Model Engine Ignitions

Help Support HMEM:

74Sprint

Well-Known Member
HMEM Supporter
Joined
Apr 21, 2013
Messages
128
Reaction score
34
Location
Winnipeg, Canada
I saw that one on Amazon after I bought mine and thought I should have gotten that one. Oh well but, it's not a majour tool for me just something handy for doing quick checks. I see that one had a better review than mine and it sells anywhere from $44 - $56 CAD depending on the options, funny how they have the same screens and buttons. They are better than lugging around a old 10kg scope. :)

Cheers
Ray
 

74Sprint

Well-Known Member
HMEM Supporter
Joined
Apr 21, 2013
Messages
128
Reaction score
34
Location
Winnipeg, Canada
Timing pulses:
Until I get a timing wheel put together with magnets it will be hard to show the auto advance of the trigger pulses.
I setup the CD4047B to show the difference between the fixed timing pulse (leading edge) and the auto advance (trailing edge) using just my hand.
Before I go further I want to explain that the actual timing and it's accuracy depends on 3 things; 1 the strength of the magnet, 2 the sensitivity of the Hall-effect, and 3 the distance of the Hall-effect to the magnet. If all these 3 stay the same then setting the timing it should remain constant. I have found and I'm sure people using Hall-effects on here also have found that with the crank moving back and forth can cause the timing to skip around a bit. I find that on a brand new engine with brand new crank ignition the timing is pretty much stable but as the parts wear the timing starts to skip around a bit. Now passing a magnet back and forth past the Hall-effect with my hand there is no way I can duplicate a running engine, my hand is simply not fast enough or consistent enough. But I have a couple of pictures to show you.

web trailing 2.jpg
web trailing 1.jpg

I know it's hard to see but if you look at the trailing edge you can see that the edge has moved a little to the left which, would mean that the timing has advanced. How much well it's about 4usec. Which if you do the math......
Lets take an engine (2 or 4 stroke it doesn't matter) running at say 5,000 RPM so
5,000 / 60 = 83.33 RPM per sec.
1/83.33 = 0.012sec. per RPM
0.012/360 = 0.0000334sec per degree or 33.4usec per degree.
So changing pulse width by 4usec with my hand which, is about 200 RPM it will advance the timing by about 1/8 of a degree. This change will become more evident (larger change) when I use a timing wheel with more speed and my big scope, with 2 channels and better triggering.

I did find out with the scope that my dwell was 40msec which is more than the small coils could handle. I adjusted that down to 2msec and will start testing different coil driver components. Also I found that the the leading edge timing (fixed) stayed pretty much rock solid as best as I could with my hand. So things look promising.

cheers
Ray
 

74Sprint

Well-Known Member
HMEM Supporter
Joined
Apr 21, 2013
Messages
128
Reaction score
34
Location
Winnipeg, Canada
I finally found some time to get to the auto-wreckers and got 4 different types of ignition coils, all 12 volt. I tested the coils for strength along with some different types of IGBTs, MOSFETs, and some transistors. I didn't actually measure the voltage or current outputs from the coils so I don't know the joule outputs. I hate when they measure coil outputs in joules or watts because those measurements mean nothing and is just a marketing ploy. There is no difference if I have 1,000 [email protected] amp per second or 1 [email protected],000 amps per second the wattage/joules is the same. But the 1,000 volts will jump a gap easier than 1 volt.

Except for the Ford coil where I used just the NGK plug, I used a 349 ohm ignition cable with the Dodge and old style GM coils, the LS1 cable had 354 ohms, and the plug I used for all testing was a NGK UR6GP sparkplug with 4.4K ohms.

The charging devices I ended up using for the most testing:
STGF3NC120HD 7 A, 1200 V very fast IGBT with ultrafast diode QTY 1 UNIT PRICE $2.45000 v-good spark
IPP25N06S3L-22IN-ND IPP25N06S3L-22 MOSFET N-CH 55V 25A TO220-3 good but obsolete

I usually don't like using transistors because the cheaper ones are usually slow, can't handle the power both switching or higher duty cycles. They are also harder to get the biasing right. Yes they do make transistors that can handle problems but they also cost a lot more. So that leaves us IGBTs and MOSFETs to play with. IGBTs are basically BJTs (Bipolar Junction Transistor) mated to a MOSFET. The MOSFET handles the base biasing and current of the BJT. This makes it easier to turn the transistor on but, makes it harder to turn off. Sometimes but not all the time you need to bring the gate voltage negative or below zero to turn it off. Most of my IGBTs were like that and only the STGF3NC120HD I had worked with zero turn off volts and had fast switching that could handle the inductive loads. The EMI from the coils didn't seem to bother it also, I guess because of the wire resistance. The IPP25N06S3L-22 MOSFET I have also worked well but, not quite as good as the IGBT. The spark colour was good but not as fat. As I said all the coils below were 12 volt, so they didn't work at 5 volts.

