Wasted Spark for 944

[barks944]

I am implementing a wasted spark ignition system for the 944 and I am looking at options for what the best way to achieve this would be. The DME will have a dedicated ouput for each channel of wasted spark. This will be a 5V/Gnd digital output or could even be a12V/GND digital output. These outputs wont have significant current carrying capacity so wont be able to drive a coil directly. There are a multitude of ways that this could be implemented. One that looks ok is:

Combined amplifier coil pack e.g: http://www.efi-parts.co.uk/index.php?productID=121

But it doesn't look like a particularily high quailty bit of kit. There must be plenty of OEM systems that could do something similar. Ive also looked at Coil On Plug but I read some of the nissan boys have a lot of trouble with them due to the heat issues. Any suggestions on good options in terms of conceptual design or for appropriate bits of hardware would be appreciated. I know WUF used some MSD ignition and COPs from a suzuki bike or something.

Cheers

Tom

 

[GPF]

Hi Tom,

I presume this would eliminate the rotor arm & dizzy?

If you need a working test bed count me in []

 

[barks944]

Yes it would . Just need to find something thats going to be nice and reliable as well as delivering a nice strong spark.

Edit: Availablity and price are also important!

 

[bmnelsc]

There must be plenty of OEM systems that could do something similar. Ive also looked at Coil On Plug but I read some of the nissan boys have a lot of trouble with them due to the heat issues

Just a comment on the nissan coils - I've had a 91 Maxima, 99 Maxima, 07 G35x and currently 08 Ex35. Only the 99 Max had a problem with coils at 99k kilometers (replaced under warranty) and then went fine till traded off at 225k kilometers. No problems with G35x at over 123k kilometers. I believe the coil failures due to heat were a problem late 90's/early 20's but I believe those issues were resolved. I think if you were buying new it would be very unlikely that you would get any old design coils. Not sure what these coils are worth now but cost over $100 CDN apiece for the 99 Max in about 2003.

Also - GM used a coil-near-plug design on the LS series V8s which used a short plug wire between the coil and the plug that kept the coil up on top of the valve covers and away from the hottest parts of the engine. Not as neat as the nissan coils but then you don't have to worry about the width/depth of the plug tubes as you are dealing with plug wire rather than a rigid stem. You still have to figure out how to fix the coils to the cam cover.

Finally - quick web search shows the GM coils are about $60 CDN and the Nissan ones are $55-$70 CDN. Of course you need 4 if you go this route. Looking forward to seeing your results.

 

[George Elliott]

Tom, what is the advantage of the setup you are aiming for?
George

 

[barks944]

I believe this type of system can deliver a stronger spark across the range of operation and operate more reliably than the distributor based system found on the 944. Hopefully there will be a performance benefit at the higher RPM's. Part of the work will be to test different systems to find out if there are actually any performance benefits.

 

[Bob]

Is it true that at high boost the spark can be 'blown out' and this is why the wasted spark type of system is beneficial for turbos?

 

[barks944]

The coil is working seriously hard at high revs. At 6000 rpm a rotation of the engine takes 10ms and the coil must fire twice in that period. That means on a 4-pot that theres just 5ms between sparks. Typical dwell for a coil is around 3ms so we still have long enough to dwell the coil properly. I'm not sure how the coil would respond to near continual charge/discharge it must build up heat pretty quick and this would probably reduce the strength of the spark due to increased resistance. The dizzy will prob drop a few volts too. Its prob ok for a road car to be honest but thats all speculation I'd like to see the result on a dyno.

 

[bertelli_1]

So you are supplying the coils with 12v from the ecu? Usually they would have a separate feed, with both pairs earthed via the ecu. I have used vb921 transistors successfully for this very task. Unless you are looking at coils with a low current, logic level input? Personally i'd recommend using a ford edis or peugeot coil (very plentiful used and new) which are very reliable - you may even be able to use the standard ford ht leads. This coil requires a high current 12v input (25amp fuse?) and both pairs earthed by the ecu. I can't think of any coils that will fit the 8v head nicely (cop style).

eta: how are you going to control charge time?

