OK Brettus started a thread in Ignition Timing that spawned a crap fest on Dwell settings. I was actually starting to get into the whole discussion but it was distracting the original threads intention.

I thought I would start a thread to continue the dwell discussion.

So to start with,

Kane,

I work with Ignition computers as I'm involved in conversion of Diesel engines in to NGV (Natural Gas). In the software I use to program the stand alone ignition computer, the dwell value can be adjusted in - % percentage. Basically 100% would mean dwell will be same duration as the spark duration. 110% means the coil will take 10% longer to charge and then provide spark with the same duration.

Looking at my SPT dwell map, could the map be battery voltage vs % time of dwell? That would mean 1208 might be 120.8% and 395 might be 39.5% of spark duration.

And then

Wow. There it is. (EDIT - Sorta. Its not a direct % of available spark time, but crank rotation time.)

So after months of speculating and even erroneous input from people who are actively involved in RX-8 calibration testing and development, it took a professional ignition calibrator's input to figure out what I managed to figure out in my garage in 10 minutes with a cheap scope meter.

I'm not picking on you, a_ahlan. You are just sharing what you know.
But its just sad that all of the whiners in this thread just couldn't be bothered with the R&D, just because they knew someone had already figured it out.
It was more worthwhile to complain about it. :icon_no2:

So now we believe/understand that the dwell table shows % of crank rotation time I am curious what people are running with what coil setups.

We have 3 common setups that I am aware of OEM, LS2 (BHR) and Bosch HEC 715 (MazSport). I am curious what people are runnning.

Kane posted some in the Timing thread:


http://www.rx8club.com/attachment.php?attachmentid=137711&stc=1&d=1240679475

http://www.rx8club.com/attachment.php?attachmentid=137710&stc=1&d=1240679475

EDIT: Updated this to a more recent map that is working much better
HEC715
http://www.rx8club.com/attachment.php?attachmentid=144582&stc=1&d=1251463576

I deleted the dwell thread I started because there was too much misleading info in it .
From that thread I learned a few little titbits but still do not understand how the mazda map relates to the actual dwell at the coils .
I did get this gem of info from someone though :

It is so unbelievably simple that a trained monkey could work it out . However no-one apart from MM ever did , so maybe you actually need to be one :dunno:

i love this community.

cool thread. lots of questions.

you're not that far from stock with the HEC. you have a shorter charge time towards the top, right?

what effect does charge time have on performance?

is a longer charge time in the later rpms better?

is a longer charge time tougher on coils?

the bhr coils have a long charge towards the top, right?

can dwell be tuned for performance without changing coils?

I deleted the dwell thread I started because there was too much misleading info in it .
From that thread I learned a few little titbits but still do not understand how the mazda map relates to the actual dwell at the coils .
I did get this gem of info from someone though :

It is so unbelievably simple that a trained monkey could work it out . However no-one apart from MM ever did , so maybe you actually need to be one :dunno:That just makes me sing the theme song for a TV show here called "Are you smarter than a 5th Grader" (replace 5th grader with trained monkey).

i love this community.
It has its moments!

you're not that far from stock with the HEC. you have a shorter charge time towards the top, right?yes. BUt that doesn't mean its what works best. In fact its one of the reasons I started this thread was to find out what people were doing with them.

what effect does charge time have on performance? I will start by saying I am the last person to be answering this but my first thoughts are that its like a camera flash, it needs to charge long enough to have sufficient power to light the bulb. If the charge time isn't long enough, the spark may not be strong enough to light the fuel. Not sure what happens if you charge to long.

is a longer charge time in the later rpms better?This is one of the things I am trying to figure out...

is a longer charge time tougher on coils? Good question, not sure.

the bhr coils have a long charge towards the top, right?In comparison to the OEM ones in NA form yeah. Not sure how they would be configured in FI form. I am making the assumption here that the same coil should be different NA v FI.

can dwell be tuned for performance without changing coils?No idea about performance but regardless of the coil on the car you can change the dwell settings.

Cheers

Andrew

No one got any feedback on the dwell of the Bosch/Mazsport Coils?

cool thread. lots of questions.

you're not that far from stock with the HEC. you have a shorter charge time towards the top, right?

what effect does charge time have on performance?

is a longer charge time in the later rpms better?

is a longer charge time tougher on coils?

the bhr coils have a long charge towards the top, right?

can dwell be tuned for performance without changing coils?

