M20B20 to MegaSquirt build

[Reck]

Class work!

[BertjeConti]

Spark setup

As mentioned earlier a'm going for a wasted spark setup.
Why? because for this setup you do not need a camshaft trigger, The ECU does not need to know which of the two in pair moving cyclinders needs to be fired. (in a 4 stroke engine each cylinder needs to be fired once each 2 crank revolutions) In Wasted spark mode simly both paired cylinders are fired at the same time each crank revolution.
This system is succesfully used on four-stroke 4 cylinder motorcycle engines for decades.

Today my coilpack arived, which i bought at ebay for just 52 euri including shipment from a germain carparts company. It is a new coilpack mentioned for a vauxhall V6 engine.
A coilpack is in fact 3 coils packed together, and each coil fires two spark plugs.
It has 6 high tension lead terminals and one low tension connector with 4 terminals.

This is the coilpack:


At each high tension terminal a number is stamped in the housing, these are the cylinder numbers for a vauxhall V6, forget these numbers because a V6 has a total different firing order compared to a straight 6.

The firing order of the BMW straight 6 engine is 1-5-3-6-2-4. 1 & 6 are paired, 2 & 5 are paired and 3 & 4 are paired cylinders.

The coilpack has a 4 terminal low tension connector, numbered 1 through 4
1 = coil nearest to the connector
2 = middle coil
3 = coil far from connector
4 = +12V supply

The coil does not have a terminal to connect to the chassis.

I've planned to mount the coilpack beneath the coolant reservoir, for that 4 holes drilled through the right inner wing.





I did not have a suitable connector, so for the time being it is wired up with small spade AMP crimp connectors. I need to order a 4 terminal timer connector for a proper connection between wiring loom and coilpack. (todo list updated)

Wiring up the coilpack is realy simple, start with cylinder 1 to a HT terminal nearest to the connector, top or bottom doen not matter , and then go around the coilpack with each cylinder, to end with cylinder 6 at the other HT terminal nearest to the connector.
At the low tension connector, connect terminal 1 to spark-A at the ecu, terminal 2 to spark-B, terminal 3 to spark-C and terminal 4 to 12V supply
Like this sceme:


For the first test run with the new ignition setup I disconnected only HT lead cylinder 1 and 6 from the old dizzy and connected these HT leads to the coilpack.
The engine would run for shure, and if the new ignition setup was correct it would run on all 6 cylinders.
And YES IT DID run on all 6 cylinders when i fired it up. (4 on the old dizzy and 2 on the Megasquirt )

After this test i wired up all 6 cylinders to the coilpack and disconnected the electronic module from the old dizzy setup.


Engine started right away and ran smoooooth. Made immediately a 10k testdrive with it. no problems whasoever.

Ofcourse i had to make a spark map for the Megasquirt. The spark map i've designed is partialy based on the distributer parameters, and based on lots of information i've gathered from the internet.

This is my initial designed spark map. The engine runs great on it:


I think this spark table can be tuned a little, thats for the coming days.

From now on the car runs on my own build Megasquirt based ECU, only ignition.

stay tuned

[Wondermike]

Respect! It's nice to read that someone decides to do it "his way"
And gets the job done!

[BertjeConti]

ECU bracket

today i made a bracket that holds the ECU firmly in place in the top of the glove compartment box.

The bracked is designed in a way that it can release the ECU box by unscrewing two scews at the front.

Here the bracket is already bolted in place, the glove compartment is removed:



and the ECU installed and wired up:



And the glove compartment back in place:




Thats all for today.

Thanks for all the positiv reactions, i appriciate it!

[BertjeConti]

In the upcoming weeks there won't be lots to tell because of some wekend and weeks camping holidays.

Tomorrow i'll pick up my caravan out of the winter storage. It needs some attention too.



it's a 1976 eriba titan which i restored almost 10 years ago. Now she gets a mover, because of my health reasons (back) and kids who stay home.



