Hi guys

Thought I would share my 440 EFI conversion experience with you. I have a low compression 440 fitted to a '74 Jensen Interceptor Mk3.

I got the car fitted with a Carter Thermoquad 4-barrel carb and dual point distributor which a PO had fitted. Its supposed to have an electronic dissy with a star wheel and magnetic pickup.

Since I've done quite a few EFI conversions using Megasquirt-2 ECU, I decided this engine badly needed optimised timing and fuelling and started the EFI conversion process by implementing "Phase 1" which is converting the ignition system to an ECU controlled wasted spark distributorless ignition system. Phase 2 will be the fuel side of this but for now it is still running with the TQ.

Phase 1 is now complete except that the timing map really needs to be tuned on a braked dyno to find the least amount of timing advance needed at every operating point.

Here's how I did it:

1. Converted the crank damper into a 36-1 trigger wheel by having 36 slots cut into it and then having one tooth removed. I had this done such that one tooth still retained the TDC mark so I could verify the the actual spark timing matched what the ECU was commanding.
2. Mounted a VR sensor (Magnetic pickup) using two of the timing cover bolts to attach a sensor bracket with.
3. Mounted 4-post coil packs (EDIS) in the V behind the inlet manifold;
4. Installed the ECU and harness with some relays into the car.

The most interesting discovery so far is that the engine seems to idle best @ 30º BTDC!!! Changing the timing (at any operating point) is now simply a matter of punching in numbers on the PC and the change is implemented immediately and on the fly, making it very easy to hear how the engine responds to different timing values. I started the engine with 0º (and it was running but not happy) and then started advancing timing by adding 5º at a time. As I went past 15º my surprise became greater and greater as I eventually settled on 30º advance!

This led to me wondering if perhaps the TDC mark on the crank damper was wrong so this morning I checked it using the piston stop method and its spot on!

Phase 2 was to do the fuelling side of things. I used an Edelbrock 7193 intake which was modified to accept injectors (Edelbrock 3598). An Edelbrock 38783 throttle body is used to regulate air intake. Please note that the 7193 needs to me machined otherwise idle air via the PCV port and the idle air controller port in the throttle body will be mostly blocked by the intake.

To avoid fuel starvation, I fitted a surge tank from a Jaguar XJS with a submerged EFI fuel pump inside the surge tank. A low pressure pump keeps the surge tank supplied with fuel.

I have a very comprehensive thread on this job here: http://www.joc.org.uk/phpBB2/viewtopic.php?f=20&t=15116

I also have a downloadable PDF document here which summarizes this conversion: https://dl.dropboxusercontent.com/u/...Dec%202013.pdf

Here is the completed job:

 

Question.... Did you try it with the Dist. ???
If so did it run with the somewhat correct timing? like 8* to 18*to start initial timing to the max advance of like 34* total timing @2800 rpm's?
What I am thinking is that if the initial setting is not responding properly you could be fighting a improperly installed camshaft? Just a hunch.
But for it to like running with 30* initial to start. sound like its going to have a hard time cranking when starting when it get hot.
Glad you came out ok with the Damper and found that it had not spun.

 

Bob, I got this car with a dual point dissy even though its supposed to have an electronic dissy. Although I could get it to run with the dissy, it never ran well with the dissy points controlling coil charging. I thus fitted an MSD6A I had lying around very early on. This has the downside of not being able to check timing with a typical inductive timing light as the MSD fires 3 sparks in quick succession for every combution stroke at idle and up to about 1500rpm. In that sense the MSD works REALLY well and I was able to get the car to run well but I could not check to see what the timing actually was because of the multiple spark discharge. (I'm still keen to know how one would go about checking timing with a MSD....)

Whilst I do not know with certainty what cam it has nor if its been installed correctly, my feel is that the cam is a standard one since the manifold pressure at idle is 30KPa (Sorry, don't know what this is in "Hg) which is quite normal for a road cam. Also, it seems to run out of steam around 4500rpm - but this is more likely due to fuel starvation as I am quite sure that the facet pump I installed can not supply adequate fuel for sustained periods. This will become quite apparent when I've got the EFI running as well.

Thank for you help Bob! It is much appreciated.

 

From what I've read, the MSD works OK with older, "non dial up" (non delay) lights.

Something I STILL cannot get my head around is the terminology MS uses.

What is the minimum board and processor version I would need for whatever term you would call their "self learn" software?

I'm currently using a Holley commander 950 ECU, and thinking of converting to MS

Seems to me MS has done a poor job of explaining the features of the various boards? versions? processors?

This for me would be towards the "small end," IE I don't need surface mount, pre-assembled boards.

 

http://www.msdignition.com/Products/...ing_Light.aspx

For best results...

 

Philip... I would be very interested in what you find as the fix to this !!!!
Keep us posted on what you find out.

 

440RR, if you give me some specifics of what you would like to build, I could guide you.

