Lucas EPIC DSA1191 / DSA1096 ECU Hacking on a '99 Smiley

[Chug]

I'm afraid Epic electrickery is not really one of my areas, but if I'm understanding this correctly you are giving an ajustable signal from the cam position sensor to the ecu to adjust the advance?

I have a pdf of the ford technical service training manual, transit epic injection system, PM me an email address and I'll send it down the line

[marcrbarker]

The van has been away all today working and I've got it back this evening so I've made a quick test while parked only otherwise would lose the parking spot.
Well here is the thing I used and it works. It's a piece of cardboard with the parts stuck on with hot melt glue. Starting with the control in the centre as the control turns to retard direction the sound of the engine gets softer as I expected, like a petrol engine, I estimate around 10 degrees of retard. Turn control back to centre then goes back to normal diesel sound. That was about 1 volt of offset. Turn control the other way and as expected the sound gets sharper and clackier and curiously as increasing reaches a point where the ECU fails to regulate the idle speed and starts racing out of control. I'm not sure why. Eventually the red engine light came on turning in this direction. I estimate at just under 8 degrees.

A multimeter set to 2 volts scale attached to red and BLACK wires. The 1K fixed resistors were removed and the pot delivers full voltage.
The pump actually has two WHT/BLU wires, this connects to the one going to pin 19. The other wht/blue already has the other device with the two LEDs "POS & PLU" connected .

IMG_20210820_223118.jpg

So now it's working I'll explain how it works. (I'm am an electronics product designer by the way)
Cam ring (timing advance) movement is represented by a 0 to 5 V analogue voltage. The ECU at all times knows where it needs to set the voltage to, (there's 'maps' and stuff in the ecu that make it all sound much more complicated than it is, but it's really little more than getting injection to occur at 2.5 ms time before TDC irrespective of RPM) . The ECU moves the mechanism about by manipulating an actuator with a PWM signal by trial and error until it gets the voltage it wants. It doesn't know exactly what amount of PWM is required, but it does know exactly what the analogue signal is supposed to be.

The ECU trusts the cam sensor is telling the truth. The above device adds an offset voltage to the the analogue signal it's as simple as that. The ECU duly manipulates the mechanism until it gets the 'right' voltage come back. What the ECU doesn't know at all is the timing has changed. That's why there's no fault code.

[marcrbarker]

Not much more has happened with the ECU experiments since my 'timing-on-the-fly' tweaking. I didn't get to try anything else. There is definately some potential to be explored though. What's happened is after a few hill climbs tinkering with timing offset I'd given the van back to the owner to try out and then he didn't want to give it back! Ever since it's been flying around the island the van is now pulling harder and going faster than it ever had done. Which is a bit weird as I expected the DSA1096 would perform less than the original DSA1191. I'm quite sure it's not because I'd accidently reassembled out by 90° inside. I'll definately need get the van back for a day later on to straighten out the pump out inside or a motor mechanic could later come along and he'll unlikely ever get his head around having to offset it on the cambelt by 6 teeth.

So for the time being there's some conclusions I've come to. Being as there's no injector needle lift sensor on this system (to automatically accommodate errors/ drifts in alignment) means everything instead hangs off a mechanical pre-trimmed datum that's referenced back to the crankshaft pos sensor as master datum. The pump has to be precision-aligned to a reference datum because the ECU is working completely 'blind' about what the timing is (no needle lift sensor) . This is why the voltage offset tweak of timing on fly is so effective. Because it's working blind means the system is so easy to fool, just distort the info the pump sends the ECU (within reason) and so alter what the pump is doing and the ECU doesn't know any different.

I've been asking around about fuel and lack of sulphur and no one so far seems to know about it nor adding 2T oil to the fuel which gets respond like "what, you mean it's like adding Calgon each washing machine load that ends up cost more than new machine" which I respond the oil is fuel too just a bit thicker and costs more per litre. People here often don't use antifreeze so I don't think it would catch on.

[marcrbarker]

I've got an update about the plunger repair. This is the ESOS thing, a finger-pokey thing valve shutoff thingy operated by one of the back mounted solenoids. It's supposed to open and shut when you turn the engine on and off. It's supposed to have a rubber end but it dissolves away and dissapears without a trace. I'd previously put a new rubber end made out of a blob of automotive silicone and shaped.

Well I didn't feel happy with it so I'd caught up with the van and removed the plunger again which is quick and easy as there's no cover on the injection pump. A good job I did because the rubber was starting to swell which would be fine if it was a gasket but it's not desired here as it's a moving part.

20210901_140857.jpg

The plan is to try get a replacement or order one online.
In the meantime I've made a much better version that would probably out-last the proper original - use for time being. So here is step by step guide how make one for next to nothing and that's probably better.