Ooooo good questions!

To establish a basis, my understanding is dwell is the overall charge time where the coil generates a magnetic field. When this field collapses it creates a discharge which is what lights the electrode on the spark plug. Someone stop me if I'm off base.

Different ignition coils have different dwell operational ranges. It mostly depends on the vehicle application. The longer the dwell the larger the charge. There are extremes on both sides.

If a coil isn't given enough dwell time it's not able to form a strong enough magnetic field which results in a weak spark. With this you're looking at incomplete or no combustion at all. On the flip side, there is a point where excessive charging doesn't result in a more powerful field and instead it's turned into heat energy.

Excessive charging will overheat the coils and shorten their life. Also you only have so much crank rotation time before you have to fire off the spark plug.

Longer dwell times only become an issue in the upper RPM band because as the motor spins faster you have less time to charge the coil.

Spark discharge needs to be its strongest at the torque peak because this is where combustion chamber pressures will be at their highest.

Dwell can be adjusted without changing the coils, however it will depend on the design of the coils in question as to what dwell settings are effective.

The coil properties should not be any different from NA to FI - they should be what the coil wants / needs to make maximum energy without burning out.

Having said that - I have confirmed the dwell test in the TSM - and OD got me this info... so it looks like I am right for wrong reasons maybe... but I really don't care about the why - only that it works.

Hey Dude--dont know if you need this or not but here it is
Volts Dwell
8 9.5
9 7.7
10 5.9
11 4.9
12 4.2
13 3.7
14 3.3
15 2.9
this is oem of course
Denny /olddragger

And again BHR needs 5ms of dwell.

This jives with the Service Manual testing procedures. 3.3/5 or 3.7/5 = 135 to 150% increase in dwell..... I don't know why this is so mysterious.

Kane, when I did my dwell map i noticed that on yours you decresed the dwell at 1,000 rpms from the stock dwell time??

I don't think so.... I may have made a mistake.... Hmm.

Looked it up and you are right.... I wonder how that happened.

Can someone please show me how to adjust the BHR coils to 5ms of dwell using AP? I'm not much of a tuning guy.

/\ that my freind is what all the fuss is about ....

You guys are fuckin killing me....

It looks like this has been solved, then? Yes?


I take it there is no longer need to dwell on this issue? Wow, that never gets old.

Good one.

is there any concern on longevity with increasing dwell 135-150%?

in the first post, how come the bhr coil values are not 135-150% of the oem dwell?
Where are these values from?

is there any concern on longevity with increasing dwell 135-150%?On stock coils? I have been told, yes. You will burn them out and there is no reason to do it anyway.

On Yukons? Again, this is the core of the whole issue, but Yukons are fine out to 8ms, I think, where they have an internal quench or cutoff.

It looks like this has been solved, then? Yes?


I take it there is no longer need to dwell on this issue? Wow, that never gets old.

does not look solved to me but i've always been a "gotta know how stuff works" kinda guy .
Did MM threaten to burn down your house if you divulged anything ?

You guys crack me up.

Take a step back and look at what you are trying to do.

It is genuinely so simple, its sickening.

yup, so easy that I updated my dwell settings yesterday.

Interesting. what does 5ms mean? Sorry I just don't know even half the stuff about coils that other people do.

yup, so easy that I updated my dwell settings yesterday.

lol!

You cheated!

if you're not cheating, you're not trying hard enough.

does not look solved to me but i've always been a "gotta know how stuff works" kinda guy .
Did MM threaten to burn down your house if you divulged anything ?

lol. You give me too much credit. I know just about as much as you do. The only thing that I know more than you is that I know that all I know is no more than you know. :SHOCKED:

Like I said. Once I can confirm what I learned from MM, I will be happy to share. Stick it to the man, so to speak. Until then, all I could say would be hearsay.

Besides, the whole drama is entertaining. Like "Days of our Lives" 20 years later I fully expect Bo and Hope to be almost getting married. And Sammy or Stephano trying to kill everyone.

I've got to the stage I don't want anyone to tell me because it would probably just piss me off .

I've got to the stage I don't want anyone to tell me because it would probably just piss me off .

Probably.

sorry to jack the thread, do you guys remember what is the dwell setting for stock coils if we are using INT-X? i forgot, i left it 2.0, because when i used mazsport, it was 3.5. can anybody confirm this, please....?
steven

Bit hard the jack a thread that has died :) Not sure about stock, but can confirm that 3.5 is what is used with the MazSport coils (ie the HEC715).

