go about making what i want.. as of right now i'm thinking the lathe, chuck up the blank, off center, turn it, then rotate the part 180 in the chuck and repeat...

better ideas? i'm all ears..

what i want to make is basically a football shaped cam, just maby not so pointy at the ends... this will spin and actuate some linkage to do what i want it to do... i'm thinking that the cam followers will have to have rollers on them, like roller lifters or something of that sort to reduce friction... this cam will be mounted on a small stepper motor so that i can control the position and speed of its rotation easily....

attached is an extreemly crude sketch.. i dont have any better drawing tools here at work...

____________________________________________________

after some more thought, i think i have changed the design, i think i would rather make a more traditional cam where the lobes are side by side, have 4 of them, 90deg apart, and then have a rocker arm configuration that rides the cam.. would the machining of the cam, or i guess you could look at it like a crank shaft as well, be about the same, chuck some round stock in the lathe off center, cut 1 lobe, re-chuck 90 degres, machine the second and so forth?
this way i could machine centered bearing bosses on each end and have the cam ride on its own, then run a toothed timing belt from the stepper motor to the cam.. this would also allow me to change the gearing between the motor and cam if i ever needed to...

at this point i really dont think that the rpm's will be that high so i dont think that balance will be that big of an issue...

i dunno, i'm kinda flowndering on this design... i'm open to suggestions, ideas, questions, links, pictures, whatever...

......attached is an extreemly crude sketch.. i dont have any better drawing tools here at work......

If it is attached you have hidden it well :) .

your absolutely right, i kinda felt like i had forgotten something...

ok, so here are the sketches of the football cam, and a more conventional cam...

really...

ok, so suggestions on how to make the cam?

JP, I've made cams before on a manual mill using an index head and a the z axis. mine was a typical lobed cam with two radii and two connecting, tangential lines. the concentric radius and flats are easy, for the lobe i worked out the math in excel with a series of divisions and z heights. your ellipse would i think lend itself to the same idea, its not as time consuming as it seems, but if you use a lot of divisions its amazing how well it'll come out - this assumes you not planning on making 1000's :D

.......ok, so suggestions on how to make the cam?

What you call the conventional cam should be simple; they look like circles mounted off-center. These you can do in your four jaw, you set them up so the off-centeredness is the same as your cam lift and then bore the shaft hole.

The 'football' cam is a bit more involved but, if it the shape is actually two blended circular radii, not too difficult if you have a rotary table for your mill. If it is circular radii then you know where the centers are located.

The blank for the cam will probably start out as a piece of round stock. Bore the center hole that will mount the cam to its shaft and drill two mounting holes an arbitrary but known distance apart about the center hole.

You know where the location of the centers for the radii that make the cam shape are relative to the center hole and therefore relative to the two mounting holes. Now you take a piece of stock and fasten it to the rotary table mounted flat on the mill table. Make sure you know the dial settings on the handwheels to align the center of the table with the spindle centerline.

Now drill and tap four sets of mounting holes, spaced the same as the cam blank, which are positioned correctly with reference to the center of the rotary so that when the blank for the cam is bolted on this piece it will be rotating around the center point of one of the radii for the cam shape.

Now you can machine that portion of the cam by rotating the rotary table.

Blending the radii will take a bit of finesse and you can cheat a little by using a small straight line segment between each pair of radii. This will cause less perturbation to the follower motion than an inadvertent dip cause by machining the smaller radius past the blend point.

