Ignition coils for small engines

[popsieParticipant @popsie]

Hi all! Happy New Year!

I have run a  standard car ignition coil by charging the capacitor to 150 – 200 volts and discharging it through the coil. this removes the need for standing current of 3amps from the battery. I’m trying to see if i can run it from scrap disposable camera flash units, as these run on 1.5volt batteries. WATCH THE CAPACITOR IN THESE UNITS as they pack a WALLOP!!!

[Stephen WesselParticipant @stephenwessel36637]

Hi. I’m new to this site. Re electronic ignition for small engines I have been using various modules on my 9 cyl rotary with a Hall effect switch. Recently I’ve been having start up trouble as follows: engine starts as normal,runs up to speed then cuts completely, all lasting about 3 seconds. You can repeat this ad nauseum but engine will never keep going. Frustrating as it has worked beautifully up to recently, flying a Sopwith Pup very realistically.

 

Lack of fuel was an obvious possibility but long since ruled out. Timing & plugs are  correct. I found today that the Hall sensor now doesn’t work at all. So my questions are:

 

a) what happens to them to cause failure? 

b) might gradual failure cause the symptoms observed?

c) how do I choose a new one bearing in mind my knowledge of electronics is almost zero?

 

My current module is a Zenoah running off 4.8V

 

Any help much appreciated. Thanks

 

Steve

[dixieParticipant @dixie]

Hi all Try Minimag .co.uk    Now under new management

[BogstandardParticipant @bogstandard]

With reference to the above post and Minimag.

 

Julian, the new owner, is resurrecting what Jim Shelley had to give up a few years ago due to ill health. That is the Minimag magneto. 

 

He has now running a prototype low speed magneto for running on hit and miss engines, which, when finished/ will solve most of the coil/battery problems.

 

He is beginning to give out info and taking Q&A in this post.

 

Minimag Link

 

Bogs

[John Wood1Participant @johnwood1]

Hi Steve, most interested in your use of Hall effect ignition.  The sensors are quite susceptible to flux caused by being too near to the EHT. I always keep the plug lead etc as far from the sensor as possible.

 

Thanks to everyone else for suggestions re coils, I now have a couple which work well sourced as spares from mopeds, not the smallest but my application is not that small so it doesn’t matter this time.

 

John

[Stephen WesselParticipant @stephenwessel36637]

Hi Dixie and John.

Thanks for the link to Minimag. Julian there was very helpful and sent me a new sensor which is now set up and ready to go. I shall remember your valuable tip John about keeping the HT well away. It could indeed have been a bit close and may have caused the trouble. Engine is now ready for start up so we shall see.

 

Steve

[Sub MandrelParticipant @submandrel]

I’d love to see articles describing how to make a trembler coil and a magneto in ME. Certainly not covered for 10-20 years?

 

I have a small electromagnet (like a single coil transformer 3/4″ across with only ‘E’ laminations). Flashing a magnet across the end gives a 4v p-p output. It’s easy to see how a step up coil could make that into a useful spark, but i don’t know how.

 

Perhaps there is an old ETW design that could be republished or updated.

 

Neil

[Ian S CParticipant @iansc]

I notice that Jan Ridders in his latest artical in ME on a simple 2-stroke I?C model engine suggests using a 6 or 12V automotive coil, and don’t use the small coils as used on scooters, mopeds and lawn mowers.Ian S C

[Richard ParsonsParticipant @richardparsons61721]

 

Steve – Hi there. This is an old old problem. Search as you might you will not find the cause EXCEPT for the plug! Get a brand new plug. A plug will work well for hundreds of hours and then fail if it is left for any time. I have had this happen many times with garden tools, scooters, and chain saws all of which had been laid up for some months.

Dick

[JasonBModerator @jasonb]

Posted by Stub Mandrel on 22/07/2010 21:23:29:

I’d love to see articles describing how to make a trembler coil and a magneto in ME. Certainly not covered for 10-20 years?

 

Modl Engine Builder had an article on making a small Mag complete with drawings, you should be able to get back issues.

 

Jason 

[Robert Miller 1Participant @robertmiller1]

On a somewhat related note, does anyone know how much energy is required to ignite the fuel air charge in a small engine?  With that figure, it shouldn’t be too difficult to make an electronic magneto to suit.  The circuit in common use in chainsaws, string trimmers, lawn mowers, and the like is found in the1979 US patent  #4,163,437. 

 

A word to the wise, these things have no protection for the darlington transistor so if there is no spark path, the stored energy goes into blowing up the transistor.  IOW don’t pull the starter cord with the spark plug disconnected.

 

Bob

[BogstandardParticipant @bogstandard]

Jan Ridders has had great success by the use of piezo units out of gas lighters, no coil, points, electronics etc, just a direct connection to the plug, fired direct from the piezo unit operated by a cam.