I hope i'll find some time for the megasquirt project too.

[BertjeConti]

Back to the coilpack

I've searched the internet to find suitable dwell settings for this coil. Dwell settings in the ECU where still default. I had no idea if these would match with this type of coil.

Internet gave poor information. max dwell was 1.6mS or 3.5mS and cranking dwell 2.8mS or 4mS, thats all i've found.
The ECU was configured with 6mS dwell for cranking and nominal dwell=3.1mS

So, the only option was measuring the coil charge current with a scope.
To do that you need a shunt resistor (approx 0.01ohm) in line with the coil and measure the voltage over that shunt resistor with a scope. At the moment the charge curve flats out the coil is going into saturation, and dwell time is to high.

I used my multimeter (current measure terminals) as shunt resistor and put it in line with the +12V terminal and the coilpack. Hooked the scope to the multimeter terminals and got the coil charging current signal on the scope.



The signal on the scope is the current through all three coils together, thats why the charge currents are so close together.

I had to increase the dwell to 8.5mS to see the coils go clear into saturation. That is a huge dwell time

Decreased the dwell to 7mS, with this dwell the coils are stil going a little in saturation:


On the other hand i tried to decrease the dwell untill the engine would misfire. I went back to 1mS but the engine still did not misfire at all. (good coil, spark plugs and leads i think )

Knowing these facts i set the nominal dwell back to 3.5mS and left the spark duration at 0.9mS, which would be for sure within the safety area for this coil and drivers.

here my settings:


The measurements gave me confidence about the ignition settings.

Each coil charges once each rev of the engine. At 6000 rpm one rev is 10mS, so when the charge takes 3.5mS and spark (discharge) 0.9mS there is time enough left over.
With a distributor the coil needs to be charged 3 times each rev, so only a 3.3mS timewindow is left to charge an discharge the coil at 6000rpm. with mechanical contact points there is a lot of safety time for discharging and a resistor in line with the coil to protect it for overcharging at low revs, this means there won't be a lot of charge current left over at high revs.

That is a strong advantage for a wasted spark setup.

[BertjeConti]

Throttle Position Sensor

A while ago i ordered this TPS intended for a Honda from Ebay. It has the same shaft shape as the BMW throttle body.

http://www.ebay.com/itm/390616558044?it ... 20&vxp=mtr

It was only $5,69 and free shipping, so i could not resist to buy it.

This TPS works great with the Megasquirt ECU, no modifications in the circuit needed.

I had to remodel the connector on the TPS, now the stock 3 terminal BMW Timer connector fits perfectly

This TPS has a perfect fit on the throttle body shaft, but that was all, the shaft is 10mm to long and the TPS does not align with the mounting holes by far.

So i had to fabricate a adaptor bracket. i used FR-2 (pertinax) for this bracket. The bracket is 10mm thick, so the shaft could remain its length.













there is still a lot to do before the carb will be swapped for injection:

- fabricate a fuelpump and fuelfilter bracket
- installing a 02 sensor
- testing and cleaning the injectors
- fabricate something to mount a E30 air filter box
- connection between the air filter box an throttle body
- things i can't come up at the moment

Stay tuned

[Jeroen]

man man man

speechless with amazement

[Wondermike]

Like this thread a lot!

By the way, your Eriba Airstream looks fantastic too

[BertjeConti]

Fuelpump and fuelfilter bracket

My car ofcourse has no high pressure fuelpump needed for a fuel injection system. The mechanical fuelpump attached to the crankcase and driven by the same schaft that drives the oilpump and (now obsolete) distributor cannot deliver enough pressure for a fuel injecton system.
For this system a fuelpump is needed that can deliver fuel at a pressure of 3 Bar or approx 45psi.

Ive planned a fuelpump and filter outside the fueltank, as usual in the old days.
Ive ordered a new fuelpump and filter from ebay (35 euro and 15 euro), and designed a bracket for these components. the bracket bolts under the car near the fueltank at the right side where the fuellines are running.

