Multi cylinder ignition systems

[Roger B]

Edit Jo: The following has been extracted from Vixen's Mercedes-Benz W165 Grand Prix engine in 1:3 scale thread so you can discuss multi cylinder ignition systems. Roger's was the first item from Mikes thread:

Small ignition systems are an interesting problem, as well as the voltage withstand you also need an energy store. This is the iron core in an ignition coil or magneto and the capacitor in a CDI system. There will be a limit on how small you can make a coil or magneto as there is a minimum energy is required to ignite the mixture.

[BillTodd]

I wondered if one could use one of those modern car type plug-cap type ignition coils with appropriate driver?

[Vixen]

Thank you Roger and Bill,

You are quite correct. There is far more to small ignition systems than just the voltage. If you and our other members wish to start a discussion regarding small ignition systems, I will open a separate topic, which will be easier for everyone to find, rather than hidden with this dedicated engine build log.

Bill, the modern COP (Coil over Plug) ignition system and multiple CDI units were developed to eliminate the need for the spark distributor completely. I am sure modern COP and CDI systems will provide fertile ground for those interested in their development. A separate small ignition system topic would be a better place to discus their design and development

Cheers

Mike

[petertha]

There is far more to small ignition systems than just the voltage. If you and our other members wish to start a discussion regarding small ignition systems, I will open a separate topic, which will be easier for everyone to find, rather than hidden with this dedicated engine build log.

Bill, the modern COP (Coil over Plug) ignition system and multiple CDI units were developed to eliminate the need for the spark distributor completely. I am sure modern COP and CDI systems will provide fertile ground for those interested in their development. A separate small ignition system topic would be a better place to discus their design and development

I would welcome a dedicated post on this subject, specifically (my own selfish interest) distributers in conjunction with CDI. Project examples, do's & don'ts, hall effect triggers, wiring, plugs, ancillaries, advance/retard considerations, supply sources.... Of course coils are another way to go, so a great place to discuss differences. There seems to be more electrical hardware options and/or supplier changes in recent years. I have harvested a few designs & tips from various posts by others with more spark knowledge & practical experience as they come up on specific builds or ignition debugging, but it would be great to consolidate more details & collective wisdom. Special mention to Mike (Vixen) who was kind enough to share some pictures & overview of his CDI/distributer via PM of just such a retrofit. Replicating that for my purposes is on my near term to do list. (Hope you don't mind me showing your nice work as example.)

[Vixen]

As Jo said in the opening post, we have opened this new topic, titled 'Multi cylinder ignition systems'. : https://www.modelenginemaker.com/index.php/topic,12062.0.html

Here is the place and opportunity for our MEM members to discus anything and everything relation to multi cylinder ignition systems. This can include, but not limited to: magnetic Hall sensors types and configurations, points, ignition coils, CDI ignition, magneto ignition, batteries, spark plugs wires and connectors. Also distributor design and alternatives to the distributor.

Over to you   

[BillTodd]

Ok, coil on cap  devices.

My thought was that these are reliable , small  and cheap (less than a tenner on eBay) , so could be used as a single coil  for a model engine.

I don't know the exact driving requirements, but it's probably a capacitor discharge circuit , which could be made quite small and efficient.

[Jasonb]

I've recently bought one of these which are cheap and will work on a couple of AA size batteries. at 60mm long excluding the lugs they are small enough to hide.

https://www.ebay.com/itm/203060489106

Graham playing with his

<a href="https://www.youtube.com/watch?v=3_U38qnyTC0" target="_blank">http://www.youtube.com/watch?v=3_U38qnyTC0</a>

And working on an engine, simple mechanical contact so no need for hall sensors

<a href="https://www.youtube.com/watch?v=lxbC2WvwXA4" target="_blank">http://www.youtube.com/watch?v=lxbC2WvwXA4</a>

[Jo]

If you buy them directly from china they are even cheaper: https://www.aliexpress.com/item/1005007271445496.html

Jo

[Jasonb]

Not by much once you have added the delivery and tax, comes out to £6.47

[Jo]

I paid £7.16 inc VAT for 2 just now (with 100 crimps I needed for £1.36 to take it over the £8 for the free 8 day delivery). If they don't turn up in 8 days I get £1 off my next order.

Jo

[Admiral_dk]

coil on cap  devices

Bill  - the Devil is in the Details 

Most modern Coil on cap devices - actually contains the complete 'Power Section' ....
So you apply +12Vdc., Ground and the 'Firing Signal' from the ECU and 

The more advanced ones even has a build in surveillance system, that can tell the ECU when all is NOT good, lower the Current through the Primary, if the RPMs are too low - etc. ...!

Oh and I have never heard of them as a CDI system internally ....

