270 Offy

[mayhugh1]

A bright red magneto prominently mounted on its front end seems to be a signature feature of the Offy 270. The quarter scale's version is really just a distributor meaning its high voltage will be generated elsewhere. Ron sneaked his voltage into the enclosure through the magneto's 'Points' terminal in order to maintain the illusion of a magneto. This innocuous-looking terminal that would have been at several hundred volts potential in the full-scale engine will ride at several thousand volts potential in the model.

The high voltage's side entry complicates the design of the rotor. A slip ring rather than the simple button contact found in conventional distributors was used to complete the path to the rotor. In order to avoid metal-to-metal wear and potential metal dust that could create leakage paths inside the distributor, the actual coupling to the rotor was accomplished through an air gap. Since this gap is in series with the model's tiny plug gaps, it robs energy from the ignition system.

My first inclination was to retain the slip ring but eliminate the air gap with a spring-loaded carbon brush assembly. After prototyping it, though, I had second thoughts about the inevitable carbon dust I'd noticed around commutators in some of the dc motors I'd disassembled.

Eventually, I decided upon a conventional rotor with its axial contact button. Similar to Ron's distributor, the high voltage is brought into the enclosure through an end-mounted terminal on the cover. However, I modified the cover to accommodate a wire running through a drilled passage between this terminal and the pressed-in button contact. One of the photos shows the voltage path through the modeled cross-section of the modified cover.

Although not obvious from the resulting hue in my photos, the cover was machined from red Delrin. The tower electrodes were turned from phosphor bronze and hard pressed into the top of the cover. The rotor contacts on their ends were exposed after boring the opening for the rotor. The side mounted screw terminal was made up with a threaded insert screwed into the cover. The interconnecting wire is sandwiched under the pressed-in button at one end and soldered to the insert at the other.

Finally, the four high voltage towers were threaded for plastic wire retaining caps similar to those in the original design. Excess material left on the cover's mounting flange will be removed later after the enclosure's details have been finally worked out. Those details depend upon the design of the rotor which in turn was waiting upon a proven cover. With it now in hand, work can continue on the magneto which for the most part will follow Ron's original design. - Terry

 

[petertha]

Very nice, Terry.

I noticed the FS engine mags had very prominent ignition wire tower caps. Any idea what purpose that served? Seems like the ones on aircraft mags of the era were slimmer.

Do you have any tricks for finishing Delrin? Yours looks like it came from a mold. Seems like my attempts just get dull & fuzzy.

 

[mayhugh1]

Petertha,

I'm not really sure. Some of the full scale magnetos had metal enclosures, and taller insulators on the towers would provide needed isolation from ground. Aircraft magnetos probably benefit from all the path length they can get in order to combat leakage due to the effects of lower atmospheric pressure and/or condensation.

I keep Delrin wet with coolant during machining. The cover was cut with an 1/8" ball endmill running 12 ipm at 5000 rpm. - Terry

 

[RonC9876]

Terry: a spark gap in series with the plug can be a good thing. Spark intensifiers were and still are sold and some people swear by them. The additional gap is said to allow the voltage to build before the plug fires. At any rate it is said that the added gap only adds about a hundred ohms more resistance to the circuit. Ionized air is basically a short circuit. All I can say is that as designed, the magneto/ distributor has worked flawlessly over the years and the ignition has been great to 13 thousand rpm. I don’t run the Offy over 10 thou now days. Had other parts fail pushing it just to see how fast it would go!

 

[stevehuckss396]

I have personally seen 10k rpm from Ron's engine. Scary but freaking awesome. The ignition system was flawless.

 

[mayhugh1]

I have personally seen 10k rpm from Ron's engine. Scary but freaking awesome. The ignition system was flawless.

I don't doubt it. I can't resist tinkering though. - Terry

 

[awake]

Incredible finish on that delrin. Looks like something is engraved on the top - maybe the cylinder numbers for each position?

 

[mayhugh1]

Incredible finish on that delrin. Looks like something is engraved on the top - maybe the cylinder numbers for each position?

Yes, the plug wire numbers are engraved on the top of the cover.

 

[mayhugh1]

The body of the magneto was machined from the same chunk of red Delrin used for the cover. One of the photos shows a model of its cross section including the gears and bearings that will be installed later. Some minor changes made to the original design included the addition of a second (outboard) bearing for the input shaft as well as positive stops for all the bearing retainers. After twenty years, the Small Parts part numbers provided in the plans for the 2:1 miter gear set have become obsolete, but Boston Gear's L159Y-G, -P appear to be equivalents.