 

[barks944]

I've designed a replacement for the digital board within the Motronic ECU. My new board supports two wasted spark outputs, these are low current 5V outputs and will need amplifying. Potentially I could use these VB921 transitors and drive a coil pack directly. In fact this would cut a lot of cost wouldn't it. If I did that I could use the existing ignition output for one bank and just have a single additional output for the second bank. Hmm, I like that idea.

Dwell time will be calculated by the ECU based on the measured battery voltage.

 

[bertelli_1]

I'm not sure about basing dwell on batt volts, my ecu uses it as a correction factor. Maybe you could simply fit an ignition amplifier (like the stock s2), there are 'twin' amplifiers available which may be suitable for wasted spark. My other car uses a 5v wasted spark setup although the coil wouldn't fit a 944, maybe there is an alternative you could use? it sounds like a 5v triggered coil may be the easiest option. Heres a link to my setup - http://www.conrod.co.uk/coil.html

 

[barks944]

Heres my understanding of dwell timing.

When a voltage is applied across the primary winding of an ignition coil the current will gradually rise towards its peak static current, this is not instant due to the inductance of the coil. This current through the primary winding creates a magnetic field surrounding both the primary and secondary winding. The strength of the magnetic field is related to the current. The higher the current the higher the magnetic field strength.

When the voltage is removed from the primary winding of the coil the magnetic field rapidly collapses. This fast change in the magnetic field generates a high voltage in the secondary winding and the spark plug fires. As the current through the primary winding stops very quickly the voltage generated in the secondary winding is largely goverened by the current through the primary winding.

Assuming a constant voltage across the primary winding time then becomes the factor which controls the current running through the primary winding. Dwell time is then the time it takes at a known voltage for the current through the primary winding of a coil to reach a known value which will generate the correct voltage across the spark plug.

So in determining the appropriate dwell time the following parameters must be considered:

• The voltage required across the spark plug gap
• The voltage that is applied across the coils primary winding
• The inductance and resistance of the coil
• The ratio of primary to secondary windings

In practice the microprocessor in an ecu will use a correction table to get the dwell time based on the battery voltage. The correction table will be specific to a coil.

 

[bertelli_1]

The important points are that 1. the dwell time is short enough to not overheat the coil. 2.the dwell time is long enough to provide an adequate spark at high rpm/boost. As you say the settings will be specific to a certain coil, in any case it will give a much better (and more reliable) spark than the stock setup! The following site is worth a read - http://www.vems.hu/wiki/index.php?page=DummyIgnition

 

[barks944]

Good link thanks. I take it you have a Mega Squirt Setup?

 

[bertelli_1]

I do use a megasquirt ecu but I have changed a lot in the time I've used it, my set up is pretty unusual - certainly regarding the coil. I reckon your project is much better suited to an 8v 944 than MS, mainly due to not having to change sensors or the crank trigger, plus I take it a lot of the idle/cranking/part throttle settings can be left alone?

 

[barks944]

I have access to all the original ECU data so I will be using that as a base for the new system.

Btw love your Mk2 engine swap! Is it running now? I have a MK2 1.4 which we are considering swapping the engine in. The 1.8T VAG unit looks spot on! What was the time and ££ cost?

 

[George Elliott]

Very interesting info "you two". []
Almost converted me to diesel, when we did history at school, they never said cavemen knew about ignition dwell.
I cant say I understand he subject in any depth, but this reminds me of the subject of twin sparks, and the benefits of two flame fronts on large bore engines.
Like the Mk2 1.8T conversion, I had a 1.8 turbo technics Scirocco Storm many moons ago (170bhp) and a G60 Golf among my VW past.
Keep us posted with developments, I am learning
George
944T

 

[bertelli_1]

Is it running??? I've done over 10k miles in it since the conversion! Currently its being used for a 44 mile commute, its ultra reliable and still gets over 32mpg average. Partly the reason its so reliable is due to its 'mild' state of tune, set up for 1.15 bar of boost. These engines are good for 1.4 bar on the standard internals. The swap took me around 100 hours, including making the loom & initial mapping (I'm now on my 44th 'tweak' of that original map......). As for costs I'm not really sure but certainly 2 or 3k if not more, these projects just seem to keep evolving & getting tweaked.... Is it worthwhile? 240ft/lb and 1 ton says yes.