604 Megasquirt conversion

robs

Active member
VIP Paid Subscriber
Fellow Frogger
Joined
Nov 1, 2010
Messages
551
Location
Blue Mountains, NSW
It's about time I posted a bit of a blurb on how converting the 604 to Megasquirt EFI has gone.

For the last 20 years or so the car had run a Holley 320 carburettor -- an improvement on the Solex setup, but nothing to get excited about. It now has a Megasquirt II ECU running MS2/Extra firmware controlling an EA Falcon dual throat throttle body injector. Feedback is from an Innovate LC-1 wide band controller with a sensor fitted in the LHS pipe. At the back I have a 505 fuel pump drawing through a cartridge type 10 micron fuel filter.

Doesn't sound like it'd be all that much work, but the Devil's in the details and the job has called for more than a few moments of musing how to make things work or make them fit. On the whole though, I'm very happy with progress, though there are still a few things to sort out yet.

So, here are some photos and a bit of running commentary.

604msgbox.jpg
Megasquirt behind the glovebox

The photo is foreshortened, and doesn't get it across, but the MS is mounted on a cradle I made and is waaay far away. There is plenty of room for the original cardboard glovebox to slot back in. You can see the main MS DB37 electrical connector on the left, and the MAP sensor vacuum line looping around it. On the right is the RS-232 connector. The brown wire is the spare analog output from the wideband which I might hook up to an AFR gauge (maybe fit in the ashtray slot which, being built to accommodate French smokers' tastes, would easily be large enough for two gauges).

604ms01.jpg
Engine bay overview

Engine bay with everything in its proper place looks reasonably stock. The Ford air cleaner doesn't look out of place and the wiring isn't too much of a rat's nest. Mind you, the Ford air cleaner took quite a bit of adapting to fit. The Ford TBI was towards the front of the engine bay. Since the 604's second carb was mounted near the back, just turning the filter around and relocating the snorkel was my original plan, but the filter ended up at a crazy rakish angle when I tried it. So it was a cut and reweld job which worked out OK (as long as you don't look too closely).

604ms02.jpg
TBI detail

I've mounted the TBI on a bit of 1/4" stainless plate with four studs. This plate bolts to the previous Holley adaptor plate which in turn mounts to the manifold through four recessed allen screws. This two plate arrangement was necessary as the Solex and TBI mounting holes overlap but don't coincide. This sandwich has left the top of the air filter pretty close to the bonnet. I haven't measured with plastigage (OK plasticine); it's not rubbing but it's close.

I bent up 5/16" copper pipe to feed fuel and take away return. The thing flows like a garden hose, so there is plenty on tap. You can see the two injectors front and centre. There are two TPS sensors on the other side. Only using one; I guess the Falcon uses the other for something like idle feedback or some such. Throttle was a bit of a conundrum. I ended up using the original Falcon ball and socket connection so I could use the compression return spring. Rigging up an extension spring for throttle return didn't look straightforward ('cause it needed to go round a corner). The throttle stop is the somewhat jury rigged bit of steel mounted on RHS in the photo. Could do with a bit of stiffening up.

604ms03.jpg
Relay board detail

Left to right we have a shiny red Bosch GT40 coil, then, mounted on a bit of bathroom trim aluminium, Innovate LC-1 controller, Bosch electronic ignition trigger box (ex Volvo) and Megasquirt relay board which gives a single point to draw all connections together. At the bottom of the relay board you can see the DB37 connector which sends a chunky cable through the firewall to the ECU. The knot of wires further to the right is the relay and thermostat for the cooling fan which has been on the car for years. Will probably get the MS controlling this and get rid of the old thermostat.

The Bosch coil seems to deliver enough sparks for me and has no bother starting the car. The ballast resistor had been playing up so a full 12V coil took that out of the equation. I have a spare Volvo dizzy and may lock it up and use the MS for full ignition control to see how that goes.