Cheers

Andrew

So just to bump one last time, does anyone have a dwell map for the HEC715's?

(DNEW/DOEM)*DOLD = DDESIRED

Christ, you people are dense...

I dont get it!

what I think he is saying is

OEM = 3ms and desired is 3.5ms so multiply all numbers on the table by 1.167

which makes sense but still does not answer the question of how the numbers in the table are derived ...

Base dwell setting (not in ATR) + number from Ignitor Dwell Time (dwell compensation table). Base dwell setting = 2.0ms.
1208 at 12.75 volts/idle = 1.2ms for 3.2-3.3ms total dwell at idle.
2.4ms dwell at 9000RPM.

Since when are the Mazsport coils Bosch HECs? The connectors for Type A (715) and Type B (716) aren't right, nor is the resistance ratings for those coils (but they are close). The Mazsport ones are .5ohm primary resistance and 7500 ohms secondary.

Base dwell setting (not in ATR) + number from Ignitor Dwell Time (dwell compensation table). Base dwell setting = 2.0ms.
1208 at 12.75 volts/idle = 1.2ms for 3.2-3.3ms total dwell at idle.
2.4ms dwell at 9000RPM.

.
Haleluya !!!
I can sleep a night now that I know this . Funny thing is , I think we came up with that concept in the other thread but went off the idea for some reason ...

what I think he is saying is

OEM = 3ms and desired is 3.5ms so multiply all numbers on the table by 1.167

which makes sense but still does not answer the question of how the numbers in the table are derived ...
Brett, I was stirring him up.

You answered one of my questions though and that was what OEM was. I actually saw that as:

(3.5/3)*1206 = 1407

This is the same as what you said but you said "desired" instead "new".

As for the HEC/MazSport I will apologise if they are not HEC715's. I was going off something someone who supposedly had dealt with them had said.

But if Atl8 is correct then what MM said is not ....

your figure would be

1206 +2000 = 3.206ms and you want 3.5ms
so the new number is 1206+(3500-3206)= 1500

Which one of you has actually adjusted the dwell with a scope?

sigh. I give up.

Well I haven't , but you have had such fun messing with me/us on this I find the logical information presented by atl8 quite believable .

I've "messed" with no one.

The calculation is so stupendously simple, its nauseating.

Its like someone asked how to get to the second-floor office and everyone rushed in to give answers that included ballistic trajectory calculations and GPS coordinates and a whole lot of other BS.
Meanwhile, I just said "take the stairs".

(DNEW/DOEM)*DOLD = DDESIRED

I took that to mean:
DNew = Base Dwell of new coils (3.5ms for the HEC's)
DOEM = Base Dwell of OEM coils (3ms from what I saw stated)
DOld = Old value in the dwell table
DDesired = New Value to put in the Dwell table.

Did I get it ? Or are these just ballistic trajectory calculations and GPS coordinates?? :fingersx::fingersx:

i love this thread! three days away, i am laughing out loud already. :)

beers :beer:

3 Days away from what?

I went with (3.5/3)*1206 = 1407 and it seems to be much better so far. Will leave it for a week and see if anything comes up and then will give the other a go.

Cheers

Andrew

So I guess you turn the coils right ? :P

huh?

3 Days away from what?

I went with (3.5/3)*1206 = 1407 and it seems to be much better so far. Will leave it for a week and see if anything comes up and then will give the other a go.

Cheers

Andrew

3.8ms might be even better - saw that figure mentioned on another thread .
What differences did you notice ?

3.8ms might be even better - saw that figure mentioned on another thread .
What differences did you notice ?
Pulls MUCH MUCH stronger and reliably now! I think I understand what one of my issues was now to. Only taken it up to 7,000 rpm as I am still way to unstable/lean there with my intake setup. Can you point me to that thread?

here ya go

http://www.rx8club.com/showpost.php?p=2956656&postcount=1622

OK I am now questioning if the MazSport ones are HEC715. Can anyone confirm this? If they can I will give it a go.

Oh wait..... deviation from a known standard.... yep - I was wayyyyy off.

***Pokes self in eye***

Oh wait..... deviation from a known standard.... yep - I was wayyyyy off.

***Pokes self in eye***

The point for me is that after all the BS it still seems that no-one (including MM)actually knows what the numbers in the table mean .