Per             

[Roger B]

A lot of modern engine management systems monitor the coil current or voltage to determine if a misfire has taken place.

If my four stoke twin gets too lean it flashes over in the distributor between the rotor arm and the cam.

[Vixen]

Using a resin printer to make a distributor cap. Some questions and answers




Photos and answers by Bill Todd

a)  What is the proposed plastic material? - Used aboive was elegoo ABS-like resin (there are many types)  . It has a texture very similar to fine car body filler and will harden more with time.

b)  What is the insulation resistance of the proposed material? You did tell me but I cannot now find that piece of information. -Non conductive, breakdown supposed to be about 35kV

c)  Can the printed material be polished, either chemically of mechanically? This may be important to prevent surface tracking.  - I could coat it in liquid resin then cure to give it a shine. It acts like a lacquer when painted on. Or, you can paint.

d)  The inside surface finish is probably more important to prevent surface tracking. I have designed in anti tracking fences on the inside to increase the distance between the cable contacts and the metal case but the internal and external surfaces will still need to be wiped clean regularly.
Mineral oils don't seem to bother it, It does soften with temperature , but doesn't melt. Might be worth looking for a high temperature resin.

e)  Is the print 100% solid or does it contain air voids? 100% solid

f)  Is the printed material easy to drill? Yes and can be threaded - works best with IPA lubricant

[gbritnell]

 I know someone started a thread on the subject but I couldn't find it so I'm starting this one. This will be a comprehensive, in depth thread about all the facets of building an ignition for a multi-cylinder ignition system. That being said it will be for an engine of two but in most cases more than two.

 A little back story as to how I got where I am with ignitions. I first started building I.C. engines with casting kits for single cylinder hit and miss engines. This would have been in the mid 70's. All of the engines had a points type of breaker, sometimes a commercial set of points and other times a fabricated breaker. My first adventure into multi cylinder engines was when I designed my inline, OHV, four cylinder engine.
This would have been in the late 70's so there was no internet, or CAD programs, at least not for me. My engine was created 'on the board' with mechanical drawing tools. I had a subscription to Strictly I.C and Live Steam so reading on how others did things was the greatest reference.
 
 The multi-cylinder engines that I saw and read about at that time all had a distributor of some sort and the ignition used breaker points. The Elmer Wall engines and engines from English designers generally had points mounted outside of the distributor body due to their size. Although they worked there was only one engine that I knew of that had points inside the distributor. This was the Challenger V-8 which was marketed by Coles Power Models out of California. With the points being internal that made the distributor oversized in comparison to the scale of the engine but at least it looked more prototypical.

 The criteria for my four cylinder engine was to make a more 'scale' distributor with internal points. I came up with something dimensionally that I thought would work. It became fiddley with the small parts and spring but it worked and is still on the engine today.

[gbritnell]

 The point arm was made from a piece of electrical panel board material. I don't know what is is called but it was a fiber reinforced phenolic material. The points came from a set of motorcycle ignition points. When I looked at the points I could see what looked like a braze line between the tungsten and point arm so I figures I would just heat it until the fusing material melted and the point would fall off. Wrong! I got it red hot and the tungsten never budged so I cut the end of the point arm off and silver soldered it to a threaded post. One affixed I then ground all the old arm away and ground the point down to the size I needed. You have to remember I had no internet that you could just look up 'tungsten'. Everything had to be made by hand by the best means possible.

 The cap was made from a material that Coles sold for the Holt engine. It was called Garolite. I had bought a piece of it and used it for both the cap and rotor. The wires were sourced from some high voltage lead wire that Radio Shack sold.

 As I move forward with this thread I will present pictures and explanations of all the ignition components I use today.
https://www.mcmaster.com/products/phenolic/shape~rod-and-disc/multipurpose-flame-retardant-garolite-g-10-fr4-rods/

[gbritnell]

 Since those early days in my I.C. career I have learned so much from fellow modelers, The internet and just by doing and experimenting.
 At this point I will move on to the techniques I use today on most of my engines.

 The distributor body: Made from aluminum with a bronze plain bearing pressed into it for the distributor shaft to ride on. My shafts are generally .125 diameter drill rod. It's dimensionally accurate and wears well running against the bronze bearing with minimal lubrication. Attached is a drawing of the distributor for my flathead V-8 engine. 
 In the distributor cavity I use a single magnet for my Hall trigger. The magnet dimensions are .125 diameter x .06 thick. I machine an angle bracket with a hole for the magnet and a hole for a 0-80 socket head screw. Some builders use multiple magnets mounted to the timing disc but I have found this to work and it simplifies the build.

