Ford 300 Inline Six

[kvom]

In thinking about this procedure as well as the fact that my rotary table+chuck weighs as much as I can lift, being able to machine the gear without the rotab tilted would be nice. Since the goal is to be able to keep the tool parallel to the groove at all time, it seems to me that tilting the spindle would accomplish this. Since my mill allows rotating the head around the Y axis, having the work piece oriented in +X should work. Each cut would be a coordinated move in -X and +A.

Not many CNC mills have a rotating head so this idea likely isn't that useful.

 

[pileskis]

Terry- I wonder if you took a video of the gears being cut?
That would be interesting to see.

Thanks, Sid

 

[mayhugh1]

Terry- I wonder if you took a video of the gears being cut?
That would be interesting to see.

Thanks, Sid

Sorry, I didn't. It ran so slow that you'd likely have dozed off. - Terry

 

[mayhugh1]

In thinking about this procedure as well as the fact that my rotary table+chuck weighs as much as I can lift, being able to machine the gear without the rotab tilted would be nice. Since the goal is to be able to keep the tool parallel to the groove at all time, it seems to me that tilting the spindle would accomplish this. Since my mill allows rotating the head around the Y axis, having the work piece oriented in +X should work. Each cut would be a coordinated move in -X and +A.

Not many CNC mills have a rotating head so this idea likely isn't that useful.

That sounds like it would work. Another approach would be a right angle drive attachment for the spindle. Chuck built one for his mill out of a set of gears from a right angle grinder. It its purpose was mainly low speed gear machining, it wouldn't have to be that precise. - Terry

 

[CFLBob]

I'm also using mach3. G1 uses the F parameter only for linear moves. The rotary's movement is done using degrees. You need to insure your rotary is capable of rotating fast enough to honor the coordinated move, however. Mine will spin 360 degrees in 23 seconds and so there was no problem. - Terry

Terry,

Sorry it took me until today to be able to get to my rotary table and double check, but there's something funny here. I did this all from the immediate interface (command line) in Mach 3. I had my rotary axis, A=0.000.

I set F10, and that appeared in the feed rate box.

G1 A90. If F10 is degrees/min, that should take nine minutes and I should visually see 90 degree movement. I stopped it after 1 minute when it had gone 10 degrees. I told it to go back to the starting point, A0.00.

I set F360 at the command line and, again, 360 appeared where it should have.

G1 A90 should take 15 seconds, 1/4 of a minute for 1/4 of a circle. And it does.

So there's some setting that I don't know that's making either yours ignore the F parameter for angles or making mine use it. Or use them differently. If I had to choose, I'd rather mine worked closer to the way yours does. You're able to do coordinated moves between the rotation and the linear axes.

The few times I've used the rotary table under CNC control, it hasn't required coordinated motion between the axes. I could increase the feed rate to like F720 (2.0 RPM), set the table to whatever angle I needed, set F back to whatever it was before and do whatever X, Y, Z movements I needed, .

No need to answer unless you happen to know why off hand. I'll be reading and trying to understand. Just thought I'd close the loop on this.


Bob

 

[mayhugh1]

Terry,

Sorry it took me until today to be able to get to my rotary table and double check, but there's something funny here. I did this all from the immediate interface (command line) in Mach 3. I had my rotary axis, A=0.000.

I set F10, and that appeared in the feed rate box.

G1 A90. If F10 is degrees/min, that should take nine minutes and I should visually see 90 degree movement. I stopped it after 1 minute when it had gone 10 degrees. I told it to go back to the starting point, A0.00.

I set F360 at the command line and, again, 360 appeared where it should have.

G1 A90 should take 15 seconds, 1/4 of a minute for 1/4 of a circle. And it does.

So there's some setting that I don't know that's making either yours ignore the F parameter for angles or making mine use it. Or use them differently. If I had to choose, I'd rather mine worked closer to the way yours does. You're able to do coordinated moves between the rotation and the linear axes.

The few times I've used the rotary table under CNC control, it hasn't required coordinated motion between the axes. I could increase the feed rate to like F720 (2.0 RPM), set the table to whatever angle I needed, set F back to whatever it was before and do whatever X, Y, Z movements I needed, .

No need to answer unless you happen to know why off hand. I'll be reading and trying to understand. Just thought I'd close the loop on this.