 

I don’t know how durable they are, but some of his engines run great using them.

 

Jan Ridders Site 

 

 

Bogs

[Sub MandrelParticipant @submandrel]

Thanks for the link to Jan’s site, I’ve only seen his stuff in ME before. Like Cherry Hill he’s someone whose work you can recognise just by the way it looks!

 

I will have to learn Dutch!

 

Neil

[BogstandardParticipant @bogstandard]

Neil,

 

Somewhere on each page of Jan’s site you will see an ‘English’ button.

 

Clicking that will allow you to read all.

 

 

Bogs

[John Wood1Participant @johnwood1]

I am currently using a coil and condenser from a moped and found that I had to increase the capacitor value by wiring a 220n component across the original in order to get a decent reliable spark, the engine now runs well.

 

John

[Anonymous]

First, to answer a question posed a few posts ago. The best I can find on the internet as to the energy needed to ignite a air/fuel mixture is a few hundred micro-joules. However, most ignition systems generate many orders of magnitude more energy than this, often 100-200 milli-joules. I suspect some this is required for reliable ignition and some is marketing bo@~%ks, particularly in the retrofit automotive area.

 

I am in the slow process of designing my own microprocessor controlled ignition system for my hit ‘n’ miss engine. So far I’ve got 6kV out of a home made coil, with a 10:1 secondary: primary turns ratio. One of the goals of the design is to use coils that do not need the more normal 100:1 turns ratio.

 

Two other questions I have been unable to answer are; how is the breakdown voltage of air affected by the air/fuel mix, and how does the breakdown of air vary with increasing pressure. As air pressure decreases the breakdown voltage characteristics are well known; it’s called Paschen’s law. But I’ve found nothing about breakdown voltage versus increasing pressure. I do have some contacts in the professional IC engine reseach area, so I guess I’ll have to fire up those contacts (pun intended) when I get back to working on my ignition system.

 

Regards,

Andrew

[Martin WParticipant @martinw]

Hi

 

Just been searching re high pressure v voltage breakdown and found this pdf file with data regarding pressure v voltage breakdown. It looks like a presentation document and page 16 gives a graph for breakdown voltages from low pressure to about 1000 bar; it also gives the low breakdown voltages for a variety of gases. Page 15 refers to the calculations and considerations re the breakdown processes.

 

I assume the data is valid as it has been produced at/for CERN , usual disclaimers etc.

 

Cheers

 

 

Martin

[Howard JonesParticipant @howardjones35282]

guys in australia building model engines often use the Bosch GT40RT coil. it is quite small compared to the canister type coils

this was just placed on top of my scanner a moment ago to show you.

those silver eyelets are 5cm centre to centre.

the coil is designed for kettering type points.

it is epoxy filled. they are about $59 Supercheap auto in australia.

[Anonymous]

Hi Martin,

 

Thanks very much for the link. The breakdown properties of air are much clearer to me now. I’m also an official idiot for not seeing what was in front of me! I didn’t spot that the X-axis of the graph is a product, pressure times distance.

 

So, let’s say we have a spark gap of 0.5mm. According to the Paschen curve that equates to a breakdown voltage of about 2.5kV at one bar (close to normal atmospheric pressure). Without bothering to measure it I assume that my hit ‘n’ miss engine has a relatively low compression ratio, let’s say 6. If we start with air at 1bar and 20°C and assume that the compression process is isentropic, and take the heat capacity ratio of air as 1.4, then according to my calculations we end up with air at 12.29 bar and 324°C at top dead centre. If we now go back to the Paschen curve we come up with a breakdown voltage of 25kV for a spark gap of 0.5mm and a pressure of 12.29bar. According to my Bosch Automotive Handbook spark voltages are normally about 25-30kV. At least this is in the right ballpark.

 

The question is though, do our ignition systems really generate 25kV? And if they don’t, how come they work? ETW talks about ignition voltages for model engines ranging from 2kV to 8kV. I suppose that what it all boils down to is that at some point one has to stop fiddling around with the maths and go and build something to see if it works!

 

Regards,

 

Andrew

[Hugh GilhespieParticipant @hughgilhespie56163]

Andrew,

 

I suspect that the answer is in the air temperature. I would guess that the breakdown voltage falls off with increasing temperature, so perhaps 2 to 8 kV is enough?

[Billy MillsParticipant @billymills]

The Paschen curves are not really applicable to a spark plug because of the geometry of the gap, they are for large parallel and smooth plates in Dry gas. Dielectrics ( fuel-air mix in this case) break down in a complex way. For example UV will cause the gap to fire at a much lower potential as will ionising radiation ( eg geiger tube) or carbon particles from previous cycles. Water vapor will also lower the breakdown voltage as will any  combustion products.