You don't need to know.... hell they could be stored in Hexdec or Binary for all we know (and I'm too lazy to find out) - if it makes the translation easier for the PCM.... kinda irrelevant.

You don't need to know.... hell they could be stored in Hexdec or Binary for all we know (and I'm too lazy to find out) - if it makes the translation easier for the PCM.... kinda irrelevant.

True . I guess if MM has scoped the changes and got the results he expected then it is irrelevant .

...scoped the changes and got the results he expected then it is irrelevant .

Quote of the DECADE.

No, if everyone would just apply this to everything, we would be in much better shape.

I know this is an Internet forum and, therefore, way over-stocked with geeks, nerds and math-tards, but we are talking about a car and all the incredible approximations that go into making it what it is.

Figuring out the math is not about coming to conclusions. Its about setting a baseline upon which you can build. It doesn't matter what the actual numbers are as long as you can repeat the effect.

Figuring out the math is not about coming to conclusions. Its about setting a baseline upon which you can build. It doesn't matter what the actual numbers are as long as you can repeat the effect.

I like this one personally.

Quote of the DECADE.

No, if everyone would just apply this to everything, we would be in much better shape.

I know this is an Internet forum and, therefore, way over-stocked with geeks, nerds and math-tards, but we are talking about a car and all the incredible approximations that go into making it what it is.

Figuring out the math is not about coming to conclusions. Its about setting a baseline upon which you can build. It doesn't matter what the actual numbers are as long as you can repeat the effect.

I agree completely . But from the very start you threw a whole lot of red herrings out there that gave me the IMPRESSION that you knew what the numbers meant and how they related to rpm etc . This is why your final revelation did not go down well with me ...

I agree completely . But from the very start you threw a whole lot of red herrings out there that gave me the IMPRESSION that you knew what the numbers meant and how they related to rpm etc . This is why your final revelation did not go down well with me ...

I do know what the numbers mean. Absolutely. There were NO red herrings.
But it is totally useless information, apparently. Because, when I explained it, all I got was puzzled looks from everyone involved.
It wasn't that they were being misled. Its just that they were being led to somewhere they hadn't seen before and couldn't make believe they understood the path.

People don't want to understand things. They want the things they don't understand to be changed and conformed into things that they do understand.


This is the real reason I don't like explaining things to people - all they will do is argue until the thing that was explained is made to look like something else that they can explain away.

I ,for one, DO like to understand things . You could accuse me of being too lazy to go and get dwell scoped for myself which would be fair .
But I did not really have any compelling reason to do that other than - I might need the info one day if I go for higher boost etc.
Could you point me to your previous explanation and I'll try see if it makes more sense to me now ...

I will second that Brettus. I absolutely want to understand things. I tried a while ago to understand a few things but got shot down for it... wrong question in the wrong thread!

The problem I see is that not everyone understands the same things in the same way. Simple as it is, I had never thought of this problem in this way "deviation from a known standard"

The problem I see is that not everyone understands the same things in the same way. Simple as it is, I had never thought of this problem in this way "deviation from a known standard"

Deviation from a known standard would not have worked if the theory put forward earlier in the thread were correct ( about a 2ms Base with a modifier table) . That is why it is dangerous to assume that theory unless you have evidence to back it up . IE scoping the results before and after .

Deviation from a known standard would not have worked if the theory put forward earlier in the thread were correct ( about a 2ms Base with a modifier table) . That is why it is dangerous to assume that theory unless you have evidence to back it up . IE scoping the results before and after .
As a senior software developer and mentor I have always followed and always teach the you should never leave something saying "That works but I have no idea why". With the RX-8 and flash tuning I have no access to the ECU source code and have zero experience with engine tuning. I am reliant on others experience and my own experience in what seems right.

I had never seen or read about the 2ms base theory until post #34 from atl8. But I have to say that it doesn't sound right to me as its way to variable for something that is very precise. On the other hand, a_ahlan's post makes very good sense, seems very logical, is backed by experience and findings from MM.

Reading back on a_ahlan's post I now have a question about spark duration :)

Cheers

Andrew

I had never seen or read about the 2ms base theory until post #34 from atl8. But I have to say that it doesn't sound right to me as its way to variable for something that is very precise.

Cheers

Andrew

I did a little test on that theory a few months back (but forgot about it when alt8 brought it up again) . Set dwell to 0 at 2000 rpm just to see if the engine would run ................. It died as rpms approached that rpm so I dismissed the theory .

