DIY CNC Camgrinder

It would depend on which shop they used.

They have a custom shop with a bunch of old Berco grinders and thousands of masters. They can set up for a custom cam in 15 minutes or less.

Same for the CNC grinders but they are typically used for higher production.

As with most projects, they are in your head and once you start talking about them, you learn and make some thought changes.

My initial requirements are to copy existing cams. For a single overhead cam, I may want to keep the cam profile and change the lobe centers.

I have a jig to measure existing cams and if I use a follower on the cam the same diameter as the wheel, the angle and X coordinates measurements should be able to duplicate the cam on the grinder.

The original issue I did not have an answer for was, decrementing the X value to grind the cam each revolution . I have 2 proprietary CNC machines, but never used logic in any machining exercise.

My thought are something like
Set K = 25 (Ending diameter)
Set S = 30 (Starting diameter)
While S > K do
G1 Xx1+S A0
G1Xx2 +S A1
G1Xx2 +S A2

---

G1Xx359+S A359
S=S-0.01
END

Or it could be something as simple as changing the absolute X 0 each revolution.

I did like the CAM turning video. Would be great for cast iron but not built up cams with 60 RC.

marsheng wrote:

Or it could be something as simple as changing the absolute X 0 each revolution.

Which is exactly what external offsets will do for you without changing the position of the shaft centre.

I Really don't think any of you have got that this does not need commanding a different axis. It just needs an offset to the x position so that the tool (or grinder) follows the profile of the desired cam automagically as the shaft rotates.

If your goal is to trace for high performance applications, you will need software to do the following:

Smooth the measured points.
If you are VERY lucky, you can use that (but don't plan on it), the acceleration and jerk curves will be a mess 90% of the time.

Software to manipulate the measured points to tame the acceleration curve shape. There is a way to do it with a spreadsheet but it is extremely time consuming, like a day to refine a lobe.

This is why it is best to just get or make a good design software.
A CAD software is not able to do this type of work.

Thanks rodw

I've looked up external offsets but I'm non the wiser.

What would the G-Code look like ?

www.performancetrends.com/ca20.htm

This is a basic software for the measurement and evaluation task.
Notice the measurements are 6 places past the decimal but the acceleration curve is still unstable.

Last time I checked, they don't make a design software.

The best place to get that is from Gordon Blair.
www.profblairandassociates.com/index.html

He has passed away but his associates still support it.
It is very pricey though,
It isn't that difficult to write a software these days compared to how it was when he wrote it.

marsheng wrote:

Thanks rodw

I've looked up external offsets but I'm non the wiser.

What would the G-Code look like ?

I'm no expert on lathes but I think you would program a simple turning operation and allow the external offset change the X axis coordinates to follow the cam lobe profile. eg at TDC of the lobe, you would be the height of the lobe away from the main shaft circle and back on it when the lobe ends due to some algorithm that accepts the rotary position as an input.

I'm not sure but there is a new hal-port pin type in master that may let you feed some external point data into a component.

You should look at the external offset sims in V 2.8 and 2.9

I did a search on Cast iron and epoxy granite and it does stack up very well.

See here

Epoxy Granite

Here is a great one

Epoxy Granite 2

Mnn Now I'm getting excited to get started

marsheng wrote:

Mnn Now I'm getting excited to get started

I can't wait to see that billy kart with a great big lumpy cam!

If all goes well it will be for a 1964 Ducati MK3 250.

My starting point,