Start off with an offcut of viton o-ring 8 mm diameter. Must be viton don't use neoprene or nitrile and don't even think about using isoprene, butyl etc.

Cut straight with razorblade a 3 mm flat disc. Put a taper profile on one end.

20210901_144045.jpg

Get the plunger and carefully put a nick into the cone shaped end with wire cutters. This is needed later.

20210901_151906.jpg

With the rubber disc, non tapered face, cut a partial hole in dead centre of with a sharp 2 mm drill taking care not to perforate it completely

20210901_151550.jpg

Then press the disc firmly over the cone shaped end making sure it goes into the hole. Now while maintaining pressure twist the disc around so that the nicked metal part cuts a recessed taper shape into the blind part of the hole, so it makes a secure retention.

Place the finished plunger standing up on a mirror

20210901_152830.jpg

to check the seating face is flat and perpendicular. I can't post more than 5 images to show end result but finishing touch is a pinhole through the centre.

Looks quite good and the end is very well secured. When the van is back in town long enough I'll fit this plunger while trying to find a new one.

[MinorMatt]

Thats a good bodge

[marcrbarker]

There's a lovely word people use here for bodge: "una chapuza" and seems to be part of the culture. I've seen some incredible chapuzas including an electric door opening control unit that's a bird's nest of electronic components I thought I had a photo but I can't find it, it's bare electronic components of the 1970s been tacked together into a 3D sculpture.

I've got a little bit more to add to ESOS plunger chapuza..
I've been doing a severe test of three types of rubber,

20210903_194309.jpg

from left to right (bottom to top): A viton before and after, B automotive silicone mentioned earlier in thread, C medical syringe plunger rubber and where it came from. I had all three submerged in 100 °C diesel for about 48 hours simmering away on the stove in a covered pot of water. The viton sample had nothing happen to it at all which is welcome news, the silicone one about the same but if anything slightly larger. The medical grade plunger rubber a surprise it has swelled to nearly double original size and seems to be disintegrating already.
I think the original ESOS plunger is meant to have a small hole through the centre. Because I have seen it on photos and the plunger solenoid itself is hollow. Mine has a hole through it and it shuts off fine with the key.

[MinorMatt]

The rubber of the pluger does not have a hole in the middle

[marcrbarker]

What I'm writing here now is for anyone who's removing the ESOS stop solenoid plunger. (the plunger has the rubber end that dissolves away over the years, then the engine runs on for 10s after the engine is turned off and goes into limp mode with engine light on.) Something to look out for while removing these pair of solenoids from the back of the pump, is the other solenoid to the plunger. There's a small O ring that can drop out unnoticed .
When that happens some rather strange things start happening, beginning with the engine suddenly failing to start for no apparent reason. There's no telltale leak either if the inner o ring's missing.
Things get even more stranger with the way the engine behaves. With the injector pipe nuts cracked open, fuel blows out normally while engine cranking, no air. Then with the injector nuts done up the engine fires up first time, but only runs for first couple revolutions, stops firing then stops, leaving the starter motor spinning disengaged. This repeats for every start attempt the same effect.
What's going on is that o ring missing completely stops the injection pump timing working. The engine would otherwise run of sorts with the timing so far out but as soon as the engine starts the ECU senses the timing way out and it promptly cuts off the fuel completely to kill the engine. A diagnostic trouble code read by scanner would say so. All this over a small o-ring .

[marcrbarker]

Update about the "Stop Solenoid Plunger" bodge repair from a year ago earlier on this thread. The bodge of made-up rubber end is still going strong no issues!

[This is the solenoid plunger at the back of the (DSA1191/DSA1096 etc.) EPIC HP injection pump, that when it goes wrong it causes the engine to run-on after you've removed the ignition key, followed red warning light and 'limp-home' mode. Then the limp-mode can be temporarily cancelled (and full power restored next start) by stalling the engine with the clutch pedal in gear before you remove key.

What I'd did was cut and prepare a small piece of VITON O-ring as a replacement end for the plunger. It was fiddly to accomplish and would probably had been better to have paid out for a whole new one. But it proves it's possible to repair things for next to nothing.]

While talking about this solenoid (pair) at the back of the pump I think I should re-mention what can happen to the other solenoid of the pair while they're removed. One of its two concentric o-rings can drop out. What happens then is the engine starts each time but after firing only run for a fraction of a second then stop. What makes this crazy is every time you test for fuel to all the 17mm slackened off injectors, there's fuel every time. The engine never runs more than a fraction of a second. Infuriating.

What's going on is both o-rings are needed in place for the injector pump timing to adjust itself while engine running, if the o-ring isn't there after the engine starts the ECU quickly realises the timing is way out - then shuts off the engine. A tiny 2 mm o-ring.