Bob

Bob,
Make sure in your General Configurations pull-down menu that your units for the A axis is set for angular. In the toolpath set up pull-down menu make sure 'use radius for feedrate' or 'use diameter for feedrate' (depending upon the version of mach) is checked. Also in the upper left hand corner of the settings screen, enter the radius of the part that's in your chuck. Due to a mach3 bug, a zero in this parameter can cause havoc. - Terry

p.s. mach always honors the maximum motor rates allowed in the setup for all the axes, and so the feedrate parameter F should be considered the max feedrate you'll get for a coordinated move if one of your axes can't keep up.

 

[mayhugh1]

For best visibility while aligning the cutter in my helical gear machining setup, I machined the gears with an opposite handedness compared with those in the full-size engine. The downside of doing this was a reversed direction of the rotor inside the distributor and a backwards order for the plug wires coming out of the cap. A small upside is that the axial thrust created by the driven gear will tend to pull the tiny distributor into the engine rather than push it out.

Fixturing an operation to bore the distributor's mounting hole in the block was a lot easier and much less risky than I had been imagining. After squaring the block on an angle plate, I realized the bore could be accurately located using a driving gear on the end of a rod in the camshaft bore and a driven gear on the end of a dummy shaft in the spindle. The final result was a smoothly meshed gear set with near zero backlash.

The distributor body was machined from 6061 and the distributor shaft from 303 stainless. I replaced the bronze bushings for the distributor shaft with sealed ball bearings to avoid having to manually oil them later. The shaft's long skinny end was turned without using the tailstock similarly to the valves (in a previous post) using short overlapping segments turned one at a time.

A rather sketchy fixture was used to drill a 3/64" hole simultaneously though the assembled gear and distributor shaft for a pin to lock the two together with zero thrust clearance. The pin will be Loctite'd at final assembly but will remain loose for fit testing partial assemblies.

The distributor uses a shuttered type Hall trigger. A machined steel cylinder with six openings rotates between a magnet and a Hall device to provide timed trigger pulses to a CDI. An alternative would be to use six magnets, but this can become tricky inside a tiny distributor where the individual fields of the magnets can interact and reduce the sizes of the flux pulses seen by the Hall device.

A shuttered trigger disk has its own limitations. Even at the ON location in front of an open window, much of the magnet's flux will be shunted away from the Hall device by the low permeability return path offered by the disk. The disk's permeability can be reduced and the flux pulses seen by the device increased by judiciously machining material from the disk. However, if too much is removed, the disk can become saturated and allow flux to pass through even a closed window. Some trial and error is usually required to obtain reliable operation that's also dependent upon the particular Hall device, magnet, and the spacing between them.

The trigger disk was machined from 12L14. It's best to use a soft steel alloy for the disk since hard alloys tends tend to acquire and retain magnetism. I modified the disk design to avoid the angled grub screw suggested in the plans to lock it to the distributor shaft. Instead, the trigger disk attaches to a flange on the distributor shaft with three 0-80 SHCS's. The bolt holes are slotted to allow the disk to be rotated on the shaft for proper timing before being tightened down.

A 1/8" x .100" neodymium magnet was epoxied into a machined aluminum holder that attaches to the floor of the distributor body with stainless steel screws. Hall devices tend to go obsolete very quickly making it difficult to recommend a specific part number. I plan to use up some mystery parts from my junk box on which I've run my preliminary tests. They appear to be marked '738S06L', but I have no idea what the actual part number is or who the manufacturer was. - Terry

 

[cwelkie]

Very thoughtful solution and excellent description using a shutter trigger with a hall effect sensor.

 

[Thommo]

Mayhugh, your skill set is truly excellent. I wish that I was 1/2 the machinist that you are fella

 

[a41capt]

Mayhugh, your skill set is truly excellent. I wish that I was 1/2 the machinist that you are fella

AMEN! That goes double for me!!!

 

[mayhugh1]

Hall sensor mounts were machined from 6061, bead blasted, and painted with black Gun Kote. A length of 32 AWG Futaba servo cable was soldered to each sensor and the assembly epoxied into a cavity machined into the front side of the mount. A 3/64" hole was drilled into the top of the mount for a pin that will later be used to locate the distributor cap. Soldering the sensors and connectors to the tiny cable required a lot of patience, and after coming up with a process I made three more using the remainder of the mysterious Hall sensors in my junk box.