604msfp01.jpg604msfp02.jpg
Fuel pump/filter

Two shots of this because I was too lazy to jack the car up and snap from further away. This is on the LHS in front of the rear crossmember behind the rear passenger footwell. I made up a mild steel mounting plate and drilled holes to mount it to the footwell. Then I was able to position everything on the bench without having to lie under the car.

On the left is a large copper pipe feeding a 1/2" hose to the inlet of the pump. The fuel filter is the shiny black thing on the right. Somewhat unconventionally, the pump draws through the filter which allows the filter to do double duty as a surge tank. I did this because I didn't want to be fitting a 1/2" outlet to the fuel tank. The pump requires a large inlet because it uses the fuel for cooling. The worry was that the fuel would get hotter and hotter... but that doesn't seem to eventuate (phew). There are four connections on the filter: from tank, to injector, injector return, return to tank. I didn't use the original tank return line because it (strangely) returns at about the 1/3 full level in the tank. I sent the return up to the breather at the top of the tank and have a check valve in the line so we don't draw air down this line. Since the return line should mostly be returning air accumulated at the filter, this seems satisfactory.

The fuel pump plumbing was in many ways the most complicated part of the setup. Pretty happy with it though. Despite the apparent tangle of hoses, nothing is actually rubbing on anything else and, in the flesh, it doesn't look any worse than the 604TI's pump and filter arrangement in the workshop manual.


Not pictured is the coolant temp sensor (tucked in to the valley in place of an existing temperature based vacuum switch), or the air temp sensor (a thermistor put in a drilled out old plastic medicine bottle, cable tied behind the front bumper).

So, how does it go? To begin with, I was astonished that, not only did it start first time, but it idled (at a very lumpy 450rpm or so). Definitely a promising beginning. I haven't yet finished tuning (waiting for the Christmas madness on the roads to abate), but the car has already been transformed. It still has all the stately graces a 604 should have, but the old dead areas of unresponsive throttle have gone. If you move your foot on the throttle, you are left in no doubt that the engine has got the message. Of course it's still no fire breathing monster, but it is really pleasing to feel it so responsive.

The other thing about the performance is that it's willing when cold (well, overnight in Sydney isn't exactly cold at the moment, but the car certainly used to be sluggish untill fully warmed up).

It'll be a while before I have economy figures. Obviously the consumption isn't going to be at its best while I'm tuning, but I'm hopeful that the economy will be very much better with mod cons like EGO feedback to help it out.

Problems

Yes there are still several problems to iron out. The biggest problem is the noise from the fuel line. Because I'm drawing through the fuel filter, there is just a 2" length of hose between the pump and the original steel line which is mounted direct to the floorpan. Every little pulse from the pump is turned into magnificent high fidelity sensurround sound with the passenger compartment making a wonderful soundbox. For the moment, I've left the fuel lines dangling, still noisy, but at least the banshee is a bit less amplified. My plan is to put a blind vertical pipe between the fuel pump and the fuel line This pipe will have air at the top which should work as a poor man's (or cheapskate's) accumulator, evening out the pulses. Kinder to the fuel line, and to my ears.

Probably the next biggest problem is the throttle throw. Having mounted the cable to the ball/socket connector, it only needs to move about half as far as the old Holley did. Consequently, the pedal is about 1 1/2" from closed to wide open. This has already led to a couple of cases of unwanted wheelspin, which is just not in keeping with the 604 image. I plan to move the pedal pivot up a couple of inches to increase the throw, but it's acceptable for driving at the moment.

I'm running full manifold vacuum (not ported vacuum) to the distributor advance unit. This seems to be working OK and I haven't had problems with pinging on take off. The engine does seem to stay cooler too, but I'm keeping a watching brief on this score.

The other problem is a whistle coming from the intake tract as you open the throttle (only while the airflow changes, it's OK on steady throttle). It's a minor irritant which I have already improved by rounding off some of the joins between adaptor plates. Can probably live with this one.

I'm sure there are a number of item's I've left out, but that's the guts of the job described. These EA Falcon throttle bodies are just the ticket when you want an electronic replacement for a carburettor. Be interesting to see how a converted carby 504 would compare with a 504 TI. Would be surprised if it didn't end up with both better performance and economy.

