S3C ECU problem
Discussion
Help!!
I've been banging my head against this one for weeks. I reflect that 'a little knowledge is a dangerous thing' but you've got to start somewhere!
The engine (part rebuilt, to replace the infamous 2.4, but that's a story for another time) won't start.
It turns over enthusiastically; there's a spark at 12deg BTDC on no.1 compression (and all the others). But any attempt to start it drastically floods the plugs (If I remove a plug I can shake two or three drops out of it!), and the engine won't fire at all.
A semi-rational program of investigation has led me to look at the injector firing, and my dinky Hantek USB 'scope gives me the following ECU output pulses on the two banks of injectors:
or:
Incidentally, the timebase increment is five minor divisions on the horizontal axis.
The triggering rise in voltage is 'switch on', and then you can see the drop as the starter kicks in, and a brief pause before the first pulse.
And that's a 30ms pulse of fuel three times a revolution (I presume, one for each PIP signal). Through each injector.
Now the wonderful Charles O Probst (whom God preserve) says in his book 'Ford Fuel Injection & Electronic Engine Control 1988-1993' : "Injection time - the pulse time that delivers the amount of fuel required - may be as short as one millisecond or as long as fifteen milliseconds."
I haven't been able to find any other hint as to appropriate injection times, but another way of looking at it from these waveforms it that the injection pulses are taking up around one-eighth of the available time between cycles; that seems far too much fuel for cranking or idle speed and enough to cause my flooding.
So what would persuade the ECU to default to such high continuous fuelling at cold cranking speed?
Well, the ECT and ACT temperature inputs at the ECU are fine for a cool day (I read that they should increase fuelling on cold cranking, but surely not by that much).
And the throttle pot and idle valve and MAP (no airflow sensors here) appear to be trying to do their jobs, viewed from the ECU.
Have I got a fried ECU? It's not the kind of failure you'd expect to see, but then, what is?
It seems much more likely that something is TELLING the ECU to do this. But what?
Any advice from any of you wonderful people with knowledge or experience of EEC-IV systems would be gratefully received.
I've been banging my head against this one for weeks. I reflect that 'a little knowledge is a dangerous thing' but you've got to start somewhere!
The engine (part rebuilt, to replace the infamous 2.4, but that's a story for another time) won't start.
It turns over enthusiastically; there's a spark at 12deg BTDC on no.1 compression (and all the others). But any attempt to start it drastically floods the plugs (If I remove a plug I can shake two or three drops out of it!), and the engine won't fire at all.
A semi-rational program of investigation has led me to look at the injector firing, and my dinky Hantek USB 'scope gives me the following ECU output pulses on the two banks of injectors:
or:
Incidentally, the timebase increment is five minor divisions on the horizontal axis.
The triggering rise in voltage is 'switch on', and then you can see the drop as the starter kicks in, and a brief pause before the first pulse.
And that's a 30ms pulse of fuel three times a revolution (I presume, one for each PIP signal). Through each injector.
Now the wonderful Charles O Probst (whom God preserve) says in his book 'Ford Fuel Injection & Electronic Engine Control 1988-1993' : "Injection time - the pulse time that delivers the amount of fuel required - may be as short as one millisecond or as long as fifteen milliseconds."
I haven't been able to find any other hint as to appropriate injection times, but another way of looking at it from these waveforms it that the injection pulses are taking up around one-eighth of the available time between cycles; that seems far too much fuel for cranking or idle speed and enough to cause my flooding.
So what would persuade the ECU to default to such high continuous fuelling at cold cranking speed?
Well, the ECT and ACT temperature inputs at the ECU are fine for a cool day (I read that they should increase fuelling on cold cranking, but surely not by that much).
And the throttle pot and idle valve and MAP (no airflow sensors here) appear to be trying to do their jobs, viewed from the ECU.
Have I got a fried ECU? It's not the kind of failure you'd expect to see, but then, what is?
It seems much more likely that something is TELLING the ECU to do this. But what?
Any advice from any of you wonderful people with knowledge or experience of EEC-IV systems would be gratefully received.