I've standardized on a male Futaba J-type connector soldered on the ends of the sensor cables in my engines. I've always used the same Maxx brand connector that's been readily available from a local hobby store. Since they no longer carry them, I ordered a kit of compatible Apex connectors from Amazon. The first thing I noticed was the gold flashing on the Apex pins was considerably lighter than what I was accustomed to seeing on the Maxx pins. Also, the plastic tangs used to lock the pins inside the connector shells were iffy, and the resulting connections were unreliable. I ultimately trashed the Apex parts.

The Maxx connectors were difficult to find, but I was eventually able to order them directly from the source (MPI). While waiting for them to arrive, I decided I wasn't happy with the just completed 'fat' mounts, and so I machined a redesigned set of four. Since I didn't have any more mystery sensors, I used the Optek OH090U's that had given me so much grief during my Merlin build. Although a bit larger in size, they worked well with the distributor's trigger disk. A day after the Maxx connectors arrived, I had eight tested sensor assemblies - many more than I hope to ever need. A nice feature of George's distributor is that the Hall device is safely located outside the distributor rather than being inside underneath a lightning storm.

Also while waiting on the Maxx connectors, I machined a hold-down fork to lock the distributor to the block. This tiny part was machined from a bit of 303 stainless bar stock epoxied down to a piece of MDF. A screw through its mounting hole helped secure it to the MDF during the final operation that cut it free from its workpiece. - Terry

 

[ajoeiam]

Hall sensor mounts were machined from 6061, bead blasted, and painted with black Gun Kote. A length of 32 AWG Futaba servo cable was soldered to each sensor and the assembly epoxied into a cavity machined into the front side of the mount. A 3/64" hole was drilled into the top of the mount for a pin that will later be used to locate the distributor cap. Soldering the sensors and connectors to the tiny cable required a lot of patience, and after coming up with a process I made three more using the remainder of the mysterious Hall sensors in my junk box.

I've standardized on a male Futaba J-type connector soldered on the ends of the sensor cables in my engines. I've always used the same Maxx brand connector that's been readily available from a local hobby store. Since they no longer carry them, I ordered a kit of compatible Apex connectors from Amazon. The first thing I noticed was the gold flashing on the Apex pins was considerably lighter than what I was accustomed to seeing on the Maxx pins. Also, the plastic tangs used to lock the pins inside the connector shells were iffy, and the resulting connections were unreliable. I ultimately trashed the Apex parts.

The Maxx connectors were difficult to find, but I was eventually able to order them directly from the source (MPI). While waiting for them to arrive, I decided I wasn't happy with the just completed 'fat' mounts, and so I machined a redesigned set of four. Since I didn't have any more mystery sensors, I used the Optek OH090U's that had given me so much grief during my Merlin build. Although a bit larger in size, they worked well with the distributor's trigger disk. A day after the Maxx connectors arrived, I had eight tested sensor assemblies - many more than I hope to ever need. A nice feature of George's distributor is that the Hall device is safely located outside the distributor rather than being inside underneath a lightning storm.

Also while waiting on the Maxx connectors, I machined a hold-down fork to lock the distributor to the block. This tiny part was machined from a bit of 303 stainless bar stock epoxied down to a piece of MDF. A screw through its mounting hole helped secure it to the MDF during the final operation that cut it free from its workpiece. - Terry

I greatly appreciate the detail you offer in your build process.
Such really helps prime my design fu AND stimulates me!

Thank you very much!!!

 

[a41capt]

Tiny, tiny, tiny work!!! Most excellent.

I’d have trouble putting the distributor clamp in place, never mind machining it! Thanks for sharing Terry.

John W

 

[petertha]

Terry, this is a good place to buy leads, pins, plug ends, connectors, kits for DIY & picky RC modelers. I've made many custom leads & harnesses over the years but you may find some other standardized plugs & accessories of interest. I've had their crimpers for many years but I'm told there are decent alternatives if yo stick with reputable tool suppliers.
http://www.hansenhobbies.com/http://www.hansenhobbies.com/products/
Not all braided wire, pins & housings are created equal as you've discovered. It may not make a difference to your particular application but good to know what you're getting anyways. Unfortunately the hobby situation is just getting worse. Some of the Ebay/Ali stuff is OK & some of it is crap. There is no way to distinguish quality from a picture. To make matters worse, there is a lot of disregard (or lack of understanding) for plug standardization one could previously count on. You might see JST or Molex thrown out that have no dimensional equivalent to documented standards or available in a reputable electrical supplier catalog. Its not all bad news, some new plug formats have also emerged, but it takes a lot of digging shed light. Sorry for the tangent, this is a thing for me LoL.