Have fun,

Rob.
 
Well done Rob, I hope you end up sorting out the minor issues. I still remember my 604, and the issues I had occasionally with it staling intraffic (and not wanting to restart!!). It ended up catching fire once it got into the hands of a "fiddler", and was destroyed.
 
.... The biggest problem is the noise from the fuel line. Because I'm drawing through the fuel filter, there is just a 2" length of hose between the pump and the original steel line which is mounted direct to the floorpan. Every little pulse from the pump is turned into magnificent high fidelity sensurround sound with the passenger compartment making a wonderful soundbox.


Rob,

A really fantastic effort.

Re the pump pulsing - Have you mounted the pump on isolating rubber mounts?


After the silly season, we'll need to compare notes.


con...
 
Always happy to compare notes Con, and in any case I hope to post occasional updates to this thread.

There is 1/4" neoprene between the pump and the clamp around it. I was worried that this might conduct too much noise and that I should use some sort of soft foam instead. From outside you can indeed hear the pump humming away. Inside, it is utterly drowned out by the fuel line faithfully conducting every pressure pulse into the cabin. Hopefully, introducing a permanent air bubble into the fuel line will smooth out the pulses and I'll be able to decide whether the pump noise is acceptable.

Sad that so many 604s went to fire. This one did have the valley full of fuel once, but got away with it.

The car hasn't been neglected, but it certainly hasn't been used as much as it might in recent years. The way it's looking since the conversion, it will make a fine backup daily driver, no longer restricted to special occasions/last resort. Might even motivate me to reinstate the aircon.

Have fun,

Rob.
 
Just thinking about this air bubble idea Rob. Wouldn't that be the same as an air lock and because the air will compress while the fuel won't you'll probably end up with hugely reduced fuel pressure if in fact it still pumps at all.
I'm thinking all the noise you are getting must be because the line is not held firmly enough, so it's akin to water hammer when not enough clips are placed on pipes in a wall. I therefore suggest a few more of those fuel line clips at regular intervals may solve your noise problem.

It sounds like an exciting project. I'm very interested in your final assessment once fully sorted.
 
I've thought of this idea with a few cars, 2 litre 505 being one.
I think an MS controlled mono point injection on a 205 Si would be excellent.
The Si suffers because of its dependance on the O2 sensor I reckon and it would be good to be able to tune for performance throughout the range.
Graham
 
Last edited:
Great report Rob, as always, just a real pleasure to read, great that the 604 has still got life in it.
 
Inside, it is utterly drowned out by the fuel line faithfully conducting every pressure pulse into the cabin.
Rob.

What about using a commodore pulse dampener after the pump?????
I've always run one on my efi and never heard any of the pulsing you mention.

Jo
 
Just thinking about this air bubble idea Rob. Wouldn't that be the same as an air lock and because the air will compress while the fuel won't you'll probably end up with hugely reduced fuel pressure if in fact it still pumps at all.
I'm thinking all the noise you are getting must be because the line is not held firmly enough, so it's akin to water hammer when not enough clips are placed on pipes in a wall. I therefore suggest a few more of those fuel line clips at regular intervals may solve your noise problem.

I hope it won't work like an air lock. I think I haven't been clear enough on the "air bubble" idea (described earlier on as a poor man's accumulator). Here's a diagram:
accum.jpg
So fuel flows along the tee, an air bubble sits above. As pressure fluctuates in the fuel, the air in that bubble compresses and releases accordingly, smoothing things out. The air should stabilise at the average pressure in the fuel line, without impeding flow at all. Some of the air will get drawn into the system when there is a sudden drop in fuel pressure, but as long as the reservoir is large enough (and I don't think it needs to be large at all in this case) an all-time high water mark will be reached. Above this there will always be air to smooth things out.

If you put a diaphragm at the liquid level in the diagram it would be pretty much the same as the pulse damper Jo suggested. If you put a mechanical spring above that diaphragm, it'd be a conventional hydraulic accumulator. I don't see what benefit there is in having a diaphragm (or spring) when (even with the considerable body roll a 604 can manage) the air bubble will always be above the fuel.