[gbritnell]

I'll step back at this point and tell the story of the ignition on my 302 V-8 engine., Trying to follow the same criteria as the 4 cylinder engine I wanted the 302 distributor to be somewhat scale. Here again this was way before Hall triggers and the information highway. I designed a set of points that would fit into basically a .93 I.D. distributor body. Mechanically they worked but electrically they failed. Here's why. The four cylinder engine naturally only has 4 flats on the small rotor and the V-8 has eight. The four has enough dwell time to charge the coil but the small difference in the O.D. of the rotor and the flats provide very little dwell. At this point I'll briefly explain dwell. Dwell is the time the points are closed which allows the ignition coil to charge. The V-8 would start and idle but wouldn't rev up because there wasn't enough charge time on the coil. I finished the engine in the late 80's and it sat on the shelf for many years until I learned about Hall triggers. Once I created the Hall distributor the engine has run great. 
The attached pictures show the 'point setup' for the 302

[gbritnell]

The next piece is the trigger wheel. The first one I ever made was from brass. When I hooked the ignition to the distributor the Hall turned on but rotating the disc provided no intermittent triggering. I thought the brass would work but what happened was the magnetic flux just passed through it so no triggering. I then made one from steel. Problem solved. Using the .125 diameter magnets I cut windows in the disc about .14 wide. This seems to give adequate dwell without overheating the coil and ignition. The steel trigger wheel absorbs the magnetic flux and only allows the Hall to be turned on when the window appears.
The wall of the disc is about .025 thick. The disc is turned on the end of a piece of stock then moved to the dividing head where the windows are cut. The stock is then put back in the lathe to be parted off. It's then turned around and faced off to the proper thickness. The trigger wheel has a tapped hole put in at an angle through one of the windows on the top. In my case it's a 1-72 thread. .073 diameter. This allows the disc to be timed properly relative to the piston position.

[gbritnell]

I make my Hall holders from aluminum. They are mounted to the side of the distributor body with 2 0-80 socket head screws. Over the years I have heard and read about people's experiences with Hall transistors. Don't mount them in metal. Don't use super glue. Don't do this and that. The Hall holder on the 302 engine is aluminum as pictured. I super glue the Hall into it. The engine has over 40 hours of running time on it and I have only ever burned out one Hall transistor. The biggest precaution when using a Hall trigger is to ground the engine back to the negative of the battery. When I run my engines I use 2 leads back to my ignition box. Just to be safe. If the spark can't get to ground properly it will go through the Hall negative and burn it out in a heart beat.
 Since I got a 3D printer I have made several plastic holders. They are identical to the aluminum ones but just made of plastic. They are on standby just in case I burn one out at a show.

[Jo]

This is really useful George 

The point arm was made from a piece of electrical panel board material. I don't know what is is called but it was a fiber reinforced phenolic material.

There were two common materials used for electrical boards that we use in our model making Tufnol and Paxolin:

Tufnol is a Phenolic resin bonded fabric material, recognisable by the fabric weave pattern in the surface and cut edges. Paxolin, also known as SRBP, is a fireproof, brown laminate sheet made from layers of paper soaked in a synthetic resin or Synthetic Resin Bonded Paper ="SRBP".

So George your fibery material you used was Tufnol

Jo

[propforward]

Tufnol is a brand name - when I was in engineering in the UK it was THE brand for this material.

https://tufnol.com/

There are other brands of very similar materials - and indeed a wide range of combinations of fibers and resins to achieve different things.

But more importantly and not to get off track, this is an excellent reference thread, thank you George for taking the trouble to write it out and post it for everyone’s benefit.

[Michael S.]

A material made from phenolic resin reminds me of my first car. (Most people called it a spark plug with a roof.) The phenol was pressed into a mold with cotton.
All body parts were made of this material. But for me it was the first car I owned in 1988.

Michael

[propforward]

Great picture, Michael.

[gipetto]

mrwolf on youtube has a lot of diy cdi examples. i've been meaning to make one because they're quite simplistic, and through hole construction.

<a href="https://www.youtube.com/watch?v=QRFsLe4-erM" target="_blank">http://www.youtube.com/watch?v=QRFsLe4-erM</a>

[gbritnell]

Next up in line would be the rotor. I have made them from several different materials, Corian, Delrin, Garolite and the material Jo mentioned, Tufnol. It's a simple part but several precautions should be taken. The conductor strip and retaining screw or rivet should be high enough above the timing disc so that the spark won't jump down to it. I usually try to make this dimension at least .08-.100. Another consideration is the hole in the bottom for the mounting post. I found that on one occasion I had gone a little too deep with my hole and the spark jumped through to the shaft. I drill the hole first with a conventional twist drill then I counterbore a flat bottom by using a drill of the same diameter that has been hand sharpened with a flat bottom. Similar to an end mill. As with the timing disc I tap a 1-72 threaded hole through the side to secure it to the shaft. For the conductor strip I use Phosphor Bronze sheet material, .010 thick. I cut a groove in the top of the rotor so that the strip will fit into it. That way it only needs to be secured at the tip. For this I use a 0-80 brass slotted screw. The height of the screw can be filed down until the rotor can be turned without hitting the terminals inside the cap. You want a minimum clearance between the rotor and the buttons in the cap.