I'm not familiar with that ECU, but given that you have the equipment to hand I'd be inclined to see whether the fueling varies with throttle pot reading or throttle plate angle, or with coolant temp reading. This would give you some confidence that the ECU was seeing valid sensor readings. I don't know how likely it is that you've got the throttle pot wired backwards or anything like that, but anything like that might show up when you vary the sensor input.
Presumably you're running the the injectors unplugged while you capture this information to avoid pointlessly flooding the engine.
Presumably you're running the the injectors unplugged while you capture this information to avoid pointlessly flooding the engine.
Thanks for the quick reply.
I'll try what you suggest tomorrow (perhaps also add a PIP trace to see how the injection pulses line up).
Throttle pot ECU input at closed and open throttle (and Vref) looked good on the DVM, as did air and coolant temp inputs. But I'll check again.
I have a suspicion that until the engine fires, the fuelling (and certainly ignition advance) are not supposed to be affected by other inputs, but we'll see.
No, the injectors are connected, Since the ECU grounds them to operate them, I wouldn't get a trace without them in circuit, and you can see the voltage spike when the transistor shuts them off. But I've removed the fuel pump relay and released the fuel pressure.
Thanks again: good to have somebody to talk with!
I'll try what you suggest tomorrow (perhaps also add a PIP trace to see how the injection pulses line up).
Throttle pot ECU input at closed and open throttle (and Vref) looked good on the DVM, as did air and coolant temp inputs. But I'll check again.
I have a suspicion that until the engine fires, the fuelling (and certainly ignition advance) are not supposed to be affected by other inputs, but we'll see.
No, the injectors are connected, Since the ECU grounds them to operate them, I wouldn't get a trace without them in circuit, and you can see the voltage spike when the transistor shuts them off. But I've removed the fuel pump relay and released the fuel pressure.
Thanks again: good to have somebody to talk with!
mentall said:
Throttle pot ECU input at closed and open throttle (and Vref) looked good on the DVM, as did air and coolant temp inputs.
My thought was that these inputs ought to be affecting the fueling, and if you the ECU adjusting the fueling sensibly in response to them then you know that whole part of the system is working.Thanks, Penelope. Yes, the pressure is regulated to 2.5 bar (which the Ford Grandad manual says is OK) and this drops a bit when I artificially put vacuum on the regulator.
Thanks, Steve: a good shout. I'll have to get hold of some suitable VR's: nothing in stock here.
I'll try with the 'scope again this afternoon, varying throttle position, and report back if anything different happens.
John
Thanks, Steve: a good shout. I'll have to get hold of some suitable VR's: nothing in stock here.
I'll try with the 'scope again this afternoon, varying throttle position, and report back if anything different happens.
John
Here are two more shots: one with throttle closed and one at WOT.
Throttle closed:
Wide-open-throttle:
You'll see that I've added a PIP (Profile Ignition Pick-up) signal input as channel 3.This is the only thing that tells the ECU what the oily bits are doing.
Incidentally, all these come from back-probing the ECU connector.
So throttle position is not affecting fuelling.
I've found more documentary info in the 1995 Car Mechanics article often quoted on this forum:
"Injectors - Duration 4-6ms hot, 8-12ms cold."
but then it says: Acceleration enrichment - 30+ms (full load)"!
So, is my ECU trying to apply acceleration enrichment at startup? What would cause it to do so?
Here's a few DVM measurements (again, by back-probing)
Vref pin 26: 5.01V That's OK
ACT pin 25: 3.45V .
ECT pin 7: 3.45V These correspond to somewhere around 15deg C according to Probst. Today it's 17deg here
TPS pin 47: 0.39V closed throttle, 4.40V at WOT. The article says it should be 0.6-4.5V
MAP pin 45: 5.1V pk-pk, 159.6Hz, 50% duty square wave. Probst says this means atmospheric pressure almost exactly! and I've previously tried applying vacuum, and got sensible results.
Could the ECU be 'remembering' something bizarre? I'm going to disconnect the battery (and Keep Alive Memory), and have a nice cup of tea while it forgets. Watch this space!