One thing you need to bear in mind is the the motion of your cam follower versus rotation of the cam does not track just the cam profile but a combination of the profile, the radius of the follower and the way the follower is mounted on an arm. This can be figured out but takes a bit of trig.

wow, yea, most of that went right over my head... i dont have a rotary table for the mill yet, as i have put all of my $$ into getting the parts to eventually do a cnc conversion, still scrounging around for stuff for that..

as for the more traditional cam, my plan was to just end up with off centered circles, i really dont need the dwell and snap of a conventional cam, one where there is a sharp flat ramp up to max lift, a small radii, then a straight ramp back down.. this dosent need to be that complex.. this just needs to move a set of arms up and down, about an inch each. actually 1/2 inch would probably do the job, and would leave more meat holding each lobe to the next.. i plan to start with 2" t6061 roundstock.. i think that would be good enough for what this is for...
make each lobe 1" wide, the rocker arm would be 3/4' wide, fingered on the end so that a 1/2" wide bearing can fit into the fingers.. this would allow plenty of clearence between the rockers

my thinking was to make my off centeredness be 1/2 inch from the stock centerline, then just turn till smooth and round, rotate 90 deg in the chuck and repeat.. one question, since the order really dosent matter to me, i was just going to put them hitting the top of thier stroke 90 deg off from each other for simplicity, but do you think that there is a better timing scheme, like do the first at 0, the second at 180, the third at 270 and then the last at 90.. would it matter for strength? i think going in order would look cool...

the other thought that i had was to not put the lobes right next to each other, but to turn a lobe 1" wide, then space over about 1/4" and then start the second lobe.. this would leave a full size 2" disk between each lobe and little more material holding the lobes together... i think... dosent hender my plan, just changes the total length of the cam...

then the rockers will have a 1:2 ratio.. 1/2" travel at the cam makes 1" at the work end...

last question for this post, what, how to make the pivot points.. my original thought was to take some 1/2" alum plate, one at each end of the pivot, standing up on edge, and have a solid shaft go through all of the rockers.. drill through the rocker and insert some sort of bearing sleve, maby brass or something, but then what would i use for the pivot rod, i know i should have a hard against a soft... bearings just would probably be too much work and expense, altho, this will have to transmit some decent force... maby needle bearings are the way to go...

i dunno, whaddya think?

no takers?
the other question that i thought of was how to accurately position the live center on the far end of the stock when its not mounted centered? i guess i could mark it off and hope i'm close enough..

JP, I confess, I don't really understand the pivot point question - are you referring to the tooling setup or the finished product?

if you turn something in the lathe off centre, you get an eccentric circle, not an ellipse, football or lobed cam - know this is fairly obvious, but i haven't followed how you are going to cut ellipse shaped cams in the lathe (unless you move the cross feed indexed to the spindle, either mechanically or electronically)

on your question about the live centre, whenever you are turn eccentric stuff, and its long and spindly, like a crankshaft, you have to have at the tailstock end centre drilled holes accurately located for all the axis you need. they need to be dimensionally accurate and properly indexed (best drilled in the mill using coordinate drilling) - this of course is required at the headstock end as well, otherwise you will not able accurately index the eccentrics. this can either as part of the original stock and machined away later or as a temporary clamp-on device.

what is it you are making? if its for a prop or wooden toy or something you might have a lot of latitude. However if for something more demanding, like machinery or an engine, usually with a cam you thinking how much lift, how i want the to accelerate and how long is there lift - this makes it difficult to substitute an eccentric and get desired results

my explanation on cutting cams with the mill was maybe too brief - did you understand what i meant? you can do your ellipse in the mill, but you will need a dividing head.

it the device is simple, not very demanding and only needs eccentrics, why not go with a built up design. drill and ream a bunch of disks, mount on a stub arbor held offset in the 4 jaw, turn the eccentric od then loctite them top a shaft. done.

hey Mcgyver.. ok, first of all, YOU STOLE MY NAME!!! hehehe but seriously, thats exactly how i spell it, and infact is my email address... has been for years and years.. anyway.. no biggie...

ok, so i got to thinking bout your built up stack of disks.. interesting idea, much simpler to build.. my only worrie would be if one of them spun on the final mounting shaft... i guess i could use something like allthread, and put a nut between each one.. that would clamp them in place pretty good i would think... anyway.. heres a couple decent pics of what i'm building... its a pump of sorts.. motor spins cam, rockers rock, pistons go up and down..

