barf

I've done a few MS2s and there will be other success stories lurking around. As kpr said you'll need a 2nd VR signal conditioning circuit if using MS2, with the NipponDenso distributor VR sensors (Ne, G1, G2 wires). If you're up to it, you can DIY or splash out and buy a 2nd VR kitset (easiest).

NipponDenso input signal documentation:

http://msextra.com/doc/ms2extra/MS2-Ext ... utors.html

the 2nd VR conditioning circuit looks like this

so, unless you can build that or the LM1815 based circuit, get a 2nd VR kitset, or the DIYPNP as that has 2nd VR signal conditioning circuit built in.

actually, the DIYPNP might be the best option. it has the same 42 pin connector as the OEM ECU.

http://www.diyautotune.com/catalog/diyp ... p-383.html

A 19 amp IGBT ignitor is builtin

holy batman!!!

to use these, you would first supply them power and grounds, then take the logic output signals, from top of R26, R27 and R29, etc, to the logic input pin A (assuming MS2/Extra).

the documentation you need is here:

http://msextra.com/doc/ms2extra/MS2-Ext ... m#copsetup

However, as soon as the throttle is opened even slightly, the vacuum shoots close to zero in a non-linear manner and doesn't vary much throughout the rest of the throttle opening. Earlier on in this thread a few other people backed this up, so I don't think you can run a MAP signal to fuel/spark it, you'll need TPS as well. That said, MAP would probably help with enrichment and other drivability things, so its always worth hooking up a tube to the manifold and the sensor, just so you have it available for tuning.

it is certainly possible to tune ITBs based off of MAP and TPS. a hybrid combination of alpha-n and speed density is required. for idle and fast idle speed-density works but as the relationship between MAP and VE becomes subtle above a MAP 90% of barometric pressure, alpha-n is used in these areas. megasquirt (MS2/Extra in particular) permits you to use any combination of alpha-n and speed-density. one table is your primary algorithm and a second table is a multiplier using another algorithm.

EDIT: here is a great resource on 'blended' tuning for ITBs

http://www.77e21.info/mstuning_blendedansd.htm

thanks. i get 11.4v during cranking. most fuel injectors are +12v ground switched, that helps explain why cranking pulse widths are higher too. the megasquirt has a battery voltage correction factor for inj' duty cycle.

Why oh why? All you'll get is shittier starting.

Why? Could you please explain that?

Now you've got 12V connected to a coil thats meant to run at 12V, producing a totally normal spark... Just like you did when you had 8ish volts running to your old coil which was meant to run at 8sh volts... But you've lost the ability to temporarily make the spark more powerful when your cranking...

How is the spark made more powerful durink cranking?

if while you're at it, you want a cheap spark upgrade, connect your spark plug receptacles to the battery negative. i think HKS sell some kit to do that. don't connect to painted stuff tho.

ground is just as important as power supply.

do the nissan MAF sensors contain an air temperature sensor?

if so, any ideas what wire into the sensor is the temperature? (i'm being lazy i could just note resistances and temperatures.)

ground your temp sensor better

i agree the code needs to be open, i'm sticking with MS2/Extra until they GPL or open it in some way. there's a growing clone market out there that they seem to be at odds with, when they really should just embrace it.

that whole 6x6 table needs to be copied from Flash or RAM to the accumulator (=clock cycles) before each group of math functions is done on it, table lookups are expensive in low-memory systems like a microcontroller.

MS2 is a freescale MC9S12 - 16-bit @ ~25 MHz

http://en.wikipedia.org/wiki/Freescale_ ... erivatives

MS3 is a newer MC9S12 (MC9S12XEP100MAL) still 16-bit but with improvements. i suspect it's mostly a code re-write(?).

http://www.freescale.com/files/microcon ... 00RMV1.pdf

@Spencer - you say "only" 6x6 as if thats a bad thing, the value output is taken from an interpolation based on this table lookup, it is smooth, there is no sudden change between values - assuming the table is 'spaced-out' correctly across the desired range of operation. bigger is always better tho

even a 16x16 VE or spark table could be 'too small' if the values on the axes were not placed in the area where the greatest transience is. hope that makes sense lol.

Befor i made this post i did a bit of searching on the net aswell as used the search function on here.

I would like to know;

Where they come from.

Where they are made.

Do they come pre assembled

Are they kitset.

Are there any NZ suppliers

What is the back up sales and service in NZ for them

Who in Auckland can tune them

Has anyone purchased one and fitted it

If there fitted by an average DIY type of guy how well will they run

Thank you.

I've ordered mine as kitsets from diyautotune.com shipping was prompt (2 weeks) and they had good after-sales support and a facebook group, although I didn't need much help with all the great documentation and forums at msefi.com.

you can buy preassembled ones at a slightly higher cost. the CPU module is preassembled always however, owing to being largely TSOPs and other SMT components. use MS2 at least, MS3 looks awesome. if you need help with them, ask. it's not a turn-key system like MoTeC or Link but the features certainly compare. maps and correction tables are in 7 or 8 dimensions if you do it right.