Throttle closed:
Wide-open-throttle:
You'll see that I've added a PIP (Profile Ignition Pick-up) signal input as channel 3.This is the only thing that tells the ECU what the oily bits are doing.
Incidentally, all these come from back-probing the ECU connector.
So throttle position is not affecting fuelling.
I've found more documentary info in the 1995 Car Mechanics article often quoted on this forum:
"Injectors - Duration 4-6ms hot, 8-12ms cold."
but then it says: Acceleration enrichment - 30+ms (full load)"!
So, is my ECU trying to apply acceleration enrichment at startup? What would cause it to do so?
Here's a few DVM measurements (again, by back-probing)
Vref pin 26: 5.01V That's OK
ACT pin 25: 3.45V .
ECT pin 7: 3.45V These correspond to somewhere around 15deg C according to Probst. Today it's 17deg here
TPS pin 47: 0.39V closed throttle, 4.40V at WOT. The article says it should be 0.6-4.5V
MAP pin 45: 5.1V pk-pk, 159.6Hz, 50% duty square wave. Probst says this means atmospheric pressure almost exactly! and I've previously tried applying vacuum, and got sensible results.
Could the ECU be 'remembering' something bizarre? I'm going to disconnect the battery (and Keep Alive Memory), and have a nice cup of tea while it forgets. Watch this space!
I'd hoped that the throttle position would affect the fueling and slightly surprised that it didn't, but I don't know whether that indicates a problem.
I understand it's using MAP for load. Do you know where the MAP sensor is physically? I don't know how much depression is normal for that engine when cranking with the throttle closed, but I would expect to see some depression, and I'd expect the amount of depression to affect fueling while cranking. Are you able to measure the change in MAP sensor reading while cranking with the throttle closed?
I understand it's using MAP for load. Do you know where the MAP sensor is physically? I don't know how much depression is normal for that engine when cranking with the throttle closed, but I would expect to see some depression, and I'd expect the amount of depression to affect fueling while cranking. Are you able to measure the change in MAP sensor reading while cranking with the throttle closed?
Yes, Thommo, these are made during cranking: you can see how the supply voltage recovers as the starter motor gets going.
It's my guess too that fuelling during cranking is fixed, but who can tell?
Presumably I've got six PIP's per distributor revolution, and so three per engine revolution. And the scan shows five rising edges over 560ms duration. 5 divided by 0.56 in 1 second = about 9 PIP's per second or 540 per minute, equals 180 rpm, at the start of cranking.
Googling "engine cranking speed" tells me that 50-100 rpm is necessary for a petrol engine: so not so bad. But I'd be grateful to anyone who would check my maths! And of course there are no spark plugs in at the moment!
Yes, Green, another good call. I can easily test the MAP sensor during cranking, with plugs in..
The tea has spurred me to think: the only thing that SHOULD be able to convince the ECU that we're demanding flank power is a signal from the Throttle Pot (although its input to the ECU looks good). Several things there to investigate, but no more time today. I'll get back to it tomorrow.
Thanks, guys.
John
It's my guess too that fuelling during cranking is fixed, but who can tell?
Presumably I've got six PIP's per distributor revolution, and so three per engine revolution. And the scan shows five rising edges over 560ms duration. 5 divided by 0.56 in 1 second = about 9 PIP's per second or 540 per minute, equals 180 rpm, at the start of cranking.
Googling "engine cranking speed" tells me that 50-100 rpm is necessary for a petrol engine: so not so bad. But I'd be grateful to anyone who would check my maths! And of course there are no spark plugs in at the moment!
Yes, Green, another good call. I can easily test the MAP sensor during cranking, with plugs in..
The tea has spurred me to think: the only thing that SHOULD be able to convince the ECU that we're demanding flank power is a signal from the Throttle Pot (although its input to the ECU looks good). Several things there to investigate, but no more time today. I'll get back to it tomorrow.
Thanks, guys.
John
mentall said:
... I can easily test the MAP sensor during cranking, with plugs in.
Well written and very interesting and enlightening. I hope you get to the bottom of the problem.On thing confused me is that I don't have a MAP (manifold absolute pressure) sensor fitted to my engine. It's not present on the TVR 2.9 wiring diagram either. Why do you have one ?
