Twin Spark, who's done it? how? Options

[brer]

I have seen the twin spark dizzies. Straight forward.

Then there are the crank fire twin pack setups. ok.


what else can be used?
Piggyback systems?




(picture is for entertainment purposes only. this is not my motor)

[ottox914]

Anyone got any good pics of some twin plug heads, on the bench or on the motor? And, how about cost to do the head mod? Anyone have a ballpark $$$ they know of?

[McMark]

Ballpark: $$$$ expensive. (IMG:style_emoticons/default/wink.gif)

[brer]

How does $100 hr. sound?

(IMG:style_emoticons/default/welder.gif)


i'm kinda slow though.

[Jake Raby]

We perfected it last year with the 257 HP, 2.3L "Mighty Spyder" engine after about 6 months of cutting, fixturing and testing the procedures and plug placement.

Here is an example(IMG:http://www.aircooledtechnology.com/galleries/msheads/slides/Arnett%20manx%20058.jpg)

This is serious business, not to be attemted by those that lack fixturing/machining experience. It requires extensive preheating, welding and post cooling of the head castings, these heads pictured has 9 hours of solid weld time into them.

(IMG:http://www.aircooledtechnology.com/galleries/msheads/slides/Arnett%20manx%20065.jpg)

(IMG:http://www.aircooledtechnology.com/galleries/msheads/slides/Arnett%20manx%20052.jpg)

To do these first heads was a HUGE task, even for Len and myself to tackle. It required buying a dozen different spark plugs andseeing what worked best physically and then we had to work out the heat ranges. The fixturing process sacrificed 4-5 heads that needed to be cut up to provide the necessary approach angle to get the spark plug out of the way of the pushrod tube, exhaust stub and allow the plug wire to be installed- all of this was done with a 10mm spark plug- nothing else will fit and provide clearance.

When we finished these heads we had them digitized fully, so now all twin plug engines are completed from these tool paths and programs via 5 axis CNC machining which is a requirement to repeat these processes at anything less than a 2,000 buck expenditure...

Now here are my thoughts on this modification.
1- If you don't intend running more than 10;1 CR on pump gas or 13:1 on race gas don't bother with the mods because you'll never gain the benefit they'll offer

2- If you have a bore smaller than 96mm you won't be able to fit both plugs, nor will the flame front of the smaller bore lend it's self to the mods

3- If you plan on doing this mod, but retaining a dizzy you need to reconsider where to spend your money and thats because direct ignition with ONE plug would be a better choice all around than spending money on a dual plug mod only to slap 1970s technology on it for firing purposes!

4- Big bores love it, all my engines with larger than 102mm bore are getting this as standard, we are building two of them now- one N/A 2.9L and another 2.5 EFI/Intercooled boosted engine. The differences that can be seen with big bores and dual plugs are apparent, even at idle. The 2.3l ENGINE THAT MADE 257hp DID SO WITH ONLY A 34 POUND FUEL INJECTOR!!!Thats totally unheard of with single plugs.

5- He who thinks that conventional timing will work optimum with these arrangements has another thing coming, we have seen engines that need as little as 18 degrees of full advance to have optimum EGT/CHT and BSFC, normally these engines with a single plug would require at least TEN more degrees of advance to reach this point. That yells efficiency loud and clear and it also means that EVERYTHING about tuning the dual plug equipped engine changes and you'll be hard pressed to ever tune one without a very well equipped dyno.

So with that being said our LE 200 N/A and LE 220/LE 230 Turbo/large cc N/A heads heads have the option of dual plugging. It adds 1K to their price tag, but the work is done spot on with CNC repeatability.

I have a custom dizzy made from a Nissan Z22 engine set up that offers twin plugs, but I'd never try it without direct fire... Why???

Well this is why, in our experience a dizzy simply cannot offer the ignition manipulation that these mods work best with. The advance curve of a conventionally fired engine is linear as RPM increases, these plugs don't like advance at higher RPM and with a dizzy its difficult to accurately pull timing away from the engine as RPM increases...

The "Mighty Spyder" engine utilized RAT/SDS Engine management with our custom dual coil pack enhancement. With this set up a standard SDS "F" system can be made to fire twin plugs easily with a clean installation...

Its a great mod if you accept the alterations it requires. I'll be happy to help anyone accomplish this mod if they purchase our CNC twin plug heads, I have several combos cooked up that work very well with these mods.

Don't think that a local machinist can do this simply, or easily- doing it correctly gave us a workout and these engines are all we touch.

Tuning the "Mighty Spyder" took us 3 solid days, the first pulls were only 200HP, we tuned another SIXTY into it.

[alpha434]

Tits.

Like Jake says, the easy thing will be distribution. Screwing with the heads for that long will be the big hassle. And if your going to be doing something this drastic, then you may as well go all the way. After all is said and done, Jakes heads may even be the cheapest way to go *gasp*.

[alpha434]

Meanwhile. I'd want independent spark control for the second set of plugs.

Megasquirt?

[brer]

I've got that photo in my files Jake. Very impressive.


So adding additional spark is best for larger displacement seems obvious looking down the throat of a 104mm barrel.

no gains for small though?

The Alfa Romeo TS series used twin spark down to 1.6L
They also used two sizes of plugs to achieve it like others here have mentioned, 10mm and 14mm
additionally, while it may not be as hot as EI, i've ridden in some of these alfas and they really move despite the Mechanical Ignition.(those were not TS though, earlier ones)

[Aaron Cox]

dude.... next engine... twin plug (IMG:style_emoticons/default/smile.gif) thats just cool stuff.

[Jake Raby]

The twin plugs WILL help a smaller bore engine, BUT at that level the benefits are not worth the cost...

As we do this more and more we can gain more data and see why some things are occuring...

I look to do a single VS twin plug comparative/back to back test in the new lab later in 2007. I have not had the opportunity or the equipment to do this as good as possible previously..

[Chris Pincetich]

Do you think a twin plug 1.7 with a heavy rotating mass, like 9.5:1 CR, and high power ignition could get 40-50 mpg on 91 octane in a stockish 914? (IMG:style_emoticons/default/beerchug.gif)

[brer]

at that level the benefits are not worth the cost...

why do people keep saying that about my 1.7L ??

(IMG:style_emoticons/default/biggrin.gif)

[DNHunt]

Meanwhile. I'd want independent spark control for the second set of plugs.

Megasquirt?

Not at present without fudging it up. Read 2 separate ECUs sharing the same crank sensor and cam sensor. Currently, there is a max of 4 ignition outputs which could do a 4 cylinder COP or cylinder twin plug (Wasted spark) per board. Advance is only 1 table per ECU so the only way to delay firing of the 2nd plug is to use a second ECU. Too much fudging and room for error in my mind.

There is another board in testing that will expand the ignition outputs to 8. I don't know if it will allow separate control of each output. Even if it does I can't imagine tuning it.

As a point of interest, tuning ignition on the street is really hard. I can't say I can do it. To tell the truth, you can collect so much data it becomes more than you can use. I suspect to do a really good job one would need exhaust gas temps from each cylinder and head temps from each cylinder along with air fuel ratio. That would mean that for every sampling and in MS you can sample each rotation you would be looking at those 9 columns and MAP, RPM, throttle opening, pulse width and advance. The thought of all of those numbers makes my head spin. Course you could only sample every 50th rotation or some other number but, then you have to wonder if you'll miss something. Fortunately I have an ignition map Jake supplied me that I have changed very little and I know it is safe. Advance comes in early and falls off at higher RPMs. I would have never found that last part without Jake's help and most tables I have seen have advance steady after it reaches max.

With Jake's permission I'll share that table with the usual warnings (If you melt it it's your lump). Sorry but, such is life when you're trying off the wall stuff.

Dave

[cnavarro]

Just as with 911's, those guys pretty much don't twin plug until they get to 98mm or larger (for street engines at least). I see plenty of guys doing twin plug small bore engines, but they are running gobs or compression in what is in essense a race engine. Not that the chambers are any bit the same as a type 4, but a good comparison point nonetheless...

[Jake Raby]

Do you think a twin plug 1.7 with a heavy rotating mass, like 9.5:1 CR, and high power ignition could get 40-50 mpg on 91 octane in a stockish 914? (IMG:style_emoticons/default/beerchug.gif)

My "Super 2 Liter" combo is based similar to this with high velocity 1.7 heads...

It has gotten 51 MPG with a SINGLE plug.

It would be very hard to do a twin plug on a 1.7 casting due to chamber shape and volume.

We base the twin plug heads on the 1.8 castings.

[alpha434]

Hey, I know we've sort of gotten away from the idea with the cable-clamps...

Would a corona charge be enough to fire the extra cylinders? Or was the idea just to trigger an ignition sequence.

[brer]

the idea was to use an inductive trigger on each plug wire to fire a second plug on each cylinder. Since the second plugs would be COP, they would each have their own individual trigger..... clamped to the corresponding wire.

[Mueller]

the idea was to use an inductive trigger on each plug wire to fire a second plug on each cylinder. Since the second plugs would be COP, they would each have their own individual trigger..... clamped to the corresponding wire.

how funny, I talked to a few people about the inductive triggering as well...what is unknown (by me and those I talked to) is the lag or time delay for the second plug, I know that for pretty much any twin plug setup you cannot get both plugs to fire at exactly the same time, but if one it trailing too much, it might not have any postive results.

[brer]

what if both plugs were COP and triggered by the same signal?

[Mueller]

what if both plugs were COP and triggered by the same signal?

there will still be a time differance, but it might not be as measureable....

I think the EDIS can trigger COP now, been a while since I've followed that stuff....I know some of the pricey aftermarket ECU's can do what you want...all it takes is money (IMG:style_emoticons/default/wink.gif)

[DNHunt]

You could use EDIS to trigger COP but, it's not as easy as it seems on the surface.

At first is would seem that a module for a 8 cylinder car on a 4 cylinder engine might give 4 cylinder twin plug but, it won't work because the sparks are miss timed. It will fire 8 times in 720 crankshaft degrees (1 engine cycle). This means it fires near TDC (good) but also near BDC (bad).

So then, how about two 4 cylinder modules? Here are some things to consider. Each module must have a signal from a VR sensor or the module will revert to limp home mode (constant 10 degrees advance). If one uses 2 separate VR sensors so each module has it's own VR sensor there will be a difference in timing because it's impossible to mount them on top of each other. So, that means the 2 modules must share the VR sensor. Fortunately, this will work, I've done it on the bench with the trigger wheel on my drill press. One can use the tach imput signal (PIP) from one module to trigger the ECU. This is known to work cause it is exactly how I trigger my engine now. The timing advance signal from the ECU (SAW) can be split and go to both modules. This is known to work. Thankfully, the modules do dwell control themselves.

So at this point there would be a current from 2 modules that could charge the primaries of 2 coils and the dwell is taken care of kind of. Here's the deal, each EDIS coil has 2 poles and each is attached via the spark plug wire to a plug. The current runs from 1 pole through the wire to the plug with a strong spark in the fuel and air, then through the engine to the other plug where it sparks and returns to the other pole of the same coil. The second spark is weak because it has to spark through exhaust gas. This second plug also has a reversed polarity. Almost all of the energy from the coil sparks in the first cylinder. If you try to spark both through air/fuel as in twin plug situation you get 2 sparks with about half the energy of a normal application. I really don't think this so hot. I know this is not COP but, I'm getting there. The solution to this is to run the wires from tower 1 of the first coil to cylinder 1 and the tower 2 to cylinder 3 and repeat with the second coil. In essence, this is separate EDIS systems running in parallel but sharing imputs.

Now here's the problems with COP and EDIS. The coil drivers in the EDIS modules expect coils of a certain inductance. If that value isn't close the dwell will be wrong and the coils will either heat up too much or have a weak spark. I don't know what inductance EDIS expects and this info is not easy to find for most coils. If it could be matched it would probably work but, it would still be wasted spark and each set of plugs would fire twice in an engine cycle and that halves one of the advantages of COP which is long coil dwell. It is certainly a compromise.

The really solution as I see it should take 1 of 2 paths.

First, there are COP coils that are smart meaning they contain the driver and dwell control in the coil case on the plug. These then only need a signal from the ECU to chargeprimary and discharge collapsing the field. These probably would get a 5 Volt signal from the ECU. It's been a while since I looked but, I believe some GM coils are smart. To do this right, the ECU would need information on crank position for timing and information on cam position to know which cylinder to fire. Likely the ECU would need 4 ignition outputs and the 2 coils on each cylinder could share the same 5 V signal.

The second way to do it would involve conventional coils. In this case the coil drivers and dwell control are in a module or in the ECU. The sensors required would be the same as above but, one would need 8 ignition outputs that would carry 12 Volts and enough amperage to charge the primaries of the coils. The advantage of this is it is more widely applicable and the coils are a lot cheaper. Likely, the dwell would need to be tuned in the software.

To me this just seems like too much work. I just can't see that I would gain much over the EDIS system I am running now.