the pivot point i was referring to was for the rockers.. in the final design the pivot plate in the picture will be open above the rockers, but in these drawings it looks like the rockers are going through the plate...

hehe, same name yeah? been my nickname for a long time, given by people who have no clue what it is i do in the shop but somehow think i could could make a hang glider out of a poncho and a couple of popsicle sticks. i let them continue to think this.

i see now what it's doing. I'd guess that because the cams contact rollers there isn't that much torque. loctite should be fine - they make several cylinder retaining compounds, better than sandwiching them on all thread with nuts. if you are worried, you could key them to the shaft, this would give a way to control alignment on assembly which will be a tricky aspect of the built up approach

how long is the shaft? on seeing this rendering, i might recant and recommend making it from solid. To make it, you need to take a round bar of stock and (after facing) use the mill the drill sets of centres in each end. use coordinate drilling; work out the math so that you have the x,y coordinates for the centre drill marks. then its a simple job of turning between centres. How do you get the centre drills points indexed the same on each end? this will depend on how long the shaft is, but basic theory is whatever set up you used for the first set, flip it over without disturbing the work to do the second set.

well i dont have a rotary table, cause that was my first thought on how to do the center points, just set it up on the mill with the correct offset, center point it, then turn 90deg.

but i had another thought, sticking with the seperate lobes idea, why not make all the circles, center drill them, and put a dowl pin between each one get them all lined up centered, and then mount them in the lathe off center and turn each one 90deg off from the dowl pin.. that way they stay aligned and i can allways take it back apart if i need to... plus by doing it this way i can add a center carrier bearing to help hold up the middle of the cam if the load tried to flex the allthread...

my question is that if i were to try and key them to the shaft, how would i do that.. i can see cutting a slot in the center shaft, but how do i make the key in the lobes.... oh wait, i think it just came to me, make the keyway in the shaft, then mill a matching keyway on the inside of the lobe, then put a loose key inbetween them half in the shaft and half in the lobe... ok, i'm on track now...

i was given that name in highschool.. long story, but lets just say that i was on a date, lost a fanbelt, used the girls pantyhose to limp her home, only to be met by her father on the front porch with a shotgun cause she left with pantyhose and didnt have them when she got home.. he didnt believe me that there was no funny business till i popped the hood on my old car and he saw the knotted ducktaped pantyhose going round and round.... it wasnt pretty, but it got her home by cerfew...

.....but i had another thought, sticking with the seperate lobes idea, why not make all the circles, center drill them,...etc, etc.

You do not need to worry about center drilling or keyways. You are coming around to the suggestion I made earlier;"What you call the conventional cam should be simple; they look like circles mounted off-center. These you can do in your four jaw, you set them up so the off-centeredness is the same as your cam lift and then bore the shaft hole."

Then you simply slide them on the shaft, align them at 90 degrees to each other and drill holes for roll pins. Or alternatively use setscrews so they can be removed.

If you want to make the alignment easier mill four flat spots on the shaft; this is easy to do in a vise just by milling one flat and then aligning this with the vise jaw for the next and so on. Then use setscrews and your alignment comes automatically from the setscrew sitting on the flat spot.

well i dont have a rotary table, cause that was my first thought on how to do the center points, just set it up on the mill with the correct offset, center point it, then turn 90deg.


a rotary table? u don't need no stinkin rotary table. :D Mr Pythagoras taught us how to do this (but don't worry i won't report you to the Mcgyver society). hold the work in a vice mounted on its side, or a v block clamped and angle plate, using an indicator or edge finder centre the work, figure out the x/y coordinates, drill the the centre holes in the end of the stock. then, leaving the round work still attached to something can be squared to to table (the vice/vblock etc), flip it over, indicated vice/vblock square, find the dia (edgefinder or indicator) and drill holes using x/y again.

now you have centre holes in each end of the stock, perfectly lined up. by successively machine between each set of centres it will be easy to do, no mess no fuss and everything is indexed properly. you just have to puzzle through, based on tooling you have, how to drill this holes in each end such that they are aligned.