For the trigger device, I'm using an Infineon TLE4095 Hall effect transistor. It's supported inside a close-fitting cavity on the inboard side of the magneto's body. A machined end cap, that will later double as the input shaft's bearing retainer, holds it securely in place. Testing showed this particular device is reliably triggered through as much as a quarter inch of air by an 1/8" diameter x 1/16" thick Neodymium magnet. A pair of these magnets will be mounted to the end of an adjustable sleeve attached to the magneto's driving shaft leaving the running gap much less than half this amount.

A cosmetic end cap was lathe turned for the outboard side of the magneto. It closes up one of the openings used to access the interior of the body during its machining. The retainer for the upper rotor bearing was turned from white Delrin. After assembly, it will remain sandwiched between the magneto's cover and a machined recess in the body after assembly.

Except for the rotor, whose tip will include a .003" 'spark intensifying' gap, this completed the machining of the magneto's Delrin parts. Its platform on the crankcase front cover was drilled for mounting screws, and the alignment of the magneto's input shaft bore with the right angle drive block shaft was finally verified.

The next steps will be to machine the rotor and to modify and install the stock gears and bearings so the unit can be exercised under high voltage. - Terry

 

[propclock]

Just my 1.414 cents worth(square root of 2) I make many ignitions for
Myself and BAEM members. I really like the Dave Sage IGBT design.
I have shrunk it to ~1"square and use SMD components.
Instead of Hall sensors I really like the use of a simple reed relay.
Not an issue in your case but halls are sometimes placed in the HV
distributor cap and they get zapped, continuously. A reed has no problem with this . The switching current is ~1ma. They are fast !
and draw NO current when not active or in the timeout mode .
And a hall draws ~10ma continuously. Finally if a ground is available,
not in your case, only 1 wire is required , gnd (the distributor) and trigger wire.
Reeds are cheap and almost indestructible, but made of glass.
I usually get mine from Electronic Goldmine but I have some super small ones made for the pacemaker industry. (not readily available)
Just my input.

 

[mayhugh1]

Propclock,
Thanks for your insightful comments.

I built my first two engines with transistor ignitions (TIM-6 variations) but then switched to CDI's because the coil, the energy storage element in a transistorized ignition, just doesn't scale very well and is difficult to hide or disguise. The capacitor in a CDI is a much more efficient energy storage element making the whole ignition considerably smaller than an equivalent transistor ignition and so easier to deal with. Some have mentioned using COP coils, but since they're designed for single plug high energy applications, they probably require a lot of current.

I've personally never had problems with Hall devices even when using them inside distributors. I can remember blowing only one, and that was due to my carelessness. I understand the issues associated with using them, though, and can fully appreciate why many would rather avoid them.

By the way, Doc1955, who occasionally posts on this forum, has built some impressive and realistic looking coils on Youtube. Although they're still a bit large for my taste, he's probably got them as small as is practical. I haven't tried my hand at winding one, because I've never ginned up the interest in dealing with the cumulative corona damage that can occur in the windings. I don't know how Doc's coils are holding up over time, but he may be getting away with just filling their containers with paraffin wax.

I've been purchasing Roy Sholl's CDI units. Even though they're pricey, they appear to be well crafted, and my experiences with them have been positive. - Terry

 

[dsage]

Terry:
Maybe I'm interpreting your drawing incorrectly It looks like the sensor top EDGE is sensing the magnet. According to the spec sheet the face of the sensor with the writing on it is the sensing face. Not to say it won't work on edge (I guess) but...

BTW I looked on Digikey for TLE4095. Not found.
Was it a TLE4096?

 

[dsage]

I think Roy says his CDI modules are rated for about 11,000 rpm on a single cylinder engine. (They are from the model airplane industry). Ive tested them as such. The high voltage is petering out at much higher rpm. The HV power supply can't recover quickly enough to charge the capacitor fully.
That would be only 2750 rpm on a four cylinder engine. Just saying...
Having said that I run one of his older modules on my Howell / Sage V8 and it runs ok, but I never rev it more than about 3,000 for fear of it coming apart.
If you're going to impress folks with the high rpm (ala Ron) you might have problems.