In the mean time:

MS101:

1. "Megasquirt" was started by two Americans . It is now very much an international effort.
2. MS started as an EFI educational tool and this objective is still very much the objective. IF therefore your objective is NOT to learn but simply to get an engine running, MS is NOT for you - in fact, you might end up being frustrated. IF however, your intention is to really learn about EFI and engine management, you have the ability to do self-study, be patient, and ask others (forum), MS is THE ONLY ECU that will really facilitate that learning process.
3. Where MS beats the pants off ANY OTHER ECU is when it comes to support. The MS documentation is better than any commercial product out there! The MS support forum is better than any other commercial product's forum out there (personal opinion of course.) The MS manuals are often criticized for being TOO COMPREHENSIVE but how can you learn when information is missing?
4. "Megasquirt" consists mainly of 4 "elements":
   1. The Motherboard (V3 = through hole PCB mounting, pre-assembled or DIY, or V3.57 = surface mount)
   2. The Processor (MS-1, MS-2 or MS-3) that plugs into the above motherboard;
   3. The firmware that the processor uses (there are two sources of firmware to choose from: "Bowling and Grippo" or "Extra"). "Firmware" is software the processor runs on which resides inside the processor.
   4. The PC software you use to tune the ECU with (Megatune = old and unfriendly or Tunerstudio payable but cheap and MOST user friendly)
5. The combination of the above gives you a "Megasquirt ECU".
6. I would not recommend MS1 given that MS2 is so much better value for money;
7. I would not recommend MS3 for an MS beginner - rather start a simpler project with MS2 (or MS1), and then tackle an MS-3 project. MS-3 is the one to get for sequential fuel injection and Coil per plug or coil over plug implementations;
8. MS is NOT self-tuning!! However, if you tell MS what AFRs you want (assuming you know what AFRs would be good...) AND you have installed a wide band oxygen sensor (correctly), it can self-tune the fuel map to that AFR map you have specified. This it can do in less than 10 minutes!
9. MS is NOT self tuning wrt ignition timing. YOU have to tune the timing map and this is best done on a braked dyno.
10. ALL other parameters have to be tuned by the tuner - and there are many. This is why you really need to know what you are doing.

If it looks like we are going into to deep MS discussion, I would prefer we start a dedicated thread on that subject.

 

Sorry I hijacked your thread but that info helped a lot. I've read some of the literature, but still was fuzzy between the board versions and the processor versions.

I'm not exactly new to this, as I've installed and programmed the Holley Commander 950 system and currently run that

. I also have no problems with electronics assembly.

I do understand that the MS is only "self learning" as far as some of the software is concerned. There are actually two third party softwares for the Holley that does some of that, one written by a guy named "Tazman" which was a big help on my system.

I have no interest in either sequencial nor coil on plug. But if I were to go with a turbo, then a system which could handle timing control would be useful.

 

1) No worries mate. If you have questions, let's start a new thread or PM me.
2) Your holley experience will come in very handy with an MS install but be warned - you WILL be learning A LOT more than what you think :-)
3) Excellent! Then you can build your own exactly the way you want it. This is one of the good aspects of MS - how you can customize it to your requirements.
4) Well, both Megatune and Tunerstudio can do the tuning for you but the B&G firmware can tune the fuel map (to target the required AFR map) without the need for having a PC connected. This is rather risky though, as you don't know what is happening. You can set limits within which the tuning is allowed to operate which can protect you if set correctly.
5) I also think sequential is overrated (but I'm told it can do a lot for emissions if that is an issue). I recon its biggest value lies in engine smoothness. There is no power or economy benefits worth mentioning. COP or CPP also holds no benefit over wasted spark other than physical implementation considerations.

 

Update:

And now the engine is running with EFI!

 

Very cool. "Sure looks funny" LOL with the plug wires coming out of the rear area of the engine.

 

Thanks, still hoping to route the HT wires down the back of the engine and under the exhaust headers to be "completely" out of sight.

Will try this when the engine is out when I tackle the ventilation system.

 

Looking Good....

 

Nice work ! FYI I had a MSD6a on a 360, standard non dial timing light. Never had a problem

 

years ago i learned how to get a duel point degrees on dwell EXACT on 42 degrees every time. i could do it in 15 min.

 

Why did it take so long ? I have an old Sears dwell/tack somewhere in the garage

 

Thanks for posting this up. Information is very helpful.
Going to be Megasquirt EFI myself soon.


Mike,

 

In the mean time I have found the idle valve that came with this Edelbrock Throttle body is a useless piece of junk. Its basically an open or closed valve. There is no way of having it anywhere in between. So I fittted a proper Bosch 0280140526 idle valve that came off a BMW E36 series.

Idle valve mounted. 3 O-rings seal the valve outlet to the new valve feed flange I made. Still have to plumb in clean air supply.

 

Having removed the TB with the intent of drilling the idle air ports larger, look what I found! The manifold essensially blocks the ports except for the small gap being left by the gasket between the TB and the manifold. The dark patch between the arrows shows where the idle valve port air hit the manifold:


Here you might also be able to see how the two idle air ports align with the two Y-shaped bosses seperating the two plenums.


Not having access to a milling machine at 5pm on an end of month Friday, I decided to do the necessary mods myself using "the means available to me". This would constitute the most vulgar of p0rn so I'm not posting pics of that. Suffice it to say that now both idle air ports can feed air unrestricted into the manifolds.