Luckyly my car has already a fuel return line which is now also in use for the low pressure fuel system to feed the carburettor. The low pressure system has a pressure regulator and 'overflow' returnline, which is not at all common for carburetted cars.

My bracket holds the fuelpump and fuelfilter with 2 hoseclamps each. Fuelpump and fuelfiters don't have any mounting flanges/holes/systems whatsoever so you'll have to be creative or use a simple clamp.

here is my bracket, made out of a piece square tubing and two pieces sawn out of a thick 50mm scaffolding pipe and 2 M8 nuts:







Now painted:



and pump and filter attached:





another job done

[BertjeConti]

Testing the injectors before cleaning

I want to test the injectors if they work properly, they come from an engine i don't know the history of. So it is important to know if the injectors are any good.

The first thing i tested was the pressure regulator, in my case the regulator from the M20B20 motronic engine was rated at 2,5 Bar.
The pressure regulator is connected to the fuelrail at the front of the engine (highest point of fuelrail) and connected to the fuel return line and connected with a thin air hose to the manifold, when the pressure in the manifold drops (vacuum), the regulated fuel pressure drops equaly.
Testing is simply done with a bicycle pump with a gauge.
The fuelrail with injectors and pressure regulator is fully assembled, connect a bicycle tire valve (cut from a old inner tube) with a piece of fuel hose to the inlet on the fuelrail.
Now try to pump up the fuelrail with air like you are pumping a bicycle tyre. The pressure gauge on the pump must halt at 2,5 Bar, the pressure regulator releases the air that you pump in to much.
My pressure regulator seemed OK.

Since i have now a fuelpump setup, working ECU and stimulator it is possible to build a injector test setup. My goal is to check if the injectors work, give a nice spray and most important if they all deliver the same amount of fuel.
Thats why i want to test them with the complete fuelrail setup assembled.

As everybody knows, a picture tells more as 1000 words, this is my bench test setup:



As the picture tells, i've used beerbottles to catch the fuel the injectors will spray. The beerbottles are secured to the tabletop with straps. Connected the fuelpump assembly to the fuelrail. Connected the injectors to a battery and the ECU, connected the fuelpump with a switch to the battery, and connected the stimulator and laptop to the ECU.

It is a save idea to do the test outside, since we work with real flamable fuel.

This is a picture of the complete setup, fueltank is a little 5 liter canister filled with gasoline.



I've made a video of the benchtest working, injectors make lots of noise, i did not expect that:



I let the ECU run at different speeds with different PW values until the beerbottles where half filled.

After the testrun each bottle is filled near the same level (hard to see due the brown glass):



After the testrun i weigh each bottle before and after emptied it.

A little table with the test results:

injector......weight bruto....weight bottle.....weight netto (squirted fuel in test run)
..1.................350..................239..................111
..2.................348..................240..................108
..3.................367..................257..................110
..4.................347..................241..................106
..5.................348..................237..................111
..6.................351..................239..................112

No so bad result, only injector 4 is a little behind the rest,
While spraying injector 4 gave another look. the squirts from all injectors hit the fuel in the bottle realy hard, as the video shows, only injector 4 did not do that.

After cleaning the injectors (they get a ultrasonic cleaning bath, and a flush with carburettor cleaner) i'll do this testrun again.

Hope to get transparrent beerbottles , so we'll be able to see more squirting action in the video.

Weighing the beerbottles:
No 1 filled:



and No 1 emptied:



stay tuned

[Jeroen]

Now that's a decent test, and the test result on injector nr 4 had a reason, curious what the results will be after cleaning!

[323baur]

Verg impressie!!! So is your airstream too

[steed]

this level of detail during bench testing pre-fitment is phenominal! most impressive work

[Thegs]

Great work Bert! Please set the privacy settings of the movie to public so we can see the test aswell