DIY CNC Camgrinder

Sorry, no progress to report.

Here is a nice video of an OKUMA in action.
I don't know exactly what the following accuracy requirement is, but I guess that it is less than 0.0001",,,maybe a lot less.

I have a Mori Sieki MV45/50 from 1984 that sat in an office building unused for decades. It was bought by a CAD CAM software company with the intention of using it to demonstrate the software but it was only used a few times.
I have been using it in my home shop with the original Yasnac control.

I am in the process of selling my house so this machine (and another 5 axis LINUX CNC machine) have to go. I posted it on Craig's List at $3,000 with no interest. So I am thinking of striping out the screws and drives etc before sending it to the scrap pile. And maybe using those parts for the cam grinder.
Still doubting the sanity of the project,

Yes, that is my website.
I changed hosting providers a while ago but have not refilled the content.

Now that I am selling my house and shop, it is not certain what I will be doing as far as machining goes. Most likely I will get a good CNC router.

I am trying to sell the house with the shop included, ( I would hate to see all the work that I have put into setting it up go to vultures) a few people have been interested bu none have been able to pull the loan together.

yeah grinding is not a roughing process so what ever the tolerances are on the cranks or cams of any particular motor that may be being built is your starting point for accuracy. so if your looking for +/-.0001 your probably going to want a resolution 3 times better than that. so now your into sub micron territory. very high end parts and some serious machines to do the rest of the build.

i personally would be intimidated trying to do something like that in a home shop honestly.

i have surface ground to .00005 and you have to have a controlled climate to work in to even get that close. heck even if you hold a part in your hand to long it will grow in size on you. not something your typical home work shop can even deal with.

surface grinding in general is a completely different animal. when your working with steel the heat will change the part, when you work with carbide it is stable but the diamond wheels grow as they get warm. pain in the a$$ honestly and i got out of it and went back to machining first chance i got.

On cam lobes, the size is not held to a close tolerance, +- 0.001 would be more than good enough.

The issue with accuracy is the shape of the lobe (offset not so critical) and the surface finish. From what I have seen with commercial cam lobe grinders, is that only the very best machines can match the finish of the master/tracer style.

Where they seem to fail is in interpolating the shape with smooth motion,the finish may look like it is faceted.

I'm think of making from scratch and not modifying some other machine. I already have 2 Cylindrical grinders but there is probably more work involved in trying to get them to work than from scratch..

The cam profiles often are not simple calculations when it comes to overhead rockers. Direct acting ie buckets, are reasonably simple.

I have a cam measuring tool and will be taking the data from there.

The previous video shows the grinder only moving back and forward and the bed has the rotary motion as well as moving from lobe to lobe.

I'm thinking of using the rotatory motion on one bed and the other two for the stone. Much like a lathe. If the bed with the camshaft moves from left to right, the working table is 2 x the length of the cam. If the stone moves, it on the 1x the length of the cam.

I do need to cater for different length camshafts.

The rigidity is one of the issues. I'm not sure how well linear bearings will work with grinding. Maybe the preloaded variety would work.

The longest cam I will probably need to grind is 500 mm so it is not a big machine.

I'm not sure that closed loop is required for this as the movements can be quite slow.

The main consideration is calibrating the 2 axis to the cam. Lobe center or TDC and the diameter. The 3rd axis can me manually moved, ie from lobe to lobe.

This is how simple it can be with copy grinding.

My machine runs on linear slides, no problem.

The mathematics for designing a cam lobe for rocker arms or buckets involves the exact same code.

The difference is in simulating flexibility. That used to be approximated by the "Polydyne method" , but no one does that these days.

I can't imagine building a machine from scratch being economical. Casting your first bed will cost at least $20K let alone machining etc. Making a spindle?

There are some beautiful grinders out there, both manual and CNC to use as a platform. There was a Studer that is still on my "one that got away" list. I paid for it and when I went to make arrangements to have it moved he gave me my money back.

I live in New Zealand.. I cant pop down to the local store and pick a second hand one off the shelf. There were very few to start with of over here and as time marched on, they have become very scare. Besides, I only want a 500 mm bed which is big enough for GS1000 cams. .

I am busy ordering linear bearings as most CNC machines uses these today. I'm going to experiment with either granite blocks or, something I have wanted to try for a while, is cast iron swarf in an epoxy. I have a few bucket loads waiting here to try.

Either way I'm sure the grinding can be tuned, if need be, to speeds and feeds outside resonance areas.

PS There was no cast iron beds in the video I posted and they seem to get reasonable results so I will just use a bit of common sense in what I use.

I anyone knows about Polydynes I would appreciate a hand with some modeling.