Toyota COP dwell curve
 

Here you go peeps:

http://pages.sbcglobal.net/jcuadra/t...OP%20dwell.gif

The internal driver current limits at 10A. The dwell curves shown are for 8.5A and 9A. (Higher curve is for 9A). These are ideal dwell curves if your wiring can deliver full voltage with no voltage drop. (Typically lose 0.2-0.4V avg over the dwell period).

I'd want to run the coils at 8.5A for safety, but given the loss in the wiring, use the 9A curve.

At 8.5A, coil stored energy is 100 mJ.

The '01 coils store 133 mJ at 6A (current limit is at 6.8A). But some of that energy will be lost in the plug wires. I don't know how much. They can jump a longer air gap than the Toyota coils.

The '00 coils can store 170 mJ at 7A but their plug wires are much longer. I don't remember how far they can jump.

By comparison the GM coils on my TEC3 only stored 50 mJ, but I think they are capable of more voltage than the other coils, as they can jump a longer gap in air.

I couldn't find the 1000x "cattle prod" high voltage probe in the lab, have to ask for it tomorrow. I can measure the coils' max output voltage. That'll be fun.

Somewhere I've seen the airgap to voltage table. 0ft altitude, 0% humidity (is that possible?), and 0bar atmosphere. That doesn't really tell how much current was in the ark though. 1kv per mm or something rough like that. I don't know if it was even linear. I doubt it.

These are what us AEM owners have for dwell settings.

Attachment 216056

Man WTF does that mean in dwell time in milliseconds for us MS users?

2.2ms assuming a max 14volt system with me assuming my alternator is amazing. 13.5-8v would probably be more realistic. So maybe 2.3ms would be good for us.

MS assumes 12v and make corrections, at higher voltages the time is reduced; lower, increased.

So it looks like 2.5-2.6ms would be best.

Jason, since the 01 coils are waste spark, does it divide the spark energy in half between the two plugs? so 133mJ becomes 66?

Matt

M@, the energy is divided into 2 plugs, but the "waste" plug gets a small fraction that the other plug gets.

The energy I quoted is that stored in the coil just before spark; I'm not sure how many % is lost in the wires.

The AEM should have implemented a dwell correction vs. MAP instead of RPM. More useful. One could then run the coil "cooler" at lower MAPs.

What coils do Supras use? EVOs? WRXs? These are high boost factory cars.

If anyone can send samples to me I can test if they're any more powerful than the miata and Toyota coils.

how does the waste plug get less juice? the coil has no idea which one is firing...

Jason SBB, thanks for all the work. It's much appreciated. You wouldn't happen to know what the 1.6L coil dwell time should be do you?

Lazzer408, do you know how to calculate the dwell time based on those AEM default settings. The manual give very sketchy details as to how to calculate the dwell time. From what I've read, it means that between 13-14V the aem gives about 1.7ms of dwell time with the stock 1.6L miata setup. The default for the toyota coils gives even less. Would this mean that we can safely increase the dwell time when using these coils.

Less resistance on an empty chamber compared to a full compressed one. Requires less work.

M@, the 2 plugs are in series; the "waste" one is practically a short circuit compared to the other one. The energy is not reduced by the fact that the coil has to spark 2x as often, provided you don't run out of time for proper dwell between sparks.

So what is optimum dwell time for a '99-00 OEM miata coil?

Check this link

http://forum.miata.net/vb/showthread...ght=coil+dwell

[QUOTE=JasonC SBB;138598]M@, the 2 plugs are in series; the "waste" one is practically a short circuit compared to the other one.[QUOTE]

This is because the waste gap isn't under compression.

It would be interesting if you could find means to measure it. The current flow is equal thru any electrical path. This includes wasted spark. The current thru both arks is equal but the voltage's across each will be different. Once the ark is formed the current will flow thru it. It's easier to form an ark in the cylinder with no compression so there's little loss happening.

Thanks for the link. I don't go go .net often. Looks like I should be running more like 3.1 or 4.1 ms constant for my '99 coils then or is that wrong? I have been running more like 2.0 just because i didn't want to fry them. Not sure if I want to use the 4A or 5A value for 12v.