Are you sure that you have the correct injectors ?
The originals I found on my 2.9 car, Bosch 0280 150 757, are rated at 15.9 Ohm and 200 cc/min.
Maybe you have injectors which are rated at a much higher output ?
I'd say the calcs are slightly out as if you consider the ignition signal, there will be 6 complete event cycles per two revolutions of the crank, so your measurement should be made between 1st & 7th rising edge (2 crank revolutions) or 1st and 4th rising edge (1 crank revolution).
Measuring between 1st and 5th rising edge covers 4 events, so 560/4 =140ms per event. So ((1/0.140)x60)/3=148rpm, still a respectable cranking speed although that will drop with the compression load of the plugs being back in place.
Some readings whilst it's running will prove all I would think. Could always hook up an arbitrary signal generator and simulate the pickup feed. That way you could dial in different rpm and see what the fuelling then does as the temp, map and throttle inputs are varied.
All interesting stuff 👍
Measuring between 1st and 5th rising edge covers 4 events, so 560/4 =140ms per event. So ((1/0.140)x60)/3=148rpm, still a respectable cranking speed although that will drop with the compression load of the plugs being back in place.
Some readings whilst it's running will prove all I would think. Could always hook up an arbitrary signal generator and simulate the pickup feed. That way you could dial in different rpm and see what the fuelling then does as the temp, map and throttle inputs are varied.
All interesting stuff 👍
Yup, that's it. The S3C has no AFM's, but its ECU calculates load from Manifold Absolute Pressure which it gets from the MAP sensor on the bulkhead.
This seems to me a simpler solution, with a less fragile sensor, and more work done by software. I wonder why the Ford engines with catalysers are set up this way but not the non-cats. I'd be interested to hear opinions.
Of course, like (I guess) most 1990ish S3C's, mine has no cats now!
This seems to me a simpler solution, with a less fragile sensor, and more work done by software. I wonder why the Ford engines with catalysers are set up this way but not the non-cats. I'd be interested to hear opinions.
Of course, like (I guess) most 1990ish S3C's, mine has no cats now!
TVRees said:
Just noticed the title of your post - is it because you have an S3 C maybe ?
S3C has cats, lambda sensors, MAP sensor, no air flow meters.S2 or S3 has airflow meters but not the rest.
Back to basics - I'm not discounting the ecu being goosed, but it seems to me there is something more fundamental going on here - if it has a spark and it has fuel, it should start. It might run like a pig, but there should be some life and it should at least try to start.
So I'm going to ask a question that I hope isn't too insulting, but we all make daft mistakes.
Are you absolutely certain that when you rebuilt the engine, you have timed the spark to occur on the compression stroke (i.e. Not out by 1/2 a turn of the camshaft but a full turn of the crankshaft)? Does the rotor arm point to no 1 plug lead on the compression stroke? (One full crankshaft turn after No1 cylinder's valves are "on the rock)
As I say, don't mean to insult, but when you've tested everything else, then whatever is left, however improbable, is worth checking again.
Thanks: I thought you might be along soon!
More of the back story: when I discovered the S3C had had a 2.4 (ACE... code) engine dumped in it, but coupled to an
88BB-12A650-LB ECU (which Fordopedia says is right for a 6/88 to 7/89 Scorpio 2.9, with Cats, Lambdas and MT75 gearbox, and which 3 other guys on here told me was what they had in their '90ish S3C's), I gave up trying to make it run properly.
I was lucky enough to be given a 'spare' 2.9 engine out of an S3 by a (very generous) guy on PH. The top end had been stripped and partly cannibalised but everything from block deck level down appeared untouched.
I measured the bores: they were part-worn and showed a modest 'step' on the thrust side, and there was no discernible rattle or lost motion in the bearings or cam drive. This was not a professional assessment, and if in the long run the engine needs reconditioning then I won't be disappointed, but I thought that it might give me a runner to play with in the short term.
