Porky EFI

Rich H · Post by **Rich H** » Tue Oct 09, 2007 6:38 pm

The K Jet is dead (Well in the shed) long live the Megasquirt! (Eventually)

Before:

The overall view:

Fuel Rail view:

Induction side:

One sender to screw in, and lots and lots of wiring to do....

Rich H · Post by **Rich H** » Tue Oct 09, 2007 6:42 pm

things to do:

Calibrate coolant and air temp senders

Fit senders (Just screw into the prepared holes)

Make some brackets for the induction gubbins

Fit FPR somewhere

Plumb in fuel lines

Wire every thing up

Plug in Megasquirt
*
Start car

Revel in improved performance and better fuel economy... Or not

Rich

*Possibly missing a few minor details like getting a basic fuel map and spark table but that's easy

BiggestNizzy · Post by **BiggestNizzy** » Tue Oct 09, 2007 7:41 pm

looking good I'm very impressed even though I know nothing about what your on about

Rich H · Post by **Rich H** » Wed Oct 10, 2007 9:11 am

Electric fuel injection, the existing mechanical stuff is a bit tired and I think the injectors are shot (Not sealing when the pressue is removed, so fuel everywhere)

Probably won't make anydifference to the performance (Thats what I'm telling the insurance co) but will probably do more MPG.

Anyone got experience of getting a base map sorted so it will run? I know how to tune all the settigns but I need a base map to work from!

Brunty · Post by **Brunty** » Wed Oct 10, 2007 10:26 am

Nice one, something I've had a few daydreams about. Which kit did you get?

Rich H · Post by **Rich H** » Wed Oct 10, 2007 12:24 pm

MS1 with PCB V3, the DIY instructions are excellent, the only complicated bit was that I wanted Spark control which requires a few mods and I hadn't noticed... Works now though!

Stim was a DIY mess by your truely form spare parts (I often have resistors, capacitors and strip board lying about...) but it worked!

I'm going to use a locked Dizzy (Ready to fit) and the VR pickup to give crank pos and Direct coil drive.

I have also looked into the K series MS but the Crank Pos is complicated as Rover used an odd trigger wheel. If you could either fit a 36/1 wheel or weld teeth into the ring (!) to make a sensible trigger it would be alot easier.

Need to work a base map now...

gdr · Post by **gdr** » Wed Oct 10, 2007 6:52 pm

I fitted MS on my Ultima and got a small block chev base map off internet (MS forum)- there are loads for US V8s of course - just chose one for an engine which was pretty close to my state of tune and it started and ran fine 1st time. SBC happy as long as it gets too much fuel, I tweaked it back gradually. Try asking on MS forum for map (assume the forum still active). Otherwise maybe just pick a similar cc and power 4 -cyl engine and go with that, advise tune rich to start with to be on safe side. Not so sure about spark, I only got as far as fuel.
Recommend you fit a wide band gauge if you don't already have one, makes tuning much easier.
Good luck!!

ps, I have an EDIS-8 and crank pick up in shed free to a good home if anyone doing this DIY EFI wasted spark ignition thing on a V8.

Rich H · Post by **Rich H** » Wed Oct 10, 2007 8:52 pm

Nice one! I'll try that. It's not exactly a refined engine anyway so a yank tank 4 pot will do! Spark map I'm going form the spark advance table in the workshop manual, once it's running I'll hire a WB lambda for a week or 2 and tune the lumps out!

robin · Post by **robin** » Wed Oct 10, 2007 10:58 pm

Are you programming spark advance vs MAP as well as RPM?

You need loads of advance for part filled cylinder (i.e. closed or partial throttle) - when there is less air it burns slower, so you need advance to get the flame front to reach the roof just at the right moment.

You also need some advance for RPM (flame front progresses at same rate, piston is moving faster, so need earlier ignition to make sure you get bang at the right time).

Assuming you haven't done anything much about the air flow then book spark advance should be OK, but you may struggle to convert it from mechanical spec to electronic spec, IYSWIM.

The bigger struggle is to get the V.E. table right. A wideband lambda should allow you to fix fuelling - once you've fixed fuelling you can probably reverse engineer the VE table and hence look at where the engine breathes better than expected (fuel increase required to level out lambda => retard ignition) or worse than expected (fuel decrease required to level out lambda => advance ignition, maybe). Remember, retarding ignition is quite safe; advancing it is less so!!

Is your injection batched or sequential? Have you got the right spec FETs and flyback circuits to handle your injectors - that seems to be the thing that plagues MS users ....

Cheers,
Robin

Rich H · Post by **Rich H** » Thu Oct 11, 2007 8:03 am

Cheers Robin, I kew you'd be along to help!

Sparky table is against MAP and RPM, both at user definable points. IIRC 12 MAP sites and 12 RPM points. Same for VE table.

Air flow will be improved as the original K-jet air flow meter is a plate type, it effectively blocks the airflow at idle near enough (No bypass) so it will flow better, how much better I have no idea.

Injection is batched, MS only has 2 injector drivers and all the injectors are high impedance so *should* work striaght out the box....

I'm using the onboard coil driver too.

I can hire a wideband setup from the guy who retails MS in the UK.

Cheers
Rich

Rich H · Post by **Rich H** » Thu Oct 11, 2007 8:06 am

Mech to electronic seems to be straight forward, the advance graph is very simple. Mech advacnce is "all in" by about 2k rpm. The lump is reportedly conservatievly timed anyway the max total advance is only 32-34 but the yanks routinely run it at 38 odd on crap yank tank juice.

robin · Post by **robin** » Thu Oct 11, 2007 1:21 pm

I think RPM vs advance on full throttle is easy enough to setup; it's MAP vs advance that will be harder to get right as there will have been many "gain" stages between the manifold and the dizzy (i.e. how tough was the dizzy return spring, how well did the tube allow the vacuum to propogate and thus act on the the pneumatic advance mech, etc.). Hence the stock setup is conservative - you probably find vacuum advance can vary by +/- 5 degrees from car to car. The good news is that once your at >2-3K you can ignore vacuum advance pretty much as you need the speed advance anyway, and advance is capped ...

Rich H · Post by **Rich H** » Thu Oct 11, 2007 2:16 pm

Sanjøy · Post by **Sanjøy** » Fri Oct 19, 2007 12:41 am

Chap who helped me with the clam has just bought a MS for his Zetec for the Elan+2. Will pass on the thread to him as he has no instructions just the MS and the looms!

gdr · Post by **gdr** » Fri Oct 19, 2007 7:50 am

http://www.megasquirt.info and http://www.msefi.com
All you ever need to know is there.