muythaibxr

At this point no chance. We're bugfixing the 3.1.x ms2/extra releases. All new dev is ms3 only.

I'm sure COP would help that. I'm unaware of any changes between 2.1.x and 3.1.x for rx8 mode, but I can check/ask James if he knows of any.

Ken

Jobro

iTrader: (2)

How bad are the release versions of 2.1.0D and 2.1.0E?

muythaibxr

Not sure what you mean by "how bad" are they.

I don't think any of them are bad, but I'm probably biased.

They have all worked for me.

Ken

Jobro

iTrader: (2)

By bad I meant in terms of known bugs.

Jobro

iTrader: (2)

I think my turbo has compromised my TPS because they are about 10cm seperated.
I'm also getting incorrect readings from my MAT sensor. I had that one out last weekend so its possible it has died as a result...

I have 2.1.0E release in the car I'm just generating a meaningful MSQ now.

muythaibxr

I believe those realeases were small bug-fix releases, so neither of them have a lot of known bugs.

Jobro

iTrader: (2)

Cool.

Well I have a no start condition. It did sound like it tried to start. It has very tired plugs in it and it was cranking very slow. So I put it on trickle charge to attack tomorrow.

I might put my new plugs in it tomorrow and zero the timing.

Cranking RPM was really low, like 110RPM.

Jobro

iTrader: (2)

At this stage I'm 99.99999% certain labels of triggering functionality has changed between 210D-B2 and 210E.

Jobro

iTrader: (2)

If I keep the same trigger settings I've always used in the 2.1.0D beta I was using,

ie

2nd trigger active on every rotation of: CAM
24 teeth on trigger wheel on crank.

Then it works and I've been driving fine for 2 years.

If I put those settings into the release code of 2.1.0E then I get half the RPM and the car won't start.

Even when I make it 12 teeth on crank I get the right amount of ignition events at seemingly the right time on all 4 wires but the car still won't start with the exact same tune cross loaded into the new MSQ format.

Frankly I've pretty shitty. I've kept everything constant and the new code doesn't work because something has changed and I don't know what.

Can you offer any insight ken?

Jobro

iTrader: (2)

I will load up the MSQ file for the 210E release code.

I can't see how you can change fundamental things like that and not know. I can't see how you change make new code that is functionally different without a change log and not expect people to get cranky.

Jobro

iTrader: (2)

OK,

I no longer have the *****.

I was correct in my original obervations.

The 2.1.0E requires

24 teeth on CAM
reset pulse on every crank.


The old software I was using required

24 teeth on crank
reset pulse on every cam.


That is the obvious discrepancy causing my car to not run and using my entire weekend...

Jobro

iTrader: (2)

Gotta say I'm pretty impressed!

#1 I tested FC, FD and RX-8 modes tonight and they all behave as expected in 2.1.0E release firmware.

#2 I tested the effects of dwell battery compensation, and it works as expected.

The extra noise / hair on the RX-8 mode scope capture is because by that stage I had cranked the car probably 15 times and the battery was getting tired.

I'm very glad this is heading in the right direction now

I still need to build the 4th LED driver / pullup circuit to test the Remaining rear rotor ignition signal. I bought some 2N3904's to do that next weekend.

muythaibxr

Actually, the correct setting for 2nd trig has been every rotatation of the crank since the beginning! But I am glad you found the problem.

Ken

chango de naranjito

can I install js11 directly to the ignitor for spark D with a pull up or I need the led driver mod to run the igniter on a 3.57 board?

Jobro

iTrader: (2)

You will need the LED driver board mod to js11 minimum.

Next problem is the 4k7 pullup resistor being pulled towards 5V, ~1mA current on a low to high transistion.

1mA is pretty

If you look at Haltech's and Motecs they have options like 20mA drive and 40mA drive. I want something similar to this.