I did a little test on that theory a few months back (but forgot about it when alt8 brought it up again) . Set dwell to 0 at 2000 rpm just to see if the engine would run ................. It died as rpms approached that rpm so I dismissed the theory .
Would love to know what it was based on.

As a senior software developer and mentor I have always followed and always teach the you should never leave something saying "That works but I have no idea why".

Which is why you and your apprentices will be sitting around, fiddling with minutiae while I'm out blasting around town in my car, enjoying the power.

I often leave things that work alone. There are plenty of things that don't work that you can geek about.

I'll even go one further -

If a man-made device works, why spend extra machine cycles trying to reverse-engineer it if you can just ask the guy that built it how it works?

But - be prepared for him to give you an answer like "It just does".

There are a surprising many things in this world, built by experts, that just work because they do.

Sure, you can dissect the physics, but you will be no closer to any true understanding once you have.
Much of it is incidental.

The problem I see is that not everyone understands the same things in the same way. Simple as it is, I had never thought of this problem in this way "deviation from a known standard"

Deviation from a known standard would not have worked if the theory put forward earlier in the thread were correct ( about a 2ms Base with a modifier table) . That is why it is dangerous to assume that theory unless you have evidence to back it up . IE scoping the results before and after .

Well a Known standard would be information that was not a educated guess... like the field service manual that lists what the dwell time is; I know I have posted that picture more than once.

"Perfect is the enemy of good" and all that jazz.

I'll even go one further -

If a man-made device works, why spend extra machine cycles trying to reverse-engineer it if you can just ask the guy that built it how it works?

But - be prepared for him to give you an answer like "It just does".

There are a surprising many things in this world, built by experts, that just work because they do.

Sure, you can dissect the physics, but you will be no closer to any true understanding once you have.
Much of it is incidental.
There are also a great many things in this world that we take for granted purely because people have said why does something do what it does.

I can sleep at night knowing that when I have understood something, there is a much lower chance that what I did wont result in unexpected outcomes.

There are also a great many things in this world that we take for granted purely because people have said why does something do what it does.

I can sleep at night knowing that when I have understood something, there is a much lower chance that what I did wont result in unexpected outcomes.

Amen to that .

I also see merit in what MM is saying about the "just do it" attitude because if we always had to know the answers before we did stuff then we would never do it in the first place .

Amen to that .

I also see merit in what MM is saying about the "just do it" attitude because if we always had to know the answers before we did stuff then we would never do it in the first place .
Sure...sometimes before you can actually figure something out you need to find what works...then you can start to theorise, etc. Been through one engine already...would now rather ask questions first ;)

Base dwell setting (not in ATR) + number from Ignitor Dwell Time (dwell compensation table). Base dwell setting = 2.0ms.
1208 at 12.75 volts/idle = 1.2ms for 3.2-3.3ms total dwell at idle.
2.4ms dwell at 9000RPM.



Sorry if I'm being a total ignoramus (I'm trying hard not to be!), but I'm really trying to figure out where my math is wrong..

According to the factory NA dwell map posted by auzoom in the first post, the value in all the fields at 9k rpm is 395. I divided 395/9000 = .04, then add 2, and it's 2.04ms dwell at 9k, not 2.4.

Sorry if I'm being a total ignoramus (I'm trying hard not to be!), but I'm really trying to figure out where my math is wrong..

According to the factory NA dwell map posted by auzoom in the first post, the value in all the fields at 9k rpm is 395. I divided 395/9000 = .04, then add 2, and it's 2.04ms dwell at 9k, not 2.4.

alt8 was saying it's 2 + 0.395 = 2.4

from the above discussion it would seem that this theory is incorrect however ........

In his example if you take the line of 12.75v, at idle the table shows 1208. He then believes there is a base dwell of 2000.

So you have:
1208 + 2000 = 3208 (3.2ms)

For 9000rpm By that logic it would be

395+2000 = 2395 (2.395ms).

I think its wrong though. Kane has shown the manual stating the base dwell is 3ms for the OEM coils.

I am running strong now using the "deviation from a known standard" calculation. Not that its hard to figure but I will post up the Dwell map I am running when I get home tonight.

EDIT: Geez I am slow at responding! ^^ what Brettus said.

Cheers

Andrew

alt8 was saying it's 2 + 0.395 = 2.4

from the above discussion it would seem that this theory is incorrect however ........