With the engine at operating temp and idle valve fully closed, she idles happily at 850rpm with air being supplied via the PCV valve and the fractionally open butterflies. If I now open the idle valve fully, idle speed goes up to 1700rpm. This is good stuff as now there is likely to be sufficient air for the engine when cold via the idle valve. This we will only know on Monday as I have no time to play over the weekend.

It seems I have to withdraw any and all previous statements regarding the idle valve fitted to the Edelbrock 38783 throttle body and the idle ports as it comes from the factory. If Edelbrock's support via the forum was as it should be, they could have responded (long ago) as follows:
"The Edelbrock 7193 intake manifold (being intended for carburettor) is not envisaged to be used with the 38783 TB. The two ports on the TB that is meant to supply idle air to the engine will be blocked except for the small gap being left by the gasket between the TB and the manifold. You could of course machine the manifold to allow air from the idle air ports on the TB into this manifold"

 

Nice! Is there a problem with MSefi's website? I could not register no matter what I tried.

 

Not that I know of, but I have not been on that site in quite a while now that I can help myself.

With spam bots being so clever nowadays, it wont surprise me if your registration is subject to admin approval but then I would have expected the site to inform you accordingly.

Sometimes these sites als have hiccups....

Give it a day and try again?

 

Now that I found the "mismatch" between the TB and the intake and having sorted that, I re-installed the PWM idle valve that came with the Edelbrock 38783 throttle body and now it works as advertised! Happy!!

 

Nice !!

 

Something else I should mention is that I changed the plugs from BP6S to BPR6S when I did the EFI conversion and after that I had a problem with the ignition module overheating ( and shutting itself down on longer drives).

I was running the car with 3ms coil charge time with the BP6S plugs.

A friend of mine suggest that perhaps because the spark duration was much longer with the BPR6S plugs, it caused much longer flyback currents through the flyback diodes (internal to the coil firing transistors) and I thus reduced the coil charge time to 1.5ms and this cured the problem.

 

That's odd... I would have thought it would of acted just the opposite!!

So the change was going from a non-resister plug to a resisted plug?
And backed up the firing time from ? to 1.5ms... Humm

After the long trip and a Hot start did the Exhaust puff black? when it started?

P.S. Phil I like this build. I just have to search all the webs to find all the info you have put out. There are some very smart guys out here! Including yourself.
Great Build...Thanks for sharing it!!
Bob

 

Mopwit

Very nice build!

I have not read your complete post's, but in your photo (post #19), is that your air temp sensor in the intake plenum?

On our Drag Pak's we run the air temp sensor externally in clean ambient air above the throttle body.

I know of a few Drag Pak owners that actually mount their sensors in the radiator support for even cleaner air.


Here is a picture of another Drag Pak (not mine, but exactly the same)with the temp sensor in the air box for clean ambient air.

Please excuse me if I am seeing something else.

Cheers

Peter

 

Correct, I changed from non-resistor to resistored and that caused the ignition module to overheat.

The time I refer to is the coil charge time aka dwell (in points terms: the time the "points are closed"). I had it on 3ms with the non-resistored plugs and then had to change it to 1.5ms with the resistored plugs.

Nope, it just died when the ignition module became too hot. After 5mins or so, I could again start it but it would idle but would not support any amount of thorttle.

Thanks, let me know if there is anything in particular you want to know!

 

That is correct Peter. On the Megasquirt forums, there is often a great debate as to where the IAT should be mounted. I have taken to mounting it in the plenum as that then takes into consideration the air temp as it is INSIDE the intake. This then allows the ECU to compensate more accurately for the air temp that is entering the combustion chamber and helps with more successful heat soaked starts.

Mounting it external of the plenum, you could in principle have extremely cold ambient air and rather hot air in the plenum (eg. after a 5 min stop for shopping). This could then result in less accurate fueling for that hot start.

Is that an Edelbrock Victor intake manifold you are using? If so, please let me know how you find its performance from idle to 5000 rpm. I was really in two minds if I should take the Victor (EFI ready) or the Performer (not EFI ready). I wanted an intake more suitable for road use and I was worried that the Victor would not work with my std cam. (I also worried about its height)

 

Philip

Have you tried mounting the sensor external to the plenum? I would be interested to see if there is any change in functionality.

Although my driving is restricted to a dragstrip and peripheral roads, my Drag Pak drives perfectly (my wife drives it around pit areas like any stock street car) at all rpm's cold or hot engine (Pacific Northwest, Canada).

Our racecars do get heat soaked while waiting tech (scales and fuelcheck) after each run.

The Drag Pak intake manifold is a Factory High Ram specific to the Drag Pak Race program.

Here is a photo of the 2500 rpm - 6500 rpm performance of my car, and a photo of the Drag Pak intake manifold.

 

The IAT sensor of the Megasquirt install on my Land Rover Discovery 1 4.6 is mounted external to the plenum (clean air side of air filter) and thus measures "ambient air temp".

I found on that car that AFRs were off-target after a heat soak due to the huge difference between manifold temp and ambient. Requires 2 - 5 mins of driving before AFRs are on target again.