 

[mayhugh1]

...... Maybe I'm interpreting your drawing incorrectly Itlooks like the sensor top EDGE is sensing the magnet. According to the spec sheet the face ofthe sensor with the writing on it is the sensingface. Not to say it won't work on edge (I guess)but...
BTW I looked on Digikey for TLE4095. Not found.
Was it a TLE4096? ......

Dave,
I have the branded side facing the magnet. I double checked the data sheet and that's what it is calling for. I purchased a number of these 4095's during my Merlin build and am not surprised they're no longer available. The market lifetime of Hall sensors seems to be pretty short. - Terry

 

[dsage]

...... Maybe I'm interpreting your drawing incorrectly Itlooks like the sensor top EDGE is sensing the magnet. According to the spec sheet the face ofthe sensor with the writing on it is the sensingface. Not to say it won't work on edge (I guess)but...
BTW I looked on Digikey for TLE4095. Not found.
Was it a TLE4096? ......

Dave,
I have the branded side facing the magnet. I double checked the data sheet and that's what it is calling for. I purchased a number of these 4095's during my Merlin build and am not surprised they're no longer available. The market lifetime of Hall sensors seems to be pretty short. - Terry

Good stuff. I guess I was interpreting the drawing incorrectly. I hate it when you find a good part and it goes obsolete. Sometimes they make it better though. Then you have to find what they changed it to.
Thanks

 

[mayhugh1]

I think Roy says his CDI modules are rated for about 11,000 rpm on a single cylinder engine. (They are from the model airplane industry). Ive tested them as such. The high voltage is petering out at much higher rpm. The HV power supply can't recover quickly enough to charge the capacitor fully.
That would be only 2750 rpm on a four cylinder engine. Just saying...
Having said that I run one of his older modules on my Howell / Sage V8 and it runs ok, but I never rev it more than about 3,000 for fear of it coming apart.
If you're going to impress folks with the high rpm (ala Ron) you might have problems.

Good point on the rpm limitations. It's worth some testing that I'll probably do even though I doubt I'll ever rev mine as high as Ron (assuming my engine actually does run). My understanding of Roy's spec is that it is 'sparks per minute' and not rpm. And so, I think (hope) you're spark calculation is off by a factor of two. - Terry

 

[dsage]

Good point on the rpm limitations. It's worth some testing that I'll probably do even though I doubt I'll ever rev mine as high as Ron (assuming my engine actually does run). My understanding of Roy's spec is that it is 'sparks per minute' and not rpm. And so, I think (hope) you're spark calculation is off by a factor of two. - Terry

Of course your engine is going to run.
Was it sparks per minute? I don't recall. Whatever. it's something to consider. Also I found there wasn't much spark energy so if you have high compression there could be an issue as well. My George Britnell V-twin has about 8:1 and I think the CDI could be contributing to it's poor running. I have yet to try one of my drivers and a full sized coil to see if it runs any better.
Whatever the case you'll work it out.

 

[RonC9876]

Guys: I am still using the old Floyd Carter designed board with a Modelelectric coil and powered by two 2 volt lead acid cells wired in series for four volts. This ignition has proven to be both reliable and capable of firing four plugs to high rpms. The difference with my engine is that I also use an electronic advance board designed by David Bowes to advance the timing as the engine gains rpm. I could never get over 8000 rpm without it. The initial timing had to be set at about 15 degrees advance. If the initial advance was set any earlier than that the engine would backfire and destroy the needle clutch used to start it. It also would hurt my wrist. This engine is powerful. With the electronic-advance board, I can set the initial timing at about 10 degrees, which helps with starting and idling, and then get up to 40 degrees of advance on the top end. I had a lot of fun running at these speeds till parts started failing. I had a gear tooth break off and destroy the whole gear train and I had a crankshaft break in half also. But I learned a lot about ignitions and timing on these small engines. From the beginning I knew this engine had the power to destroy itself. I proved it several times!

 

[dsage]

LOL.
You certainly do push it to the max. Always makes me cringe when you give it "the Beans"
But such a wonderful running engine.
Thanks Ron.

 

[tornitore45]

I have blown several Hall Effect devices until I realized my spark box was connected wrong: feeding the HV to the chassis and the ground to the plug tip.
As a professional power engineer I am ashamed to to confess it.
No more failure after that.
The secret in pacific coexistence of sensitive and power electronic is in understanding and controlling the GND net. Return currents from power elements should never be allowed to run through a wire extended to a low voltage sensitive element.