[gbritnell]

The next part in the chain is the distributor cap. I have only used two materials for these, Delrin and Garolite. The cap is turned as a disc to the overall height with the inside bored out to the required dimensions. As with most of my caps the O.D. is 1.06. The disc is parted off and then faced to the required height. The next step is important. The terminal posts will need to be machined. To do this the cap will need to be held with a chuck like tool. I have a dividing head made from a chuck I had from my very first lathe, a Unimat. I mount this in my mill vise to hold the cap. It's important to make a small disc of metal to fit snugly inside the cap. Tap a hole in the center so it can be withdrawn from the cap when done. The disc is clamped securely and the center found. To cut the terminal posts I have made several annular cutters, single, double and triple fluted. They are made from drill rod and hardened. I drill the hole in the center first. This will be a little smaller than the terminal diameter. I then bore the hole with 1 degree reverse taper. The flutes are then cut on the mill and the edges honed sharp. I first drill and counterbore the holes for the wires. For my radial cap the holes are .141 diameter. I then drill the center hole for the brass terminal. This is .047 diameter. the annular cutter is inserted and the posts are cut. After cutting the posts there will be material left between the posts so it will need to be removed with a small end mill.

[gbritnell]

When I made my first caps I made a small brass post that fit inside the wire hole. It had a short .047 diameter stem on it that would protrude into the inside of the cap. A small brass washer was made with a .047 diameter hole that would fit over the stem. To assemble the terminals I would insert a steel rod into the wire hole up against the brass post and then rivet or solder the stem to the washer. The washer is always left thicker than the finished dimension so that after assembling the cap can be put in the lathe and the terminal cut to the proper depth. The clearance between the rotor and cap is critical for good spark transfer. When the wires were inserted into the cap they would touch the brass post in the bottom of each terminal. With the possibility of a wire becoming loose I came up with another way of making the terminals. Rather than just insert the brass post in the bottom I now make a brass insert that is the full depth of the terminal hole. I first drill and counterbore it with the flat bottomed drill as was mentioned earlier. I then turn the O.D. to whatever size is required. I make them with a .012-.015 wall thickness. Similar to the first method I make a post with a stem that is inserted from the inside of the cap and this is riveted in place to hold the brass insert. With this method should the wire move a little it will still be in contact with the brass sleeve.

[gbritnell]

These are some other examples of caps I have made.

[gbritnell]

The cap will also need a way to be located relative to the distributor body. I have done this two different ways. One is when making the distributor body to leave extra height on the wall then mill it down leaving a small projection on the wall. The cap will  have a corresponding notch cut in the side. The other method is to drill a small hole in the side wall of the distributor body to accept a pin. The cap will have a notch cut to locate over the pin.

[gbritnell]

The cap will also need a way to be secured to the distributor body. The methods I prefer which follow full size practice are to make holding straps, one type flat and the other from small gauge wire.
I don't throw away many materials that I can possibly use for model work. This is the case with what I use for straps to hold the cap down. On some windshield wiper blades they have a thin stainless steel strip on each side of the rubber wiper. Over the years I salvaged some of them and cut strips from them to make hold down straps. I bend the one end slightly greater than 90 degrees. This will be on the top of the cap. The other end I bend with a double curve. This allows it to slide under the body of the distributor and fits into a pocket that I hand grind for this purpose. The other is to use small gauge wire that is bent in a similar fashion. I use the wire type for caps with 8 terminals and not enough room for the strap type.

[gbritnell]

Wires: I mentioned early on that I used test lead wire. That was before Roy Sholl (S@S engines) and Jerry Howell. Roy sold 2 sizes of wire, .100 (10KV) and .125 (20KV). The wire is black. Jerry Howell sells a white wire .125 (15KV). I'm sure there are other sources available. The peculiar thing about this wire is even with the wires in close proximity to each other and metal parts on the engine I have never visually seen a spark jump through the insulation, even given the small diameter.
For the distributor end I place the wire in the cap and slide a short piece of shrink tubing over the wire and up against the distributor cap terminal. I then shrink it in place. I then take a larger piece of shrink tubing, one that will fit over the terminal post and long enough to cover the other piece of shrink tubing and shrink it in place. This provides a boot to hold the wire in place and insulate it at the same time.
For the spark plug end I use a commercial part made by Dorman. It was brought to my attention by a fellow modeler, Ron Colonna years ago. It's a vacuum elbow. I trim it down on both ends. I make a brass insert to go inside to hold the wire and snap over the plug. (Drawing attached)
There is a set screw in the center so that when the wire is inserted it can be tightened to hold the wire in place. These caps and boots will fit the Rimfire type plugs. For larger plugs and home made plugs I use a screw on type terminal. I make up my own wire terminal ends and solder them onto the wire.