 

 For a spark plug the electrodes have relativly sharp edges and accumulated very small carbon particles so field emission will set in long before simple breakdown as the voltage rises accross the gap. The ionisation  will then avalanche the channel accross the gap.

So you don’t need so many volts but rough points help a lot. That is one way in which abrading the points really helps.

 

Regards,

Alan.

[Anonymous]

Hi Hugh and Alan,

 

Good points, thanks! A further search on the internet found an approximation to the breakdown voltage that takes into account gas temperature. This brought the breakdown voltage for the previous parameters down to 12kV. However, as Alan has pointed out there are many other factors that fortunately conspire to lower the breakdown voltage.

 

So, I suspect the best we can say is that we need a few kV and a few mJ to produce a spark that will run a model engine. It is probably time to stop wielding the calculator and start building something. Bit of a relief actually, as the high voltage probe I bought for the oscilloscope is good for 15kV, so I don’t need to find a higher voltage probe.

 

The first thing to do is knock up a small adjustable spark gap and see if my ignition coil setup will create a spark in air at STP.

 

Regards,

 

Andrew

[BowberParticipant @bowber]

I think your over complicating things.

Points and coil systems operating from battery voltages have been in use for over 100 years and work!

Now we have transistorised units that perform the same basic functions as points at battery voltages but you still need the high primery to secondery ratio to get a good spark.

If you want lower coil ratios use a power transfer system like a magnito or CDI, but then you need a higher primery voltage, and so the cycle goes on.

 

Steve

[Billy MillsParticipant @billymills]

Andrew,

The spark gap voltmeter is easy to build and a good rough and ready indicator of what an ignition system is doing. If you use balls from a race that are not too chewed you will get around 2.5KV/mm for balls around 5mm. The advantage of balls is that they are more reproducable than points and less subject to wear. ( you get field emmission with a few volts and a sharp enough needle) Some people put the gap inside a glass tube to  a. jazz it up  b. improve the stability  c.pretend that they are Dr Frankenstein.

 

I would not use a scope probe unless you run into a gap to avoid the risk of flashing the series resistor of 1000M ohms ( if it’s a 1000:1 probe). Most  “Car Ignition Scopes” use a coupling capacitor of a few pF ( turn or so of well insulated wire around EHT lead or a fancy clip) into a few M of coax with -say- a reliable 2.2nF capacitor across the coax to make a 1000:1 capacitative divider. You don’t use a 1000M resistor because there are no low frequencies to see, the -3dB point being around 70Hz.  You calibrate the thing with the scope cal sig or an audio oscillator  or anything else handy- eg a multivibrator hooked up to the EHT lead ( coil disconnected) then adjust for 1000:1 division or whatever.

 

Normally you see a fair overshoot spike before the gap fires, you get to see what the mixture sees. The overshoot before ignition will jitter, it’s the random delay before the channel in the gap ionises so the bigger the spike the less delay and the smoother she runs. Usually people trig the scope from the points/ Hall device. For more than one plug you bug the coil lead so you can see the different plug breakdowns sequentially( with a distributor). For seperate coils you run seperate taps to the coax inner.

 

It is a nice idea to always use DC scope coupling with a hv probe. If you go AC the series cap can charge toward   the  full direct voltage level ( 25KV on a CRT display) before it ionises and takes out the input stage of the scope.

 

regards

Alan.

[Anonymous]

Hi Steve,

 

It may be a bit complicated, but I like to try and understand what I’m doing before building stuff. I have very little workshop time, and I don’t want to waste time building stuff that will not work, where I could have deduced that by thinking about it in the first place. Just because coils and contact breakers have longevity doesn’t mean I want to use them. They might work but they are not reliable. I remember being out there in the wet fiddling about with eroded contacts when the car wouldn’t run properly! On my hit ‘n’ miss engine I am planning to use a magnet and hall effect device. Neither do I particularly want to buy an ignition system that uses a Darlington transistor and a few wire-ended resistors. A Darlington transistor is a bipolar device and therefore has both majority and minority carriers. The mobility of minority carriers in the semiconductor lattice is slower than the majority carriers, so bipolar devices are relatively slow to turn off. I’m using a MOSFET and I let the flyback voltage from the coil avalanche the drain-source junction diode. This occurs at about 800V. So, 800V on the primary times a 10:1 step-up gives several kV. Which is roughly what I see on the oscilloscope. Once I’ve got a microcontroller in the system, it becomes easy to implement things like speed related advance and retard, multiple sparks and other things I haven’t thought of yet.

 

I did think about a CDI type ignition, but I didn’t feel like designing the step up converter. It would probably have to be a flyback style converter, and it’s a pain getting small quantities of gapped ferrite cores.

 

Finally I think it is a nice juxtaposition to have an old style hit ‘n’ miss engine together with a surface mount microprocessor ignition system.

 

Regards,

 

Andrew

[Author]

Posts