If you go built up, like Geof says, you don't have to key them, overkill imo for this app. Were you to key, you cut the internal key with a broach or in the lathe by using it as a shaper if you don't have a press. as i said loctite would work and would be easy, although i like Geof's idea of flats because a challenge with built up is getting it all indexed properly.

..... you just have to puzzle through, based on tooling you have, how to drill this holes in each end such that they are aligned.....

You also have to puzzle through how you machine each cam lobe without chewing into the ones either side. What you would probably finish up doing is using something like a grooving tool, or parting tool, to plunge in very close to the final size then use a left hand tool to turn up to the adjacent lobe on one side and a right handed tool to turn up to the adjacent lobe on the other side. In some respects the turning is more challenging than the center drilling.

good point, I've made a lot of eccentrics so didn't really give much thought but it does add to what a newbie will have work through. still, a piece of cake compared to those three throw 3/8 dia crankshafts with a 1/2" throw and 1/2 between webs and 9" in length :D for those you grind parting tools into left and right hand combination knife/facing tools using magic marker to pick up the cut - too flimsy to for a parting tool so you creep up on it. for this job i think you nailed it with your approach, part, left, right, its looks fairly solid.

one of the things in the back of my mind is that while built up cranks seem really easy to make in your head as you're falling asleep (yes it is a disease), in practice there's just a lot more mating surfaces and hence stuff to go wrong. too much clearance on a hole, not enough on another, tough to perfectly align First time you realize that centre journal is off is the last built up crank you make :) (Where built ups come into there own is on stuff too large to turn from solid, but to overcome the challenges, for a really bang up job, you've got to use shrink fits and a fixtures, a lot more than what we talking here i know) Yet the from the solid cranks I've made, have turned out perfectly, although granted they take a bit more thought up front. I'd recommended built up as well, but with more time on the fence i think i falling to the from the solid side. i know he's not making a crank, but the principal and challenges seem the same

hmm some good points... well i too have that thinking while falling asleep dease, and i had a thought... take my solid.. smooth it along its length, cut cookies off it, then take each cookie, chuck it up in a 3 jaw, smoothe and re round, and face the end, flip it around and finish smoothing and facing the other end.. do this to all lobes, then put in the 4 jaw, setup the off centerdness, put each one in in turn, and drill through it using the tailstock chuck...

i really liked the square shaft and setscrews idea, thats something that i can definately do, drill and tap the thin side of the lobe into the shaft hole.. take a stick of square solid and turn the ends down round on the lathe for the bearings and stuff, and then drill and tap the end to hold on the driving pulley.. and just make sure that the driving pulley is trying to tighten the screw rather than loosen it... i could allways put some locktight or jb weld or something between the lobes if i had to...

ofcourse, if i do go the setscrew route, why not just use some good straight beefy allthread.. i understand the flats advantage, and could mill them into the allthread.. just thinking about trying to use what i have rather than having to buy stuff... and i think that the setscrews would dig into the threads of the allthread and hold in place pretty well.. dont you?

also, the timing on this isnt absolutely critical.. if its off a few deg then i dont see that it would really hurt anything.. just the more accurate it is the smoother the output pulses from the pump will be...

....and i had a thought... take my solid.. smooth it along its length, cut cookies off it, then take each cookie, chuck it up in a 3 jaw, smoothe and re round, and face the end, flip it around and finish smoothing and facing the other end.. do this to all lobes, then put in the 4 jaw, setup the off centerdness, put each one in in turn, and drill through it using the tailstock chuck... ..

Umm....isn't this what I told you to do way up near the top of the thread?

What you call the conventional cam should be simple; they look like circles mounted off-center. These you can do in your four jaw, you set them up so the off-centeredness is the same as your cam lift and then bore the shaft hole.

i think your right, it just took me this long to fully wrap my head around it...

ok, well i have a few other things i need to work on for this comming weekends fishing trip, but maby i'll have time to start on this today... thanks for the input, i'll let ya'll know how it goes...