[petertha]

I know someone started a thread on the subject but I couldn't find it so I'm starting this one. This will be a comprehensive, in depth thread about all the facets of building an ignition for a multi-cylinder ignition system. That being said it will be for an engine of two but in most cases more than two.

Hi George. Yes I believe this thread got initiated as a dedicated topic offshoot from some ignition posts in Mike (Vixen) engine construction thread, so you are definitely in the right spot. Fantastic pics & description, thanks! Actually I'm a bit late to my own party because I was going to upload a generic design I've been plunking away on, in order to solicit comments. Now I'm going to re-read your material in conjunction with some contributions others were kind enough to share. I SHALL ULTIMATELY BUILD A FUNCTIONING MULTI-ENGINE DISTRIBUTER!! haha

[RReid]

The postings George has contributed are a very useful and practical collection of information and experience, and directly relevant to my current project.  I now think I can “downsize” my initial distributor design significantly.

Thank you very much for all that good stuff, George.

[gbritnell]

Spark plugs:
The last link in the ignition chain but undoubtedly the most important is the spark plug. The plug for model engines was developed in the late 20's. AC, maker of auto plugs was one of the first, later followed by Champion. My first introduction to I.C. engine was with the hit and miss type. I bought castings for one and it came with a Remco spark plug. I couldn't find much history other than they made engines and spark plugs starting in the early 50's. Picture #3
The next plug, a Stitt, Picture #2 was made by the Stitt company which made commercial plugs for large stationary engine.
The next pictures are of 2 Champions, a 1/4-32 and a 3/8-24. These were made by the Champion spark plug company up until the advent of the glow plug which killed the market for them. Late 40's
AC also made model engine plugs. The one shown is a 3/8-24
In the mid 90's Paul Knapp started marketing his own line of miniature spark plugs under the name of Rimfire. (last picture)
The NGK CM-6 is available but only for larger hit and miss type engines with a short reach.
When I got into into I.C. engines, late 70's to early 80's I only had access to the Remco plugs but being as they were intended for model airplane use the reach was quite short so if you needed something longer than .187-3/16 you had to make your own.
My first plugs used Teflon for the insulators. There are two types of Teflon, mechanical and Virgin. I found the mechanical material wasn't as homogeneous as the Virgin and would break apart when machining. I found that as the diameters got smaller the Teflon would flex so the first thing I would do was to drill the hole for the electrode and push a piece of wire in to stiffen it up.
Years later I heard that some fellows were using Corian so I tried it and that is what I use today. It has good heat and electrical properties and machines well.
There are machinable ceramics but the material is expensive and fragile. I was given a piece of it years ago but it's still in my box. I never tried it.
Now with the advent of the Chinese makers getting into everything you can buy their miniature plugs.   

[gbritnell]

I have made spark plugs from 1/4-32 thread size all the way down to 6-40. They all work but with the tiniest ones you have to be careful with your fuel mixture as any richness will foul the plugs very easily.
The last 3 pictures are of the NGK CM-6 plug.

[gbritnell]

When I make my 1/4-32 plugs I came up with an idea to be able to reuse the plug when the center electrode wear down. I drill and tap the top of the insulator for a 3-48 thread. I turn up a brass fitting with the #3 thread on one end and a 0-80 thread on the other for attaching the plug wire terminal. When the electrode wears down I simply unscrew the electrode wire, heat it to soften the solder and pull the wire out a little. I then regrind the length to the proper size.
Attached is the drawing for the spark plug for my radial engine.
I turn the Corian to a close slip fit on both the smaller tip extension and the body diameter. I don't worry about gluing the insulator in. I leave a small ring around the top of the spark plug body and after inserting the insulator I crimp the ring over with a simple pressing fixture. 

[petertha]

George, thanks so much for this info. I have lots of questions but I'll start here. I'm practically electronics illiterate, so bear with me. I have seen distributers like this sketch. They contain a disc attached to the shaft which holds small magnets, one per cylinder. When the Hall sensor sees the magnet it tells the CDI/ignition module to fire. That event aligns to the rotor position much like your 'upstairs' assembly. Energy comes down the center wire, along the rotor arm conductor & out to a cylinder wire.

But your system (if I understand) has slots, no magnets? So is the sensor seeing the presence of metal 'frames' between the window cutouts & that signals the module? I've heard the term 'shuttered' trigger, is maybe what this refers to? If so, is the timing dictated by the edge to edge of the metal portion? If so, what type/class of hall effect sensor would a person be looking for because this would seem to be quite different than a magnet sensor? If I'm already way off base then stop me now, otherwise what are the pros & cons of this system vs magnet sensing?