12v Coils tested:
  1. DODGE 1998-2003 RAM 1500 PICKUP Primary 1.1k ohms Secondary 12.92k ohms V-Good spark
    1. SKP SKIC522 {#56027966, 56028172, 56028172AB} STANDARD MOTOR PRODUCTS UF198T CAD$8.68
  2. LS1 used from 1999-2007, coil blacks (-) red (+) blue (sig) CAD$17.94 to CAD$121.24 excellent spark without using a transistor but 12 volts is required
    1. Found mostly on SILVERADO 1500 4.8-6.0L
  3. GM 4 WIRE old style COIL GM 1985-1994 Primary 0.45 ohms Secondary 8k ohms STANDARD MOTOR PRODUCTS DR37T CAD$14.03 Good spark
  4. FORD 2.3L FORD FOCUS 2003-2011 coil on plug Primary 0.6 ohms Secondary 5.4k ohms CAD$9.57 good spark
  5. 3 different CDI coils had spark but very weak. These coils usually need between 100-120 primary volts to work.
The Dodge coil to me was the best overall for price and performance. The problem with it though is getting a new connector which was $23 CAD so best to get a coil and pig-tail from the wreckers or just a pig-tail. You can also use 2 small width female blade (red) connectors and glue them in. I don't see why people talk about the Ford COP being so good. It uses a long spring to contact the sparkplug, I guess one could solder a lead onto it after cutting it down. The Ford COP also has an inductor built into the spring for noise suppression. The old style GM 4 wire split connector that secures from the top was ok. This coil along with the Ford COP would make good coils for a CDI ignition. The GM LS1 smart coil was the easiest to hook up and make work and it had the best spark overall. This is probably because it has the transistor built right into it but, using just the Hall-effect wasn't strong enough to trigger it. More testing needed. This LS coil was also the most expensive and it has a pencil point projection for the coil wire to attach but, if you squeeze down on a normal coil wire female terminal you can make it work. So if your going to get one from the auto wrecker make sure to grab the coil connector and a plug wire. It is very easy to get this coil working just a few simple components is all that is needed. And it will work with either points or Hall-effect.

I was disappointed by the Ford coil's spark until I measured the secondary coil and found it was on the small side. I'm not sure but, I think if I were going to use 12 volts and wanted a simple ignition I would probably go with the LS1 smart coil from the auto wreckers. The current draw when not firing is just 23ma avg. 19-25ma range so this is excellent. If I were building a multi cylinder distributorless engine I would go with it. But the LS1 smart coil does present me with a bit of a problem and that is deciding if I should go with an TCI or CDI distributorless ignition for my race car. The LS1 Smart Coil Will Not Work With a CDI Setup.

Did you notice that no where have I ever mentioned a ballast resistor? Well that's because my circuit doesn't need one, it has adjustable dwell. For this testing I turned the dwell up to 3msec. to be fair on the testing. On every car that I have worked on that used a ballast resistor except the Fords with the built-in coil resistor, a stupid idea, bypassed the ballast resistor/wire during starting. This gave full voltage and current during starting to the coil and this would fall back when you let the key go.

Hall-effect triggering is based on it's Gauss value. What this means is that if you use a H-e with a gauss of 50 it will turn on sooner than with one that has a gauss of 350. In real life this means that if you use a strong magnet the H-e could turn on when the magnet is 2" away instead of a 1/4" away. On a small diameter wheel this would matter a lot and most likely drive you nuts trying to time the engine. I ordered and tried about 8 different Hall-effects and the one I like the most is the A1102EUA-T by Allegro, it's easy to find, popular, and is fairly cheap. I would like to find one that does have a higher gauss trigger point that is just as easy to get as the A1102.

Next up is to make a spinning timing wheel and some movies.

Cheers
Ray
 
Last edited:
Top