I've enjoyed working on the magneto, and despite my efforts to stretch out its construction, it's sadly coming to an end. Like the water pump, it's one of the few sub-assemblies that I'll be able to check off as fully tested before the engine is completed.
One of the photos shows its internals just before being installed inside the body. The thickness of the spacer behind the gear on the input shaft was trial/error trimmed to set the pinion depth for smooth operation and nearly zero rotor backlash. The end of the magneto's 3/16" input shaft was machined with the male half of an Oldham coupler. It meshes with the female half machined into the end of the 1/4" shaft sticking out from the drive block.
The white Delrin rotor has a relatively thick brass tip electrode and a .003" running gap between it and the tower electrodes. A flat phosphor bronze spring provides positive electrical contact between it and the high voltage button in the magneto's cover. An 0-80 flat head screw holds the pieces in place inside a close fitting slot in the rotor. The rotor is a snug fit on its shaft with a 2-56 grub screw adding insurance against slippage. A witness mark engraved on the top flange of the magneto's body will be used during final timing adjustments to locate the rotor directly under the #1 tower electrode.
Ron's last post resonated with me. Although I don't plan running above the full-size engine's 6k rpm rating, I'm having second thoughts about the current lack of timing control especially during starting. Even with a modest 10 deg fixed advance, I get occasional attention-grabbing kickbacks when starting my Knucklehead. Not having been through Ron's original development work, I'm not yet as comfortable as he is with the engine's starting system whose teething pains included the nose of a crankshaft that I don't want to machine more than once.
Ron mentioned that his engine's timing is controlled with a board purchased from David Bowes. David designed and sold a PIC-based controller for his EVIC-111 engine published in the final issues of Strictly IC magazine during 2001-2002. As expected, after contacting him, I learned these boards are no longer available although his website is still here:
http://rbowes1.11net.com/dbowes/index.htm
Hardening the controller electronics against ignition EMI was probably no easy feat, and I'm impressed Ron's board is still working after all these years. Although it would be an interesting challenge to design an electronic advance, my last minute decision was to simply add a manual control to the magneto.
This change involved scrapping the already completed end plate and replacing it with one that can manually rotate the Hall sensor 20 (distributor) degrees. This will allow retarding the timing during starting but provide up to 40 degrees (crankshaft) advance while running. The photos show the new advance-able end plate that will now house the Hall device. Since this rather complex part required some involved fixturing, I made several spares while still set up.
I wasn't confident in the friction between the Delrin body and the new advance plate being able to maintain its setting under vibration, and so I added a fixed metal back-up plate between the two. The plate was machined while sandwiched between a couple pieces of sacrificial aluminum, and so there was no effort involved in adding a couple extra pieces to the stack for experiments. The photo shows the piece of .005" thick stainless shim rolled into a beveled washer that I finally used. Although the advance arm can later be controlled by the throttle positioner, an obtrusive non-original linkage would be required.
I didn't do myself any favors with the tiny space that I left for soldering the connections to the sensor. If I were doing it over, I'd make this portion of the arm a little longer. After coming up with a soldering fixture, I installed sensors and cables in all the plates I made and backfilled them with JB-Weld.
The Oldham coupler between the magneto and the gear tower's right angle drive block is concealed by a faux aluminum housing that contains a pair of trigger magnets. Three setscrews lock this housing to the drive block's shaft in order to fix the ignition timing to the engine's camshaft. Preliminary testing showed the magnets triggering the sensor as expected, and so the next step will be to come up with the ignition module so the entire system can be finally tested with actual spark plugs. - Terry
One of the photos shows its internals just before being installed inside the body. The thickness of the spacer behind the gear on the input shaft was trial/error trimmed to set the pinion depth for smooth operation and nearly zero rotor backlash. The end of the magneto's 3/16" input shaft was machined with the male half of an Oldham coupler. It meshes with the female half machined into the end of the 1/4" shaft sticking out from the drive block.
The white Delrin rotor has a relatively thick brass tip electrode and a .003" running gap between it and the tower electrodes. A flat phosphor bronze spring provides positive electrical contact between it and the high voltage button in the magneto's cover. An 0-80 flat head screw holds the pieces in place inside a close fitting slot in the rotor. The rotor is a snug fit on its shaft with a 2-56 grub screw adding insurance against slippage. A witness mark engraved on the top flange of the magneto's body will be used during final timing adjustments to locate the rotor directly under the #1 tower electrode.
Ron's last post resonated with me. Although I don't plan running above the full-size engine's 6k rpm rating, I'm having second thoughts about the current lack of timing control especially during starting. Even with a modest 10 deg fixed advance, I get occasional attention-grabbing kickbacks when starting my Knucklehead. Not having been through Ron's original development work, I'm not yet as comfortable as he is with the engine's starting system whose teething pains included the nose of a crankshaft that I don't want to machine more than once.
Ron mentioned that his engine's timing is controlled with a board purchased from David Bowes. David designed and sold a PIC-based controller for his EVIC-111 engine published in the final issues of Strictly IC magazine during 2001-2002. As expected, after contacting him, I learned these boards are no longer available although his website is still here:
http://rbowes1.11net.com/dbowes/index.htm
Hardening the controller electronics against ignition EMI was probably no easy feat, and I'm impressed Ron's board is still working after all these years. Although it would be an interesting challenge to design an electronic advance, my last minute decision was to simply add a manual control to the magneto.
This change involved scrapping the already completed end plate and replacing it with one that can manually rotate the Hall sensor 20 (distributor) degrees. This will allow retarding the timing during starting but provide up to 40 degrees (crankshaft) advance while running. The photos show the new advance-able end plate that will now house the Hall device. Since this rather complex part required some involved fixturing, I made several spares while still set up.
I wasn't confident in the friction between the Delrin body and the new advance plate being able to maintain its setting under vibration, and so I added a fixed metal back-up plate between the two. The plate was machined while sandwiched between a couple pieces of sacrificial aluminum, and so there was no effort involved in adding a couple extra pieces to the stack for experiments. The photo shows the piece of .005" thick stainless shim rolled into a beveled washer that I finally used. Although the advance arm can later be controlled by the throttle positioner, an obtrusive non-original linkage would be required.
I didn't do myself any favors with the tiny space that I left for soldering the connections to the sensor. If I were doing it over, I'd make this portion of the arm a little longer. After coming up with a soldering fixture, I installed sensors and cables in all the plates I made and backfilled them with JB-Weld.
The Oldham coupler between the magneto and the gear tower's right angle drive block is concealed by a faux aluminum housing that contains a pair of trigger magnets. Three setscrews lock this housing to the drive block's shaft in order to fix the ignition timing to the engine's camshaft. Preliminary testing showed the magnets triggering the sensor as expected, and so the next step will be to come up with the ignition module so the entire system can be finally tested with actual spark plugs. - Terry
Last edited:
