Ignition Thoughts/Ideas/Recommendations

[NoDOHC]

So now that I am nearly certain that I have squeezed every single possible horsepower out of my engine with the stock ignition. I am looking for ignition options.

I have a limited budget for this (I do not want to exceed $200.00 for new ignition).

I have heard mention of LS1 coils, Some have mentioned that MSD coils work pretty well, but I don't know if I could use one for the rotary.

Explanation:
Because the leading plugs are wasted spark, the leading plug coil must fire 2 times per engine revolution. Because the coil requires approximately 0.5 ms to fire and at least 3.5 ms to charge (or the spark is very weak), The leading plugs can not reliably fire above 250 Hz 1/4ms = 250 Hz.

{** IGNORE THIS PARAGRAPH IF YOU ARE NOT AN ELECTRICAL ENGINEER**}
The current through the leading coil peaks around 4.8 ms, indicating saturation. Because the current through the inductive coil increases at a nearly linear rate, the coil energy (0.5LI^2) can be computed based on charge time. If maximum charge occurs at 4.8ms, the current at 3.5 ms will be 3.5/4.8 times the peak current or 0.729 times the peak current, 0.729^2 gives 0.532. Thus at 3.5ms charge time the coil contains 0.532 times as much energy as at 4.8ms charge time.

Unfortunately, 250 Hz equates to 7500 RPM. (7500 ignition events/rotor/minute = 15000 ignition events per minute = 250 ignition events per second). This means that the leading coil is not charging adequately to light the fuel mixture. The trailing coil is similar as the charge signal must travel a common line for both rotors, even though neither coil is close to saturation.

Now I can give some actual hard data for mechanical guys who don't care about EE formulas.

I have attached a graph of normalized torque output for my engine.
Normalized Torque = Wheel Torque/VE of the engine as a decimal (IE: 1.1).
Most engines will have a very constant normalized torque as it is basically a measure of chemical efficiency (which depends on combustion chamber dynamics, engine displacement, compression ratio, Air/fuel ratio, etc.), all of which are fixed for an engine (if the AFR is tuned consistantly).

A good number for a piston engine that I have heard is 1.0+((CR-8.0)/30) lb-ft/cid at 100% VE (1.0).

The chart is included as a .jpg. These numbers are listed for anyone who wishes to check my work.

Engine Speed (RPM),Fuel,AFR,VE,Torque,Normalized Torque
5000, 4.85, 13.3, 103.05, 145, 140.7
5500, 5.02, 13.3, 106.67, 151, 141.56
6000, 5.21, 13.2, 109.87, 155, 141.07
6500, 5.4, 13.1, 113.02, 160, 141.57
7000, 5.5, 12.8, 112.47, 154, 136.92
7500, 5.3, 13.3, 112.62, 148, 131.42
8000, 5.1, 13.4, 109.18, 139, 127.31

From this trend, it is clear that the engine makes very consistant torque/VE below 7,000 rpm, but this number starts to decrease abruptly as the speed exceeds 7,000 rpm. (Interesting that this is a very similar speed to that computed above, isn't it?)

Interresting Side Note:
Using the formula above, the 100% VE torque of this engine should be:
1.0*((8.2(Compression ratio)-8.0)/30)*160 (effective displacement in cubic inches)= 1.007*160 = 161 lb-ft at 100% VE (1.0).

If you use an 20% drivetrain loss calculation (standard for Mustang Dyno and IRS car) you get a flywheel torque number of 176 lb-ft at 100% VE
Thus this rotary engine makes 10% more torque than an equivalently sized piston engine (with the same compression ratio) would make.

{** IGNORE THESE 3 PARAGRAPHS IF YOU ARE NOT A DATA ACCURACY FANATIC **}
The injection time is actually shorter than the injector energization time that is used for these calculations. Therefore the engine volumetric efficiency is probably not as good as the table above would indicate. However as the injection times listed are all within 10% of the mean injection time, the introduced error from an energization time offset in these values would not be appreciable.

The torque/VE claim would actually be higher if actual injection time were used as this would decrease the implied VE without modifying the measured torque.

The drivetrain efficiency would have to be 90% (which few people would accept as valid) before the torque numbers would fall into the piston engine domain.


Oh yeah, I need to compute what this engine should make for power if it didn't have ignition issues. No, actually I will need the VE curve for 8,000-9,000 rpm to do that (which I do not have). At 8,000 it is pretty simple 141 Wlb-ft/VE * 109% VE * 8000 rpm / 5252 gives 234 WHp (Probably a little higher at a higher rpm).

What should I use for ignition?

[C. Ludwig]

Your ECU is not capable of direct fire ignition which is really what you're asking for. BDC did a write up on the evil forum on how to use another trailing coil and toggle in place of the leading coil to enable direct fire with the E6K/X. This replaces the wastespark set and effectively halves the duty cycle of the wastespark setup allowing full saturation of the coil.

LS1 and LS2 coils want approximately 5.5-6.0 mS of charge time for full saturation. Just switching to those and still running wastespark won't benefit you much. You could use those combined with the toggle as above though if you wanted to get industrious. The LS2 coil is fairly hot. The LS1 not as much. Both are available on Ebay for cheap.

A 6A or similar on the leading coil only would also solve the problem and fit your budget though I am not a fan of CDIs until they become absolutely necessary are you're no where near that point.

Think your VE problem is still going to be a mechanical limit though. Prove me wrong. Great job so far!

[NoDOHC]

The toggle doesn't help me at all, as it still uses a common line to send out the ignition signal. This means my coils still get a drastically too-short charge time, even while being nowhere near to being saturated.

At 9,000 rpm, I get a maximum charge time of 2.8 ms (3.3 ms - 0.5ms discharge time). This will barely fire the engine.

Example of ignition signal:

3000 rpm
------------Leading -------------------Trailing--
___________|---------|___________|---------|

6000 rpm
--Leading ----Trailing--
_|---------|_|---------|

9000 rpm
-Leading-Trailing--Leading-Trailing
_|-----|_|-----|_|-----|_|-----|

Both the leading and trailing coil take a similar output from the ECU (just differing in the split angle), Thus the toggle signal is meaningless for the signal.

What I need is a "Smart" ignition system (that controls dwell independant of signal charge time). I can easily build a circuit that divides a signal to multiple ignitors based on the toggle output.

[TitaniumTT]

Paul Yaw has a set of Bosch coils that he has dyno proven will make more torque at equivalent charge times. I plan on upgrading to these over the winter. Aside from that little tidbit of information, I really have nothing else to input on the coils themselves. I plan on contacting him in the near future.

LS1 coils are quite robust. I have had only one problem firing a wet plug (which I pulled, blew out, re-installed and have run for >1100 miles now) BUT they do like alot of dwell time and running them in wasted for the leading is just not going to happen. I'm pretty sure I'm going to be out of coil when I up the rev limit to 8k

Renni coils are something else to consider as they are designed to fire a 9k rpm rotary........... although you have that wasted spark to deal with.......... although you said you could build some sort of electronic trickery to toggle the signal.......

FWIW I was speaking with Motec about the shortfalls of the LS1 coils. We discussed three possibilities.

Yukon LS2 coils - good
Renni coils - Good - he rec'd them to me a while back but I went LS1 for ease of procurement and when I suggested them again is when he mentioned Paul doing a back to back dyno of renni coils vs the Bosch ones and the Bosch made more torque.

So it's either Yukon LS2 or Bosch over the winter for me.

[NoDOHC]

I want to go LS2 really bad, but my ECU doesn't support them.

[C. Ludwig]

I hadn't thought about the "direct fire mode" not really being direct fire with the 3 channel systems like the E6X/K. You are correct though.

The best inductive coils your going to find are going to be the Bosch Motorsports coils or the Mercury IGN-1 and IGN1-A coils. The Mercury A coils have a built-in ignitor for direct connection to your ECU. The Bosch and IGN-1 coils will need an external ignitor.

The LS2 coils will certainly work with the E6X. Why would you think they're not supported. However, with any of the inductive solutions as I stated in my first post, you'll be near the limits of saturation since you'll still be driving the leading in wastespark. I know the Mercury coils require a lower charge time than the LS1 or LS2 coils. You should see good saturation at around 3.5-4.0ms. Not sure what the Bosch coils need. Any of these solutions though is going to topple the budget you stated in your first post. Again, the MSD 6 or comparable on the leading is about the only thing that will be an upgrade and fit your budget.

[NoDOHC]

I'm sorry, I meant that the e6X is not capable of individual coil-per-plug on a rotary. Wasted spark with an LS2 coil is identical to what I have now.

I have another option that I am researching.

I will summarize it at the risk of flaming.

I think I will keep the stock coils and run them at 24V DC. This will effectively cut charge time in half, allowing a good hot spark at 9,000 rpm and costing me nothing.

(I can easily reconfigure the E6X to give a 2.4 ms charge time instead of a 4.8 ms charge time.)

[TitaniumTT]

Originally Posted by C. Ludwig

I hadn't thought about the "direct fire mode" not really being direct fire with the 3 channel systems like the E6X/K. You are correct though.

The best inductive coils your going to find are going to be the Bosch Motorsports coils or the Mercury IGN-1 and IGN1-A coils. The Mercury A coils have a built-in ignitor for direct connection to your ECU. The Bosch and IGN-1 coils will need an external ignitor.

The LS2 coils will certainly work with the E6X. Why would you think they're not supported. However, with any of the inductive solutions as I stated in my first post, you'll be near the limits of saturation since you'll still be driving the leading in wastespark. I know the Mercury coils require a lower charge time than the LS1 or LS2 coils. You should see good saturation at around 3.5-4.0ms. Not sure what the Bosch coils need. Any of these solutions though is going to topple the budget you stated in your first post. Again, the MSD 6 or comparable on the leading is about the only thing that will be an upgrade and fit your budget.

I believe there are Bosch coils with built-in ignitors available. In speaking with Motec who turned me on to them, I mentioned this as a consideration.

Mercury Coils are going to need by design a very low dwell time. They are in fact designed for actual 2-stroke engines. However, I would be a little leary of their spark energy. I don't know of any 2-stroke O/B's making all that much HP. There are 300+ ones running around all day @ 9k and up using stock.... in this case Merc.... coils.

The 24V idea is interesting......... how would you do that? Wire it up.

[NoDOHC]

Short term, I would cut some connectors out of a harness on an old car that is destined for the junk yard, connect 24V to them (using an extra battery in series without any alternator or anything feeding it).

If this fixes my high end bogging, I will purchase a 12V to 24V DC-DC converter and feed my coils with that. (This should get me by until I decide to install the GT37). Once I install a turbo, I will need better ignition, as the open circuit voltage of stock coils is limited to ~30,000 volts (from what I recall). I think LS1 is ~45,000 and LS2 is near 50,000 volts.

[C. Ludwig]

The Mercury inductive coils have an output of somewhere between 120-150mJ depending on who's literature you look at. They're what Electromotive use for their DIS systems and they claim 150mJ. That's the hottest inductive coil I'm aware of. I've used the TTL version with great success on a couple rotarys. Love these coils! A 3.5ms charge time will get you 90+% saturation. Should have more than enough energy even at partial saturation to ignite an NA mixture. The LS-2 coil is supposed to be around 120mJ but requires around 6ms of charge time to get there.

Bosch Motorsports does show several inductive coils with built-in ignitors. The highest output rating is 100mJ and it's a COP coil. Would be interesting to try.

[NoDOHC]

If I could get a set of Mercury IGN1-As for a decent price, maybe it would be worth going E8 (that will have to wait until next years budget).

I won't have a full 6ms for an LS2 coil (I intend to rev to 10,000 on my next iteration, if I can light the charge).

Does anyone know what the saturation energy of the stock coils is? If not, I will get an oscilloscope and find out. I would like to be sure that I am significantly upgrading ignition, as the stock coils seem to ignite very well at lower revs and I can run them at higher voltage to get over the rev issue.

[RXTASY]

Just bought 4 of these to run with my P1000 and E85.
http://www.diyautotune.com/catalog/i...oil-p-394.html

[RICE RACING]

I have run my own experiments comparing inductive to boosted inductive then to CDI based, as per a good Mazda test I found more power, lower EGT, leaner AFR (all indicating better more complete burn), on CDI than any version of inductive all other variables held constant. There is also a nice increase in RPM, again most cant test this on road/track.

Most who trash CDI or dismiss it have never run proper tests to back their views.

The best set up is COP, and CDI on all plugs, and wasted spark on leading

[RICE RACING]

Originally Posted by NoDOHC

The toggle doesn't help me at all, as it still uses a common line to send out the ignition signal. This means my coils still get a drastically too-short charge time, even while being nowhere near to being saturated.

At 9,000 rpm, I get a maximum charge time of 2.8 ms (3.3 ms - 0.5ms discharge time). This will barely fire the engine.

Example of ignition signal:

3000 rpm
------------Leading -------------------Trailing--
___________|---------|___________|---------|

6000 rpm
--Leading ----Trailing--
_|---------|_|---------|

9000 rpm
-Leading-Trailing--Leading-Trailing
_|-----|_|-----|_|-----|_|-----|

Both the leading and trailing coil take a similar output from the ECU (just differing in the split angle), Thus the toggle signal is meaningless for the signal.

What I need is a "Smart" ignition system (that controls dwell independant of signal charge time). I can easily build a circuit that divides a signal to multiple ignitors based on the toggle output.

Forget inductive.

What you NEED is complete CDI system. Always running very high chamber pressure, with lots of fuel and water injected I have trialed every CDI offering on the market (or so it seems) also all the "hottest" inductive types, and many things in between from boosted primary to "Don Mega" coils snake oil and various guru's claims.

In the end CDI on a Wankel Engine wins hands down. You only need to follow Mazda to see this but some of us need to find out our selves too

If you want to know my recommendation feel free to PM me and I will gladly pass it on.

[TitaniumTT]

Originally Posted by RXTASY

Just bought 4 of these to run with my P1000 and E85.
http://http://www.diyautotune.com/ca...oil-p-394.html

Linky no worky...

Originally Posted by RICE RACING

I have run my own experiments comparing inductive to boosted inductive then to CDI based, as per a good Mazda test I found more power, lower EGT, leaner AFR (all indicating better more complete burn), on CDI than any version of inductive all other variables held constant. There is also a nice increase in RPM, again most cant test this on road/track.

Most who trash CDI or dismiss it have never run proper tests to back their views.

The best set up is COP, and CDI on all plugs, and wasted spark on leading

I did I did a back to back test of an M&W CDI system vs my inductive bosch coils. The CDI picked up 3hp but the idle blew donkeys... For the price savings of close to $1500, I'll stick with my inductive setup at this point. In the future when I start looking for north of 500 honest bhp with either an E85 or a water injected setup, then I think the CDI will hold merit. But at 400whp and below, especially n/a as NoDOHC is, the CDI just isn't worth the bread.... that's just my opinion.

I will say, and I do need to call C. Ludwig to tell him what happened with the QuickSilver coils that he rec'd, those are some stout fucking coils. Blow alot of things out of the water, LS1 included. The QuickSilvers added 4whp to my car when she was running pig ass nasty rich and getting ignition breakup AND they cleaned it up, lower egt's more hp, L was about the same but the breakup was gone. The ONLY drawback to the QuickSilver coils is their size but damn do they light it off. Thanks again to MC

Originally Posted by RICE RACING

Forget inductive.

What you NEED is complete CDI system. Always running very high chamber pressure, with lots of fuel and water injected I have trialed every CDI offering on the market (or so it seems) also all the "hottest" inductive types, and many things in between from boosted primary to "Don Mega" coils snake oil and various guru's claims.

In the end CDI on a Wankel Engine wins hands down. You only need to follow Mazda to see this but some of us need to find out our selves too

If you want to know my recommendation feel free to PM me and I will gladly pass it on.

Where he is, he doesn't NEED a CDI.... I would like to see what a CDI would pickup on an N/A car.... I would be surprised if it was more than 2rwhp. Don't forget Peter... not all of us are running the serious chamber pressures that you are... and water... last I checked water doesn't ignite very well... perhaps you should give bdc and hc a call and get a lesson in meth... after all... meth is a fuel