If I wanted to drive the ignition use a CDI module like S/S or others (as opposed to a coil) would your system work as described? Or is the module designed to expect a particular kind of signal in terms of on/off & that must be considered beforehand?

[gbritnell]

Actually if you will look back to reply 15 you'll see that I do have a magnet, but just one. The shutter wheel exposes the magnet to the Hall sensor and turns it on. As far as ignition modules I use 3 types, an S&S, a Jerry Howell and a Sage Gedde. They all work fine with my design.

[petertha]

Sorry George, I misunderstood. So the shutter wheel / openings would be orientated & rotating something like this?

- Is the hall sensor specific to the S/S kit? Is there a preferred offset distance or that's in the specs?
- Can you elaborate on the Sage Gedde system? I think I saw it mentioned on HMEM. Maybe I'm confused with something else, I thought it was a design requiring making a board from scratch?
- Have you tried any of the Chinese CDI modules like what Stirling offers for tabletop engine kits?
- I recall some discussion, maybe on one of Terry Mayhugh's engines? on HMEM, where the 'normal' S/S module had some recommended discharge limit where you had to factor cylinder count and rpm. If it approached or exceeded some amount then S/S recommended a different module. I don't want to put you on that spot speaking to that product, but did you go down that path with your V8?
- and to confirm, in the systems you have used, none of these have advance/retard integrated into the module like (to the best of my knowledge) RCEXL type CDI systems for RC engines? You are accomplishing this by swiveling the distributer body?

[petertha]

I forgot to ask, what are these 2 openings for on the shutter wheel?

[gbritnell]

Hi Peter,
I'll answer the easy one first. The two windows in the top. One is to access the set/Grub screw and the other is just for balance. More could actually be added.
The Hall sensor isn't specific to any particular ignition module. I don't have the model number handy for it but I will look up my last order. There is a supplier out of England, Minimag, that sells ignition components. The Sage-Gedde ignition module is one that was developed and improved by the named designers.  They posted a schematic for it. It is a robust module with safety features built into it. The drawback is that you would have to build it.
My 302 has 8 cylinders and will rev to 7,800 rpm with all three ignition modules , although it starts to run out of breath with the S&S. By that I mean at the higher rev range it's just not as clean as the Sage module. In all fairness the Sage module will operate on 12 volts whereas the S&S is 6 or less.
I bought 2 of the Chinese ignition systems,  one for a single fire and one for a V-twin withe a waste fire. Initially I found that they triggered as soon as the Hall sensed a magnetic field whereas the others trigger when the magnetic Flux is cut off. Similar to a point type ignition.
If you have any other questions please feel free to fire away.

[petertha]

You show your spring plate type contact for the center terminal. Did you determine this was necessary (as opposed to air gap like the cylinder contacts) based on specific arcing problems, or just good practice? Is phosphor bronze chosen for friction or electrical reasons?

I see some distributers (well Terry's in particular) position the cylinder terminals within the cap in order to corner notch them, I think to promote spark direction. It looks like you use flat ends & your overall cap diameters are smaller yet. I'm wondering if its related to air gap distance. You mention keeping the gap distance small - do you have a recommended dimension?

I thought when I make my distributer I would like to bench run it. Is there anything to see or measure in this regard other than number of healthy sparks should equal number of cylinders? haha

[gbritnell]

As far as the terminal design it could go both ways, notched or flat.  The important thing is having the smallest gap possible.  This would mainly depend on the accuracy of the machining.  As far as the Phosphorus Bronze center srip  I just follow automotive practice. One less air gap the better.  Plus the Phos Bronze is springy compared to regular yellow brass.

[Vixen]

I used a similar shutter wheel, one magnet, one Hall sensor system as the CDI/coil trigger on the two 1/3 scale BTH SC-9 cylinder magnetos fitted to my big Bristol Jupiter engine. You can find the details of the design and build of this shutter wheel trigger in post #243 of my "By Jupiter" build log. https://www.modelenginemaker.com/index.php?topic=7559.200
Post #243 is towards the bottom of the page, so you will need to scroll down a long way.

I built this  test rig To test the magnetos and to confirm they worked with a selection of CDI units and traditional ignition coils. Two turns of the handle (simulated crankshaft motion) produces nine fat sparks in that circle of spark plugs. Like George, I discovered some CDI and ignition coils triggered on the leading (rising)edge of the trigger pulse while other triggered falling edge of the pulse.