There are 4 channel line driver IC's available that can drive high by 32mA and drive low by 64mA. I think I'm going to end up building the led onto JS11 and then use the pullup and the negative lead to drive a higher current line driver, which will then in turn drive the ignition system.

I have in the past unsucessfully used 270 and 330 ohms pullups to drive VB921 ignition chips before. My feeling is the drive is just not there.

chango de naranjito

in the manual says you can conect it directly to js11, Im using ms2 3.11

http://www.msextra.com/doc/ms2extra/...ion.htm#3coils

Jobro

iTrader: (2)

That could work too. However you are relying on the Spark Output Inverted to work which I did not test. You are also relying on the microcontroller driving high and then driving low, rather than driving high and then releasing the pin (make it high-z input).

You are also relying on the microcontroller pin to drive the igniter. They are only rated at 1-3mA typically in this family of chips. You should strive for 20-40mA drive like I am.

Jobro

iTrader: (2)

Ken,

I've come up short on outputs. Here is my resource usage (current and future).

Fidle = PWM idle (future, but used in past hardwhere built)
LedA spark
LedB spark
LedC spark
JS10 = CAS 2nd trigger
JS11 = spark d

iac1 fan relay (switched earth)
iac2 boost control or (water injection on/off)

are there any outputs left?

I have realised that if I use full sequential ignition i will need a dash tacho driver because all negative coil wires will be at half speed.

muythaibxr

I'd have to go look at the charts myself as I don't have it all memorized.

The chart I'd check is here:

http://www.msextra.com/doc/ms2extra/...htm#ms2options

Jobro

iTrader: (2)

Hey thanks for the pointer!

So if JS4 and JS5 (spark output pins used for SparkE and SparkF) are not selectable in the software, its just a little bit of firmware to make the tacho out go on them?

Jobro

iTrader: (2)

In the end I made IAC1A,B IAC2A,B pins on the back panel available for ignition wires. I replaced my 4.7k pullup resistors on the LED's with 220 Ohm resistors. I soldered another 2n3904 npn transistor into the proto area, 1k from the processor pin to the transistor base. I put a 220 Ohm pullup to +5V. No LED built onto this extra ignition driver.

I'm also thinking about changing the 330 ohm led resistors to be larger, like 1k ohm to save a bit of current incase the 2n3904 transistors are now having a hard life fighting 220 ohm pullups.

The car is now setup like this, and I have driven it with 220 ohm pullups and still using the standard FC ignition coils. I'm sure I was told this would never work and that the resistor needed to be 4.7k.

Anyway when I get some time I will get a 4channel scope trace of the RX-8 mode. Maybe tomorrow if I get time!

muythaibxr

If there is any such thing as "just a little bit" when it comes to firmware.

Ken

muythaibxr

The problem I had with them at 1k vs 4.7k was that the TPS and several other signals would go wonky and trigger random accel, etc...

It's certainly possible that problem was specific to that car, but using 4.7k resistors there fixed that problem.

Ken

Jobro

iTrader: (2)

Oh sure its possible because they share the +5v bus. I haven't had the laptop on it to check for new noise.

Anyway I'm daily driving with 220 Ohms there so it seems ok. I will scope it tomorrow!

I'm going to post a new thread about injector deadtime and deadtime compensation and it is definitely something you might know the answer to.

wankelbug

hey guys im adding ls2 coild to my 13b powered bug. i am currently using the stock fc setup. but my coil died so i decided to upgrade. i need someone with experience to answer a couple questions.

here is what i got so far.

all coils wired per the diagram on mega manual. including the 2 capacitors.

now on the computer side the way i understand is is
front rotor leading signal output A d14 with 4.7 pullup
front and rear trailing (firing as a wasted spark) are outputs b and c to 2 coils signal wires
rear rotor leading output D (js11)

then i run it in fd mode with a dwell of 5.5 and set to "going low" correct?

im also not running the stock rx7 tach so i can now just grab a signal from an injector wire correct? or should i go into the megasquirt and wire my tach in there?

thanks
Andrew