 

[mayhugh1]

Terry, this is a good place to buy leads, pins, plug ends, connectors, kits for DIY & picky RC modelers. I've made many custom leads & harnesses over the years but you may find some other standardized plugs & accessories of interest. I've had their crimpers for many years but I'm told there are decent alternatives if yo stick with reputable tool suppliers.
http://www.hansenhobbies.com/http://www.hansenhobbies.com/products/
Not all braided wire, pins & housings are created equal as you've discovered. It may not make a difference to your particular application but good to know what you're getting anyways. Unfortunately the hobby situation is just getting worse. Some of the Ebay/Ali stuff is OK & some of it is crap. There is no way to distinguish quality from a picture. To make matters worse, there is a lot of disregard (or lack of understanding) for plug standardization one could previously count on. You might see JST or Molex thrown out that have no dimensional equivalent to documented standards or available in a reputable electrical supplier catalog. Its not all bad news, some new plug formats have also emerged, but it takes a lot of digging shed light. Sorry for the tangent, this is a thing for me LoL.

Thanks for the tip. Their 28 gage servo cable looks good also. The 32 gage I used is tough to work with. - Terry

 

[petertha]

Forgot to mention, they sell the plug housings in many other variations like 1x4, 1x2, 4x4 etc. (standard RC plug would be called 1x3). Not sure you have a requirement but something to know about. Potentially neater way to bundle different cable count combinations for other electrical do-dads but using the exact same male/female pins you already have. I found crimping these is a bit fiddly when you first start out, but like most things, gets better with practice.

 

[CFLBob]

Thanks for the tip. Their 28 gage servo cable looks good also. The 32 gage I used is tough to work with. - Terry

With 40 plus years in industrial electronics manufacturing, we called anything finer than about 30 ga. "frog hair".

 

[mayhugh1]

Bob,
Did you ever resolve your rotary's slow speed problem?
Terry

 

[CFLBob]

Bob,
Did you ever resolve your rotary's slow speed problem?
Terry

I haven't tested it. I've been working on the camshaft on the lathe, and some other stuff. I copied off your recommendations to look at next time I have that computer on.

 

[mayhugh1]

I happened across an NOS distributor cap left over from my '72 project truck. I noticed the direction of the cylinder numbers on its top agreed with the rotor direction I'm expecting for my model's distributor. It was pure serendipity, but I ended up cutting my helical gears in the same direction as those in the full-size engine after all.

The model's distributor cap was machined from black Delrin. I don't normally use black Delrin for ignition parts since its color is a result of carbon added to the material which can affect its electrical properties. Natural Delrin has excellent electrical characteristics, but manufacturers don't specify them for the black material. The original Ford cap was black, and since other builders have been getting away with using it in their ignitions, I decided to give it a try.

After machining the cap's top surface, brass inserts were pressed into the high voltage towers. These inserts were drilled to accept some gold connector pins from my electronics scrap collection. The pins were soldered to the ends of the plug wires and will later be covered with rubber boots fashioned from automotive vacuum fittings. A small through-hole allow trapped air to escape during the pressing operations. Machining the cap's i.d. exposed the ends of the embedded inserts which will wind up in very close proximity to the tip of the rotor electrode.

The cap is secured to the distributor body with a pair of spring clips similar to those used in the full-size distributor. These simple looking parts took on a build life of their own. Because of the cap's convex top surface, it was necessary to machine recesses into the sides of the cap for the clips to snap into. Similar recesses were added to the bottom of the distributor.

The starting material for the spring clips was a piece of bandsaw blade. The .030" blade was somewhat thicker than I wanted, but my only other on-hand choice was some .010" spring steel that proved to be too flimsy. After annealing the blade, the clips were formed from a thin strip cut from it. The finished clips were brought to red heat with a torch, quenched in oil, and tempered at 500F. After the shop gremlins made off with their share, I turned the rest of the material into spares.

The rotor was the final part of the distributor and was machined from white Delrin. A grub screw will secure it to the distributor shaft once the distributor is timed. A slot was milled into the top of the rotor for a brass electrode that was secured with a pair of 0-80 flat head screws. The tip of the rotor extends beyond the Delrin to prevent burning of the plastic, and its final length was determined by trial-and-error for near zero clearance with the tower electrodes. A strip of .005" phosphor bronze provides a rotating contact with the cap's center button.

The final assembly was bench tested using a bank of six plugs and a CDI module and functioned as expected. The boots will be added during final assembly. The next steps will include work on the camshaft. - Terry