My first wankel engine 'bits'!

[itsme]

Hi there,

I have now finished machining the eccentric shaft for my engine. I decided to take a chance with this engine and leave the shaft unhardened. We'll see which part wears out first (assuming it actually works...). I also roughly assembled the parts so far to see if everything fits together as it should and everything does. The rotor turns smoothly in the housing, which means my eccentric shaft has the corect eccentricity.

The rotor gear is still causing problems, but I'm sure I'll eventually get there. I decided to make a dividing plate for the Compact 8, but due to the 140mm pulley on the back of the spindle, the plate would have to be bigger than this. This is all fine until you try and squeeze 160mm out of the y-axis on a micro mill - it just doesn't work. I'm now thinking about using a stepper motor to drive the spindle with a timing belt and eliminate the dividing plate completely. With a bit of luck this should work and the positioning accuracy should be good as well. I haven't paid much attention to the small cross slide yet - I think I'll get the dividing attachment done first.

Devin, please let us know if the couplings were the problem on your mill. I can't see it being backlash if it could still make a round circle.

Hmmm, now back to work on the gear...

Regards
Warren

[RotarySMP]

How hard is it to remove the pulley? You don't need it in place for the indexing operation.

A stepper driving the spindle with a timing belt will not be rigid enough to ensure the accuracy of the teeth, besides you would have to remove the pulley to install the timing gear, so you may as well install an index plate.

[devincox]

Guys,
I don't know if this would work or not but, how about if you used a faceplate with the 21 holes drilled in it and then drill some holes to use to mount the rotor on the faceplate , across the 3 apex (Much like you would if you mounted it on the mill)
This way the pulley is not a problem.

Just a thought.

Devin

[itsme]

Hi Mark,

I'll give the stepper idea a miss if it won't be rigid enough. I think I will use a dividing plate, but I will find an alternative gripping method, as I really don't want to mess with the bearing preload. In the meantime, I have started machining some parts for the actual shaping attachment.

I have managed to get a proper dxf file for the 21 tooth internal gear. If you remember originally I had a dxf for a 60 tooth internal gear and that is what I had based my tooth shape on (and therefore my tool shape). I have compared the shape of the teeth from the 60 and 21 tooth gears and it is slightly different. This means I will have to regrind the cutting tool. Fortunately it isn't too serious, as I will only have to remove material from my current cutter - I won't have to start from scratch.

I have also decided that this is going to be the first and last time I will make an internal gear this way. It will do the job for now, but I'm thinking that a purpose built machine would be better, especially if I'm going to be making more of these (which I intend to). Once this engine is done and I've done some other upgrades on my mill, I think I'll have a go at building a 'CNC shaper' specifically for gear, spline and keyway cutting. The idea is that I can set the work piece and tool up and then just watch the gear being cut.

I'll try and get some photos of the eccentric shaft up tomorrow. Once the gears are made, there isn't too much left for me to do...

Regards
Warren

[devincox]

Well,
the couplers were not the problem.... I checked everything I can think of except the step pulse width and I don't know how to calc this for my stepper motors. I sent an email to the guy I bought the system from. I will let you know.

Devin

[RotarySMP]

Pulse width is not a characteristic of the motor, but of the driver electronics. Some drivers need a minimum duration for the signal to stay high before the state change (step) is counted. This can help reduce the effect of interference. You should be able to read the minimum pulse wide in the driver specs.

What about interference? I know on my set up, if I lie the printer cable across the stepper cables ->lost steps!

Devins suggestion of putting the indexing plate on the front by the work is a great idea. To ensure concntricitywith the rotor, you could machine the index plate with a circular relief which the rotor just slides into, then all you need is three clamps and a dodad which grips the bed and holds the indexing pin.

The Swiss gear shaper I saw was like a rotary table mounted flat on a Y axis, with a vertical shaper. There was all sorts of tricky swiss watch gearing to enable helical gearing be cut.

I was looking at the internal gearing in the Mädler catalog, and it looks like it would be a good idea to choose a standard gear and then design the engine around if.

[devincox]

Mark,
How would I go about calculating the pulse width?
I do have my power cord running parallel to the stepper motor cables. I guess this could be causing some interference. I will move things around tonight and see if this helps.

Do you know of a source for internal gears that might fit this application? I would be willing to change my rotor and/or housing in order to not have to cut the internal gears.

Thanks,
Devin

[itsme]

Hi Devin,

If there was a source for an internal gear with a 21mm PCD, then I can assure you that I would be using it. The smallest internal gear that I have come across has a PCD of 30mm. Using a gear that size would make the eccentricity 5mm which gives an engine that is significantly bigger than the SW92. I would guess that the capacity would rise to somewhere around 20 or 30cc.

I haven't contacted any of the companies that make these gears to see if they could do a 21 tooth gear, mainly because I'm scared of what it might cost. If you really don't want to make the gear, then you could always have it wire EDMed. I got a quote to have the basic rotor profile with the gear cut into it for $85 including the material (for 1 rotor or $140 for 2). I kind of figure that this will be my second choice if I really can't do the gear myself, but I think I have solved most of the problems (thanks to everyone's help ). I also don't have $85 lying around to throw at it at the moment.

I've started making the gear cutting attachment (both the shaping and dividing parts). The final setup has ended up being a bit of a hybrid from all the ideas that have been thrown in here. I'll post some photos of this setup as well as the eccentric shaft in the next day or two, as soon as it's done (I know I said e-shaft photos would be today, but they have to wait...).

Regards
Warren

[devincox]

Warren,
Mark just mentioned something about a "Mädler catalog" and I thought he may have found some suitable gears. I'm always looking for the easy way out. (c:

Devin

[RotarySMP]

Devin, www.Mädler.de does have internal gears, but not 21 tooth 21mm PCD. They have a 20mm PCD gear but it is brass and 40tooth, so you couldn't do 3:2 phasing with it.

The smallest steel gear would be a 30 tooth 30mm gear.

In the states you might want to try a local Boston gear distributor.

You can't measure the required pulse width, as this is inherent to the driver design hardware. generally the low cost drivers require no minimum pulse width, as they are not sophisticated, but I think the geckos need something around 5 ms minimum.

[RotarySMP]

Warren,

What grinder RPM, feed and speed did you use for grinding the trochoid? I am having trouble finding a good setting for my Proxxon as a tool post grinder on my lathe.

[itsme]

Hi Mark,

I had the grinder on maximum speed (which is a claimed 32000 RPM, but probably less with the extension). The wheel I used for the finishing pass was 19mm in diameter and the feed speed was about 90-100 mm/min. I found the feed depended a lot on the size of the cut - a big cut had to be fed slower and left a poor finish (probably because the RPM dropped significantly).

I also reground the gear cutting tool today. I was using a 15mm wheel at full speed with a feedrate of 75mm/min - the finish is excellent. Again very small cuts were taken.

The gear cutting attachment is coming along nicely. It should be ready by tomorrow morning and then I can start cutting gears.

Regards
Warren