In his example, he cites 1208/1000 (1208 is derived from the map value of 12V @ 1k rpm). This gives you 1.2ms, which is then added to the base value. Right? Ugh. Nevermind.

Is there any evidence of this 2ms base dwell time, or is it just an assumption?

Is there any evidence of this 2ms base dwell time, or is it just an assumption?

seemed liked a pretty logical theory at the time but there is no evidence i know of to support it

As promised. Dwell I am now running using HEC715's.

Cheers

Andrew
http://www.rx8club.com/attachment.php?attachmentid=144582&stc=1&d=1251463576

I dunno exactly what your target dwell was there, but you did it wrong.

Are you talking across the board or just the 6000rpm and above?

As promised. Dwell I am now running using HEC715's.

Cheers

Andrew
http://www.rx8club.com/attachment.php?attachmentid=144582&stc=1&d=1251463576


Uhhh, is that 3D graph accurate? :confused:

Are you talking across the board or just the 6000rpm and above?

Across the board.
You've called for a different dwell time at different RPM points and most of them are under 3.6ms.

(DNEW/DOEM)*DOLD = DDESIRED
I took the OEM dwell table, I then took the Bosch base dwell of 3.5 and divided it by the oem base dwell (3ms) (3.5/3 = 1.67) and multiplied the dwell table values by this amount.

What did I miss?

Cheers

Andrew

What did I miss?

Cheers

Andrew

didn't you say you only adjusted the dwell from 6000rpm upwards ? I think that is why MM is ?

So far I have only adjusted 6000+ to 3.8 as the change to 3.5 up to there was fine. It was struggling above there though.

Bosch dwell of 3.5 divided by OE Dwell of 3 equals 1.16.
Just multiply the whole table by 1.16.
Sheesh.

Sorry, you are right. I miss typed. I had 1.16666666667 and I think I went cross eyed :).

Other than the difference between 1.1666666666666666666666666666667 and 1.16, that is what I have.

So basically the BHR Coil Dwell should look something similiar to this

148321

Looks close to mine - what was your M value?

Looks close to mine - what was your M value?

Yeah 1.666


Offtopic: Hey just noticed your location changed, Welcome man. I think there is a meet going on this weekend I'll hit you up on pm w/the details. :

Yeah 1.666


Offtopic: Hey just noticed your location changed, Welcome man. I think there is a meet going on this weekend I'll hit you up on pm w/the details. :

I did 1.45 per Ray's advice on this new build.... either way they have an internal safety of 8ms...

Cool man; I'll need a ride though - or I'll be rolling in a buds busted 240...HAHA

Hmmm interesting Imma play w/the 1.666 & watch it across the board & adjust from there

I did 1.45 per Ray's advice on this new build.... either way they have an internal safety of 8ms...

Cool man; I'll need a ride though - or I'll be rolling in a buds busted 240...HAHA

I just want to state for the record that this thread sucks ass. My brain still hurts and I've read the damn thing 3 times in as many days.

The best advice I've gotten out of it was to follow Jedi's example and grab Jeff by the neck! :scratchhe

----------

So somewhere between mid page 2 and the end of page 3 I thought I had it all figured out.... Really simple shit for my LS2 coils too...

LS2 base dwell = 5ms
Stock base dwell = 3 ms

5/3 = 1.66666....

Open up ATR, go to the dwell table and multiply the whole thing by 1.666666

Then Kane's comment, apparently recommended by Ray to multipy by 1.45 instead?

WTF people?

To make matters even more confusing LS2 coils apparently have their own termination, assumedly for self-preservation, so then I can't help but wonder why we don't just max out the friggin table!

Have the BHR guys posted a dedicated dwell thread for the BHR coils that I simply haven't been able to locate? Can someone link to it if so?

The idea is you want the shortest dwell possible that doesn't missfire. The same way you want the lest fuel possible that doesn't run lean.

The Mazda service manual has the dwell at approx 3.8ms dwell at idle and the LS2 is at 5....so if 1.45 gets then great, if not do 1.66.

This is the table from another source.
Volts Dwell
8 9.5
9 7.7
10 5.9
11 4.9
12 4.2
13 3.7
14 3.3
15 2.9

I originally ran the +66% to hit 5ms. After some chatting with Ray I went down to +45% and it worked just fine. Since switching to much richer tables I just did +55% across the board and called it a day. :)

I'm at +45% now and digging it. May revisit in a couple weeks when I put on this bolt on.