Here's the dwell curve for the '00 coils:

EDIT: the old graph was wrong:
To reach 5.5A:
10V: 5.9 mS
11V: 5 mS
12V: 4.25mS
13V: 3.75mS
14V: 3.3mS
15V: 2.9mS


and for the '01 coils (use the 6A curve):
http://pages.sbcglobal.net/jcuadra/t...ell%20plot.gif

You can find the dwell curves for the NA 1.8 and 1.6 by looking at the Hydra maps.

Anyone have the hydra maps for the 1.6 and 1.8?

he just posted the numbers.. it's easy enough to fill in. btw, FM has started using jason's numbers above in their maps.

Judging by that dwell curve for the '00 i should be running more like 5.2-5.3 on my '99. That is a pretty long dwell and I was under the impression that the NB coils weren't very robust for that kind of current.
So that would be great if I could safely run a dwell of 5.2 and not fry my coils.

Hmm... I must have missed where the 1.6L and 1.8L NA coils are posted. Tried a search, but couldn't find it. I just wanted to be sure as I have conflicting information. The AEM tells me dwell times are in the 1.7ms range, other post have said they are running 5ms for the 1.6L.

I'm suspecting that AEM probably didn't test the coils to come up with optimal values. They may have just taken the stock mazda ecu and put it on the bench to measure what the dwell outputs are. If I can maximize the dwell settings and get a better spark, I might not need to change anything on my car to cure the misfire.

I am running 5ms @ 12v on my MS/1.6. That is the value Jerry @ DIY Autotune scoped out from the stock computer.

With these much shorter dwell times--2.5ms for the toyo coils vs 5.0ms for the mazda coils--does this mean that under higher load it is possible for a hotter spark, as the coil had enough time to completely charge. Let me put it another way, with a dwell of 5.0ms and wasted spark, is it possible that spark events could be called upon before the mazda coil had the required 5.0ms to charge fully?

Wow, if the stock ECU actually has a dwell period of 5ms, with these toyota coils, wouldn't they be overheating? Your charging them twice as long.

You didn't say NA 1.8 :)

I would think there would be a good chance of overheat. Luckily most of us run MS, AEM, or Hydra so not a problem.

Ive put 300 miles on mine on the stock ecu so far. longest trip has been 150miles, I've had 0 problems.

More data. I will type it here for posterity. If I keep it in a notebook I will more likely lose it.

Toyota coils
------------
primary L: 4.7mH below 5.2A, 1.6 mH above
current limit: 10A
delay from signal to beginning of output voltage rise: 6 uS
rise time to 20 kV: 12 uS
peak voltage: 37kV
rise time to 37kV: 48uS


2001 coils:
-----------
primary L: 7mH
current limit: 6.5A
delay: 6uS
rise time to 20 kV: 25 uS
peak V: 25kV
rise time to 25 kV: 65 uS

'00 coils:
--------
primary L: 10 mH
current limit: 6A
delay: 8 uS
rise time to 20 kV: 13 uS
peak V: 36kV
rise time to 36kV: 38 uS

TEC3 coils:
----------
primary L: 3mH
current limit: 5A engine idling with the TEC running
rise time to 20 kV: 9uS
rise time to peak of output: 30 uS
peak output at 4A of current is 35kV
peak output at 5A of current is 40 kV (limit of the high voltage probe)


The TEC coils appear to have a very high output voltage indeed. It seems to have a high turns ratio, and low capacitance. This would also translate to a lower current spark but long duration (with a high peak voltage capability of igniting hard-to-ignite mixtures)

TThe '00 coils are good.

The Toyota coils are good but don't have the same energy storage as the '00 coils.

The '01 seems lousy in comparison, in terms of peak voltage capability and delay. At 7200 RPM, there are 23 uS per degree of timing. If this translates to any real-world performance loss, I don't know.

What is the deal with 97 coils?

Lazzer408,

FYI, I just logged the dwell time on my AEM for the 1.6L default settings. The dwell time is 3.5ms, but it starts to go down at about 4400 rpm. I believe this is because the max dwell is set at 0.5 teeth. So it progressively decreases with the RPM. I'm not so sure why AEM set it like this. Actually I'm not so sure why they would have a max dwell time dependent on the crank angle. It might be the reason why I was getting misfires on the higher rpm boost ranges.

Just to make sure that i've got this all right before i change my settings.
Would a dwell of 4.25ms for 12v be ok to use for my dwell with my '99 coils or is it going to cooke them?