You will notice the left hand distributor cap and trigger unit cover have both been cut away to let me to see inside the unit and allow the position of the rotor arm and trigger point to be set accurately. Once set, the cutaway items are replaced with working components. I found three pin JR  plugs/ sockets worked well.JR plugs will be familiar to those who also dabble in Radio Control.  I included the newest Bosch 8 cylinder Magneto for the Mercedes engine in the above photo, for a comparison of size.



Cheers

Mike

[Roger B]

A lot of great information     

Rather than a spring contact on the rotor like George I chose to have a sprung centre contact on the cap and a flat rotor contact. The build is here:

https://www.modelenginemaker.com/index.php/topic,7883.150.html

[Vixen]

There are many ways to make spring contract between the distributor cap and the rotor arm contact. Spring arm or sprung centre contact, as described by George and Roger above

Another option is to use the phosphor bronze contact arm and points, salvaged from an old relay.


Here are four, I made earlier.

Mike

[RReid]

Hi Mike. More great info being collected here, this thread is becoming a real treasure trove for those of us with the need/interest but not the experience!

Just to let you know, the link back to your "By Jupiter" thread back in reply #43 is broken by a gap in the spelling. I got there by cutting/pasting into a new tab and eliminating the gap, but I'm sure you'd want to fix it properly.

[Vixen]

              Fixed

[gbritnell]

Ideally a small piece of carbon rod that is spring loaded for the center contact would be the way to go.  It just adds more height to the rotor and therefore the cap. As with a topic like this it's great to have multiple ideas to work with.

[petertha]

Wires: I mentioned early on that I used test lead wire. That was before Roy Sholl (S@S engines) and Jerry Howell. Roy sold 2 sizes of wire, .100 (10KV) and .125 (20KV). The wire is black. Jerry Howell sells a white wire .125 (15KV). I'm sure there are other sources available.

My understanding (maybe wrong) is kv (kilovolt) refers to the maximum voltage rating mostly or entirely related to the sheathing, not the conductor wire itself? For a model distributer if we could wave a magic wand it would be scale diameter for aesthetics, but of course has to meet the electrical requirements & maybe that's the tradeoff. I don't consider ~0.1" too large for my intended application, but I'm just wondering more about these related details. What kind of voltage levels are we expecting or seeing in coils and specifically CDI's (my likely path). How would an under-rated kv rating manifest itself?

But I'm confused by the conductor specs, which I don't think relates to KV? I have a basic understanding of resistance specs & that is related to current carrying capability. Too much current & the conductor heats up, then bad things happen. So by selecting the KV & the nominal wire OD, are we indirectly defining a wire that just so happens to also meet our ignition voltage & current conductor wire requirements? I'm guessing the amperage must be quite low based on the acceptable cross section of the rotor conductor bar & terminals etc? What kind of amperage range is expected?

Another point of personal confusion. I think carbon based core/coated? automotive ignition wires are to help suppress electrical (EMI) / radio (RFI) interference. ie. that the carbon (or whatever it is) doesn't really improve electrical properties. I can understand if conventional stranded wire kicks out 'noise' to the extent of RFI problems for RC models if it interferes with the receiver reception & thus model control. I 'think' that's why we see those big braided cables, metal/grounded plug caps, integrated resistors... all that stuff on RCEXL type CDI modules used in RC applications? So on this basis are we saying conventional stranded wire is fine for a bench running engine? I've heard opinions that as the distributer & related distances shrink on our scale engines, arcing or cross-firing can happen or is likely to happen. But the counterpoint is many examples of small distributer + high cylinder count which seem to run just fine. So is it safe to say there must be other contributors to these problems (maybe gap distance). Or maybe person A used a coil & person B used a CDI brand X?

Why are capacitors used in FS distributers & we dont seem to need them in model distributers?

[gbritnell]

I believe S&S supplied a small capacitor for their ignition when using points rather than a Hall sensor.  My Holt has a points trigger but I have never used the capacitor and haven't had any problems.

[Roger B]

The voltage rating of a wire is complicated. A piece of typical PVC panel wire will withstand around 20kV short term, 6 kV long term but will carry a rating of 1000V. For long term voltage rating the quality of the conductor surface and the lack of an air gap between the conductor and the insulation is important to control the corona discharge which will cause long term breakdown.

Adding some resistance to the ignition wires or plug caps reduces the rate of voltage rise and the amount of RF interference produced.

Capacitors are used with traditional points and coil ignition systems to reduce arcing at the points, as the points open the capacitor holds the voltage across them to zero until it charges up. By the time the capacitor charges up the points should be open enough to withstand the voltage. The capacitor also allows for some resonant oscillation to prolong the spark.

[petertha]

Mike, if I recall your Jupiter distributer correctly, you initially considered both a shutter wheel type & a magnet disc type. You went with a shutter wheel & confirmed it functioned well on a bench rig so ended it there, at least for now. And you have also made a magnet disc type on different project.