I've installed a few, just for mates in Christchurch and we tuned them ourselves using Innovate LC-1 wideband controller with Bosch LSU 4.2 exhaust gas-oxygen sensors (although they actually detect the amount of nitrogen). we didn't have a dyno and tuning is a completely different matter to installation.

does it still go hard OK?

if so idle control valve maybe? they can become fouled with oil from PCV, soaking in diesel worked for me

the 20v itb's vacuum signal isn't shit - it's application is different (they combine AFM and MAP - if am I correct?), actually it has a very precise vacuum sensor position but it is denied a MAP reading when the throttle is closed as noted, there is a small vacuum-sensor hole occluded by the butterflies when shut, if you look closely. will document how we got around this in a new project thread, hope to post that up soon as poss. basically you need idle control system plumbed into your itbs - got photos you can put up rob?

In the case of Megasquirt 2 - TPS without MAP works OK using a speed-density algorithm but you'll likely miss out on features like acceleration enrichment, X-Tau (wall wetting compensation) and have no chance of electronic spark advance derived from manifold pressure. there are two main fuelling algorithms, speed-density (for throttle-body injection ie; V8 holley TBI manifolds) and alpha-N which is the multi-factor algo you want to get it nice and smooth.

out of curiosity, how do you ppl achieve proper spark advance control when the vacuum signal has been neglected? does the Link have a simple RPM + TPS table?

good luck with your project.

you can go sequential injection and 8-cyl COP with MS3 for well under $1000 - if you can solder. bank-fire fuelling algorithm in MS2 accounts for wall-wetting so it's a fine difference (don't bother with MS1). one project coming up is a 4k with twin-downdraught throttle body injection robbed from a 1S-iLU

whats best on valves?

thanks

head is off but i don't have a valve spring compressor, just been turning the cam to open valves and clean it up :-\

I've seen kitsets by MSD to replace points, not cheap, I suggest address the ignition system as a whole. you're not just replacing a part here, you're moving from kettering points to electronic ignition systems.

I like the megasquirt II (ignoring the fuel injection features) it works great in a 'spark-only' application and they're cheap and reliable. they work with almost any crank or cam-speed timing signal, the challenge here is getting a timing signal, use a sensor from a dizzy, install laser-cut disc (+pickup) onto front crank pulley,etc.

i lost compression in a cylinder because of loose coke interfering with the valve seat, whats a good/safe solvent to use on, or way to clean an alloy head to remove coking? basic question i know but, first rebuild here, she was burning a lot of oil before the rebuild.

already had the acid dip and leak down test was OK but there was still lots of gunk on top of the valve when i got it back from the head shop. turning the engine over was enough to loosen it off when the valves moved and i don't want to burn that shit in clean cylinders!

my vote is for the 1LR-GUE by Toyota/Yamaha

or the 3K-R

in chronological order

Massey Fergusson 168 (with PTO and two hydraulic circuits)

Ford Escort Panelwagon

Toyota Celica GT (RA63)

Toyota Corolla DX (KE70)

Datsun Bluebird SSS (Z18) (schedule A)

Toyota Starlet S (KP60, two KP61s) (schedule A)

'77 Hillman Hunter (caught fire outside our house, plastic fuel hose. swapped him for the fire extinguisher)

BMW 323i (E30 5-speed)

Mitsubishi Lancer GSR (CM5A)

Toyota Celica GT-T (TA63)

and will one day restore the 1936 Jaguar SS100 sport coupe sitting in the garage

so I'm frustrated and want some advice.

trying to disassemble the oil pump on the 4G93 to de-coke it, the bolts are supposed to be torqued to 10NM but after 270,000kms they seem pretty stuck. they are a plus-head countersunk bolt and i'm not keen on mincing the bolt head if i can avoid it, theres no nut they just fasten into the oil pump. whos got the best pro-tips?

do i reaaaaalllly have to drill and eziout these tings? would heating/cooling help?

i think you might not observe anything on that wire from the ignitor because it doesn't output a voltage but switches to ground instead, there was a 5v pull-up in the ECU on that diagram for this purpose.

well you might be lucky here, if the signal from the toyota ignitor was the same or similar enough (can't recall) then it could just wire straight up OK.

the theory is the same in either case, a tacho identifies a revolution by observing a momentary break in the ignition circuit. logic level tachos may invert this signal and/or represent it by frequency of a square wave

forgive the pbrush.exe diagram

looks like that ignitor outputs logic levels and my dash (CM5A) measures RPM as logic-level pulse from ignitor/ECU - not from coil. so that tacho output from the ignitor probably won't drive your tacho, that circuit with the zener is the ticket.

what sorta tacho are you hooking up? vehicle's original tacho? (what car?)