If I understand correctly, these Toyo coils have a smart dwell function that will force a discharge if dwell time is too long, preventing them from cooking due to too long dwell.

If that is the case, should I still use a fixed dwell setting, or should I allow the COP to set its own dwell?

Thats a good question. I originally had 4 COPs that were slightly longer than the ones everyone was using that were running on the stock ecu for about 2 weeks before I started building the ms2 a while back. Well I somehow burn one out testing that out so I ordered another set. These ended up being the shorter ones everyone is using right now.

Here is were the story gets interesting. I only replaced the one COP that burnt out with one of the new ones I had gotten. The stock ecu ran everything great. By this time I had decided on MSI instead of II and had already built what should be a working version. The problem was that I couldn't get it firing on a cylinders. Well I had had enough after messing with it for 2 weeks and just threw the stocker back on. That weekend I went home and hung out with brad and scott while they "installed" his torsen while I messed with MS some more in the driveway. I eventually found out that the old COP fired on inverted and the new ones on normal. Very strange indeed that I was able to run the car perfectly on the stock ecu with that mix without ever killing a single one.

So, I'm not really sure how I was able to kill one with the msII but I was able to run them mixed with 3 of the originals surely charging the entire time they should be off.

No, it doesn't "force a discharge", it current limits, and will still fire normally.
HOWEVER, during current limiting, the ignitor will get hot, and will burn out if you do it all the time.

If it "forced a discharge" it would cause the spark to occur way too early with a possible bent rod.

Thanks Jason. That's very good information, and exactly what I need to know. I've set my MS for 2.5ms@12V. Unfortunately, I can't control the rate of dwell change with voltage change, so it is some what of a WAG as to what's happening at running voltage of 14V

Ben, MS assumes 12v and make corrections: at higher voltages the time is reduced; lower, increased.

of course. but how much correction? we know it changes, but we don't know the rate of change.

look it up in the code? I'll take a gander myself.

Good lord, jason, you rock. Sweet curve. If I were still using stock coils, I would totally care. You don't have those for the toyota COPs, do you? :-)

edit: Sorry, totally missed the FIRST post. Abe McDouche, reporting for duty.

In other news, Mr. McDouche kindly explained post #18 to me. So instead of my recently updated cranking/run dwell settings of 5.0 and 4.0, I'm going to try 6.0 and 4.5 .... right now!

Here is a screen shot of how I modified my AEM EMS Cal for the toyota COP's. I used the formula from the AEM forum as follows:

Dwell time = (0.02 * Dwell Factor * Dwell Vs. RPM * Dwell Vs. Batt Voltage)/1000

I used Excell to figure out how to modify the Dwell Vs. Batt table after using the Coil Wizard to select the toyota 04 up prius.

I'd be interested to see how you guys with AEM's modified your tables to set Dwell. Seems like you could change any one of the 3 variables to achieve the same goal. I also just made an assumption that the curve would follow smoothly below 9 volts.

I have not started the car yet, so I cannot claim this worked yet.

Infiniti COP dwell curves
 

So that all the dwell measurments I made are in the same place, here are the Inifniti COP dwell curves.

Marking says
Hanshin
MCP-1350

It has a 3 wire connector
The COP has labels "IB G +" which refers to input, ground, 12V.

The input requires at least 2.5V and 3.6 mA to that it will current limit at at least 6A. More drive results in higher current limit but it's diminishing returns.
For example it will current limit at 7A if driven to 3.2V / 23 mA.

The input should not be driven without a current limiting resistor. From a 5V drive 220 to 680 ohms will work. From a 12V drive, 1k to 1.5k will work.

Note that some ECU outputs will have a current limited drive and this may need to be considered.

To trigger a spark the input needs to go below 0.7V minimum. A high impedance low state is OK.

The primary inductance is 3.1 mH
The resonant freq appears to be 31 kHz

Here are the dwell numbers for 5.5A peak current (94 mJ primary energy)

8 V 2.5 ms
10V 1.75 ms
12V 1.43 ms
14V 1.2 ms
16V 1.06 ms

At 5.5A of primary current and using a 1 kV zener on the output, secondary energy is ~50 mJ.
Peak secondary current is 55 mA and falls to zero in 2 ms.