Are there pros, cons or limits where one system would be net better performance than the other? My own naïve assumption, at least for a distributer of say 1" diameter or larger & 4-9 cylinders max, is a magnet disc is possibly a tad easier to fabricate. Disc requires N holes for N cylinders equally spaced. Magnets are spit cheap so buying 1 vs buying 7 more for an 8-cylinder doesn't seem like a concern. I've seen neodiums available as small as 1mm dia x 0.5mm thick. Maybe that's too small or maybe the quality / consistency factors in this? The shutter wheel requires a bit more careful milling compared to N spotted holes for a magnet disc, but not a huge difference either way. Maybe crowding too many magnets in too small a space on its disc can create issues? Or spinning an array of magnets in close proximity to the upstairs rotor assembly is less desirable for some reason & requires more vertical offset distance between them?  I think I'm hearing the hall effect sensor is essentially the same if not identical. They are also a relatively inexpensive component & require about the same consideration/methods to mount. I am also considering making 2 types just to see, but wondering out loud why not magnet disc type first?

[gbritnell]

The only issue with a magnet disc is there is no way to control the dwell. With a shutter disc you can control the dwell with the size of the windows.

[petertha]

Thanks George. I have no idea if my sketch is a figment of my own imagination so hopefully its understandable enough for this purpose, or correct me as needed. I've read some hall effect sensor specs & I understand ~20%. Using your V8 shutter & magnet dimensions as example, are you saying that the aperture width would constrain (reduce) the interval time the sensor sees compared to a bare magnet in a disc something like so?

I assumed the way CDI modules worked is they had some kind of internal logic; when the hall tripped ON based on some minimum magnet strength, they applied their own 'on time' spark duration based on that signal point mainly so it met the requirements of the capacitor discharge. But maybe it will attempt to deliver for as long as as it detects ON & that's where shutter controls? Now maybe I'm getting messed up with similar discussion on the RCEXL type modules that keep track of pulses & time & therefore know RPM & how to electronically apply advance/retard... they are a different animal apparently even though they also use hall sensors. I should probably address these questions to S/S but bottom line is I think their base configuration assumes a magnet disc? But I think I've also read where people use a shutter with S/S.

https://www.cncengines.com/ic.html

[Vixen]

Peter,

I am not sure you have got all that Hall trigger stuff quite right.

Lets go back to first principles. The image below represents the magnetic field surrounding a simple bar magnet (dipole). Note: the magnetic field is normal (vertical) to the two end faces (poles) of the magnet, but only in a very small part near the centre of that face. Everywhere else, the magnetic field is inclined at an angle. A Hall effect sensor acts as an electronic switch. It will switch ON, ONLY when an applied magnetic field is a) strong enough  b) normal (vertical) to the face of the sensor c) of the correct magnetic polarity.  One face of a Hall effect switch will only respond to the North pole of a magnet. Flip the Hall sensor over and the reverse face will only respond the the South Pole. Therefore the Hall sensor and magnet's orientation is vital.





We know it is possible to make a Hall effect multi cylinder ignition trigger systems by rotating an appropriate number of small rare earth past a single Hall sensor. Alternatively it is possible to use a single static magnet and Hall sensor with a rotating iron/ steel shutter wheel between. Both will 'sorta' work, but the shutter when has the potential for more precise timing. It seems unreasonable to assume that the 8 'cheapo' rare earth magnets in the rotating disc will all have identical magnetic fields. Some timing dither is to be expected. Which method you use depends on the design restraints and space available for your ignition trigger. The rotating disc could perhaps be smaller, or fit the available space better than the shutter disc. Sometimes it's the other way round.

One advantage of the shutter wheel (as George pointed out) is the ability to control the mark/space ratio (blade to gap ratio) of the output pulse from the Hall effect sensor. The mark/space ratio can be used to alter the 'dwell angle' of your ignition device. The 'dwell angle' gives time, especially at higher RPMs, to fully charge the capacitor in a CDI unit, or time for the field to fully build up in an ignition coil.  However, you need to be cautious. Coil systems fire when the points open i.e. the Hall sensor switches state. Commercial CDI units can fire either when the Hall switches On or when it switches to the OFF state. You never know which until you buy and test one. It's generally not a problem, as you can usually alter the angle where the Hall sensor triggers to suit your chosen CDI or coil.

When it comes to choosing the mark/ space ratio for your shutter wheel, you need to consider how and when the chosen coil or CDI fires. One will favor a longer mark to space ratio, whereas the another will favor a shorter mark and a longer space to give the desired 'dwell angle'. It therefore pays to optimise your design for ONE type of ignition system and stick with it. Swapping between ignition systems and hoping, may not always produce the best results.

Cheers

Mike