Knowing stock 3.8 helps a bitunderstanding wise. I missed that.

Thanks Kane, Jedi, and Flashwong :)

In the case of the BHR Coils

Soooo scaling the entire ingitor dwell time table isn't hard...
I've have 2 tunes saved now ... one scaled at an m value of 1.66 (66%) and another at 1.45 (45%)

Neither of these look anything like the BHR Dwell table shown in the first post by the OP of the thread. I'm guessing this should be updated or deleted for clarity?

Kanes statement in post #91 confused me slightly just when I thought I had it all fingered out (yes I said fingered)
Understanding that you want the shortest dwell time possible without a miss fire, as mentioned already, Mazda aims for 3.8ms dwell (not base dwell, correct?) and the LS2 coils aims for 5ms.

Confirmation of math and values:
To arrive at these values you add the value on the dwell time table to 3.0 ms?

Example OEM = @ 12.75 volts at 1000 rpm dwell time is 1208 .... therefore 1.208ms + 3.0ms = 4.208ms

Example BHR +45% dwell = @ 12.75 volts at 1000 rpm dwell time is 1752 .... therefore 1.752ms + 3ms = 4.752ms

Example BHR +66% dwell = @ 12.75 volts at 1000 rpm dwell time is 2013 .... therefore 2.013ms + 3ms = 5.013ms

Holy crap I think I get it.
Yes?

I can't believe this thread persists.

The OE table yields a dwell setting of 3.5ms. Everywhere.
That is all the numbers mean from a practical standpoint.

If you don't believe that, take a scope to the coil trigger and see for yourself.

Now divide your target dwell by the OE dwell (ie 5ms/3.5ms=1.428) and multiply the WHOLE TABLE by this number.

Done.

^^^^ AAAAAAA HA!!!!
I'm glad I asked!

So it's not 3.0ms its 3.5ms, I was never completely confident that I was using the right value.

Thanks MM that's what I did, except I used the wrong base value.
Honestly I tried to find it in the Mazda Shop Manual.

Done and DONE.

+1 done

Man...Reading threads like this make me realize how much im going to poo on myself when I try and tune my car. I will definitely be creating a thread for that. It will be called, Noob with accessport = ???!

I can't believe this thread persists.


bumping this (as well as starting controversy) just for you, pal :lol2:

I've since concluded that using a multiplier on the existing Mazda dwell map, while it is easy to do and works, is not really ideal.

instead, I took the attached factory D585 coil dwell table and using the same multiplier concept came up with a new dwell table that closer approximates it in a general manner along with consideration for the Renesis output characteristic and rpm limit considerations. My reasoning is that IMO the Mazda dwell table is a bit off in this regard.

Team puts flame suit on

The Mazda dwell table has built-in "error correction" of sorts that you are defeating by modifying the scaling to be more mathematically correct.

I verified the real-world dwell values using a scope.

Well there's no doubt that it does this in several areas, but so far my data is not suggesting this for the dwell :dunno:

How did you measure the actual dwell?

It should be possible to figure out the correct scaling for this table. Most table references in Mazda roms actually define the "real world" scaling and offset.

For whatever reason, Cobb chose not to use them in all cases.

For example, here is a fuel table with Mazda's scaling:
http://www.rx8club.com/attachment.php?attachmentid=177034&stc=1&d=1316225701

Closed loop targets make more sense with the Mazda scaling, IMO:
http://www.rx8club.com/attachment.php?attachmentid=177036&stc=1&d=1316226414

Unfortunately, the dwell table is one of a few which use a different data structure with no scaling, but here it is with the same scaling as the fuel table. 0.00390625, which is used for about 40 tables, and also for Subaru dwell tables in 32bit ecus. (in ms.) Make any sense?

http://www.rx8club.com/attachment.php?attachmentid=177035&stc=1&d=1316225701

Oh, duh. 0.00390625 = 1/256 = .5^8

Looking at the code, all this table seems to be used for is to set a 16-bit down-counter for one-shot pulse generation, pretty straightforward compared to other subroutines.

I thought there was 6.6ms of time between ignition events @ 9000 rpm?
if that last table is a delay then wouldn't the stock dwell be 5.12 ms?

No, I think those are the actual dwell times. IOW, multiply your desired dwell in ms by 256 to get the value.