With an open secondary peak secondary voltage is 35 kV at 5.5A primary current.
Spark delay is 12 us signal delay, + 35 us output voltage rise time.
Rise time to 20 kV is 13 us.

If someone can paste the coil driver schematics of the MS I can tell you if the Infiniti coils are compatible.

I didn't know compatibility is an issue. Are you saying that if they're not compatible, I can work around that somehow ... or that I'd be S.O.L. and wasted my money on the coils and hardware? :(

Worst case you mod the MS circuit to make it work. Schematics, anyone?

Schematics are here: MegaManual Index ... I'm not sure which is the correct one. (DIY built mine - v. 3.5.7)

Update - I called DIY, they said the coil driver is on drawing 7 (of 8) at V3.57 Main Board

Should be fine.
So should Joe's modded circuit here:
https://www.miataturbo.net/forum/t33964/

Great! Those "three little words" are the last bit of data I need to move forward with the COPS. :yippee:

Big public "w00t!" to JasconC.

Thanks. What did you find with using the '01 COP shafts?

Not sure what you mean. The Infiniti coils stick up about 1/2-inch too far, but as ya'll said, that is not the end of the world. I'll see about putting some kind of spacer / o-ring combination around the exposed part for extra protection against debris and moisture.

Just to add to the above in one location for posterity's sake, here are test results of the Infiniti coils provided by Evank.
They had long skinny shafts and a larger body than the Toyota COPs, and are a bit clumsy fit in the 99 cam cover, seemingly hopeless for the 01 cam cover. Their markings were "Hanshin MCP1350", no Nissan logo or part number
Brief disucssion of fitment was here
https://www.miataturbo.net/megasquirt-18/01-cops-ms-32827/

In my own language lol

Input is proportional (e.g. not comparator’ed / schmitt’ed)

Current limit can be 6-8A depending on drive

e.g.
5A if 2.2V @ 1.7 mA
6A if 2.5V @ 3.6 mA (not recommended to go below this drive)
7A if 3.2V @ 23 mA
7.7A if 4.2V @ 53 mA

drive requriements, assume 5 mA @ 2.5V (output impedance with 5V drive <500 ohms)

Off drive minimum is ~0.8V (drive output V must go below this)

Delay time, trigger fall to start of steep rise in HV – 12 us

With large currents, rise time of HV is 9us

Resonant frequency is 31 kHz


6A Lab supply with 10mF bypass testing

Cur lim = 6.5A
Initial inductance is 3.1 mH
Current rise is very straight.

Dwell times, for 5.5A primary peak current: (94 mJ input energy)

8 V 2.8 mS
10 V 1.78 mS
12 V 1.44 mS
14 V 1.17mS

secondary waveform testing
With 1 kV TVS load

5.5A pri ~55 mA peak sec current, 2 mS ramp down time, ~55 mJ.

(calc est 100:1 turns ratio)

2A primary peak current yields 13 kV pk, unloaded
3A 20 kV
4A 25 kV
5.2A 32 kV @ 35 uS rise time

Fry your COPs yet ?
 

Speaking of dwelbs :bowrofl: Did you guys fry your COPS yet ?

I was running mine for a while on the stock ECU without a dwell reducer.

Recently I went back to using the MS and while switching MSQs on my the motor immediately started missing.

After switching to the OEM coils back again everything was fine then I tried the COPs again and two are not firing

The wiring and connections are fine far as I can tell

Are you sure it wasn't your MS setup incorrectly causing them to fry?

Very likely that I switched to an MSQ that had the dwell set to 5 instead of 3.5

that wouldn't kill it. (it should be 2.5ms anyways).

I'm thinking more along the lines of the spark output set to inverted instead of normal or something.

I say that can kill em, if run waste-spark, at sustained high RPM, high temp operation.

Maybe all the MSQ switching corrupted the file and told the MS to make the coils do something obscene.

In any case - what is a good place/price for toyota denso 90080-19015 ?

+1. I don't have the .cal files handy but I blew a few coils early last year after switching to the AEM - I had a 190% modifier on the dwell page that I glossed over, which brought the actual dwell up to the ~6ms range. I am still on wasted spark. I popped one coil at MRLS, another right after that, and finally figured the problem out.

Having said that, at least two of the coils in my car lived through that and appear to be just fine.