In any case, it's cheap and easy and I'll soon give it a go. If my DIY physics experiment doesn't work out, the Commodore pulse damper will probably do the trick.

Thanks all for sharing your thoughts. This is a project I've had in mind for quite a few years and I'm glad there is interest in it here.

Have fun,

Rob.
 
Hi -Was rather sad that the dodgy -were they Lucas fuel pumps-failed -Cortinas had the same problem as the inlet/outlet tubes were only a press fit-after time heat or whatever the much metal would not hold the tubes & Up she Would Go.

If anyone out there has a 604 they are using i would strongly advise replacing the mechanical pump with an electric one & adding a prissure reg. while they are at it.

Cheap Insurance.

CheersPekay.
 
Last edited:
Cortina problem I thought was people not replacing the crimped hose clamps ,perfectly satisfactory but need to be replaced after removal of hose.

Combination of Ford saving a few cents and the owner's/ mechanics stupidity.

Graham
Hi -Was rather sad that the dodgy -were they Lucas fuel pumps-failed -Cortinas had the same problem as the inlet/outlet tubes were only a press fit-after time heat or whatever the much metal would not hold the tubes & Up she Would Go.

If anyone out there has a 604 they are using i would strongly advise replacing the mechanical pump with an electric one & adding a prissure reg. while they are at it.

Cheap Insurance.

CheersPekay.
 
Had a 4cv catch fire this same way ,the metal tube came out of the alloy pump housing spraying fuel all over the distributor, first thing i knew about it was smoke coming out of the demister vents on the corners of the windscreen, managed to put it out but the plug leads and some of the wiring was melted i had heard that 604 were prone to catching fire but had never found out why PUGS
 
excellent write-up, very thorough. this is a very clever mod!

with regard to the noise issue -- whatever other fixes you put in place, you should also consider additional sound dampening underneath the carpet pad inside the car. i've pulled out my carpeting to replace it, and underneath the carpet pad i've lined my floorpan with dynamat, which is essentially self-adhesive asphalt sheets (not dissimilar to roofing material) with a foil backing. it's commonly used here in the US as part of high-end audio installations. here's an in-progress shot from my installation:

753b0693.jpg


i did my entire floorpan, this is massive overkill and adds a bit of weight, you could only apply it right above the source of the noise.

i am not sure if dynamat is sold in australia but i am sure you have something similar there. just a suggestion...

andré
 
Just a quick update on the "physics experiment". As per the diagram, I teed off a blind hose from just after the fuel pump to a reasonably high spot. Fuel line was nice and quiet after that -- for the first, second, third time. But gradually, the noise was returning. I guess that the sudden change in pressure when the pump stops manages to cause enough of a splash in the line that a little more air is drawn out each time you turn off.

Not wanting to admit defeat, I started having thoughts about putting together some sort of gadget with baffles and so forth, but then it dawned on me that, no matter how fancy I make it, air can dissolve in fuel so (to an increased extent when it's under pressure). So any solution that doesn't involve a membrane of some sort is going to fail eventually. Mind you, I don't see that anything terribly fancy is called for; a (fuel resistant) balloon in a cannister would work.

But Jo's right -- no need to experiment when the clever people at Bosch have off the shelf units built for the purpose. Still, I've enjoyed thinking about it.

Apart from the experiment, no other progress on the conversion, but will soon get stuck back in and get it finished.

Have fun,

Rob.
 
The story continues...

Jo was indeed right. I have fitted a pulsation damper and the fuel line is now silent. The line is clamped hard against the body again and looks tidy. Unfortunately, this hasn't banished all the noise.

I'm pretty sure the fundamental problem is that I chose the wrong fuel pump, and have exacerbated the problem by putting a fuel filter before it. So I'm now troubled with cavitation on the pump's suction side. This isn't nearly as noisy as the fuel line pulsations, but cavitation is never good news.

As I said, I chose the wrong pump. The EA Falcon manual's fuel pump tests stipulate 100kPa and at least 90l/hr fuel delivery at that pressure. That's pretty modest by any standard, and the 505 pump I chose (in order that one spare pump would do either car) is meant to be capable of something like 400kPa and 160l/hr. The Falcon's regulator seems to be coping with the flow, but it's obviously churning the petrol at an unnecessary rate.

Continuing on the physics experiment theme, I've been trying out dropping the voltage to the pump by putting various resistors (well, OK, I don't have a stock of power resistors, but things like headlight globes will do for an experiment) in series with the pump. My limited experiments show that the pump still runs (and very quietly) at 4.5V and is still quiet at 6V. I only did one flow test. At 4.5V it delivered the Falcon's pressure, but only about half the flow volume. All that pretty much exhausted my enthusiasm for experiments with jury rigged resistors, but I'm hopeful that a voltage of something like 8V will deliver the goods, while still having a good safety margin against pump stalls. I'll see if I can recover my enthusiasm and run a couple of headlights in parallel. If I can get close enough we'll find out if Nirvana is indeed at 8V.

I'd be interested if anyone has any "NO!!! DON'T DO IT!!!" advice. I can't see any problem with it myself, but "there's none so blind as those who can't see" (or something like that :).

Have fun,

Rob.
 
Rob,

Rather rooting around with headlight lamps and fragile filaments, heating power resistors consider using one of these:

http://secure.oatleyelectronics.com...d=625&osCsid=d89f8af740141043879a6199068fb712
(K252 - DC MOTOR SPEED / HOT WIRE CONTROLLER KIT)

This kit uses PWM (pulse width modulation) to control the motor speed and the speed is constantly variable via a pot.
You can set the minimum speed via a series resistor. There is practically no heating and much improved efficiency.

The kit is cheap and easy to build. I've built this kit a few times to replace burnt up aircon fan switches/ resistors in older cars.
 
I'm no electronics wiz and was actually a bit proud of thinking of using a headlight globe for the experiment, at any rate there was a warm glow coming from somewhere when I turned it on.

I'm also a bit wary of PWM -- my mum always told me not to play with the light switch and there must have been some reason. But I'll happily take your word for it. I've not quite ordered the kit yet, it's in my basket waiting for them to get back to me with what size of box I should get to put it in. With a bit of luck I'll have it by the weekend.

Many thanks for the suggestion -- even I get sick of rooting around eventually.

Have fun,

Rob.
 
It's very small about 50mmx 50mm.

Mine usually come in jiffy padded bag.

Expect endless emails from them and eventually mailing advice.

Use a hot clean soldering iron and get some solder wick to clean up any dags. Solder the mosfet last and avoid touching the leads as much as possible.

Btw the rooting around jibe was in good faith - I've realized it can be taken two ways.
 
would it be possible to get a fuel pressure sensor and use to control the voltage to the pump closed loop?
then it would run the pump only as hard as is needed to get the pressure that you have set
 
would it be possible to get a fuel pressure sensor and use to control the voltage to the pump closed loop?
then it would run the pump only as hard as is needed to get the pressure that you have set
That would seem to be the deluxe approach, though it strikes me there might be a matter of lag between speeding the pump up and the extra pressure becoming available at the injectors. Might be better to let the mechanical regulator regulate pressure. Instead, take pump feedback from flow in the return line (if there is such a gadget). A chance for more physics experiments...

Hmmm. Just googled and turned up http://www.futurlec.com.au/Flow_Sensor.jsp. The $12.90 gadget could sit in the return line and you adjust pump duty cycle to try to keep a flow of something like 15l/hr. Counting pulses from this device seems easy enough and one microprocessor could clearly do this and intelligently drive PWM for the pump at the same time. Sounds like fun.

For now, getting my pump to dumbly (as in silently) do what the original dumb in-tank pump did will be pleasing enough.

And don't worry robmac, I wasn't at all offended at the rooting around remark. It was pretty much what I'd been doing after all. Pretty much what I do most of the time come to that.

Have fun,

Rob.
 
Top