I didn't try to check the cam timing directly, but No 1 compression TDC (tested carefully with a screwdriver down the plughole) appeared to match the TDC mark on the pulley when I was setting the tappet clearances. I rebuilt the heads replacing missing parts with bits from the from the 2.4 (mainly, the pushrods, fuel rail and injectors) which I checked were compatible on Fordopedia. Incidentally, it's a solid-lifter engine, as was the 2.4.
I checked the base ignition timing, adjusting the distributor position by 'barring' the engine over with a socket and long handle. There's a spark as accurately as it's possible to get it, at 12deg BTDC on No 1 plug. And I checked again that it's actually compression TDC, by sticking my thumb in the plughole and feeling the compression.
The sparks comfortably clear a 25mm test gap in air. But I replaced the coil with a new one anyway.
The engine DID run (badly, but for long enough to get warm and to check the spark timing with a timing light) after I first put it back together in the car a couple of months ago. It wouldn't tick over without a wedge under the throttle stop, and it wouldn't pull. Diagnostics, and some reading, suggested it might be a TPS issue. I'd checked the pot, and it tested OK, but I fitted a new one anyway. The TPS input at the ECU connector was clearly wrong, so I dismantled and refitted the throttle linkage from pedal to pot: this must be a very common story! So I now have a smooth progression, visually tight throttle discs and a convincing reading of ECU input (see above).
And a really gross flooding problem!
I have rechecked using the old throttle pot, and the problem doesn't go away. And I sent the injectors away for cleaning: they came back with a good report, but haven't solved the problem.
Reading all the above convinces me that the problem must be at the throttle pot or its connections (including earth) to the ECU. Nothing else but a misinterpreted TPS value could convince the ECU to deliver such gobs of fuel at the wrong time. And it seems such an unlikely mode of failure of the ECU itself. I think I know where to go next.
Thank you for reading so far: you'll understand that I am having this enjoyable discussion with myself at the same time as with you all!
And I'll never resent kindly-meant suggestions: I'm more aware of my ability to c@ck things up than anybody else could ever be!
Thanks again
John
More of the back story: when I discovered the S3C had had a 2.4 (ACE... code) engine dumped in it, but coupled to an
88BB-12A650-LB ECU (which Fordopedia says is right for a 6/88 to 7/89 Scorpio 2.9, with Cats, Lambdas and MT75 gearbox, and which 3 other guys on here told me was what they had in their '90ish S3C's), I gave up trying to make it run properly.
I was lucky enough to be given a 'spare' 2.9 engine out of an S3 by a (very generous) guy on PH. The top end had been stripped and partly cannibalised but everything from block deck level down appeared untouched.
I measured the bores: they were part-worn and showed a modest 'step' on the thrust side, and there was no discernible rattle or lost motion in the bearings or cam drive. This was not a professional assessment, and if in the long run the engine needs reconditioning then I won't be disappointed, but I thought that it might give me a runner to play with in the short term.
I didn't try to check the cam timing directly, but No 1 compression TDC (tested carefully with a screwdriver down the plughole) appeared to match the TDC mark on the pulley when I was setting the tappet clearances. I rebuilt the heads replacing missing parts with bits from the from the 2.4 (mainly, the pushrods, fuel rail and injectors) which I checked were compatible on Fordopedia. Incidentally, it's a solid-lifter engine, as was the 2.4.
I checked the base ignition timing, adjusting the distributor position by 'barring' the engine over with a socket and long handle. There's a spark as accurately as it's possible to get it, at 12deg BTDC on No 1 plug. And I checked again that it's actually compression TDC, by sticking my thumb in the plughole and feeling the compression.
The sparks comfortably clear a 25mm test gap in air. But I replaced the coil with a new one anyway.
The engine DID run (badly, but for long enough to get warm and to check the spark timing with a timing light) after I first put it back together in the car a couple of months ago. It wouldn't tick over without a wedge under the throttle stop, and it wouldn't pull. Diagnostics, and some reading, suggested it might be a TPS issue. I'd checked the pot, and it tested OK, but I fitted a new one anyway. The TPS input at the ECU connector was clearly wrong, so I dismantled and refitted the throttle linkage from pedal to pot: this must be a very common story! So I now have a smooth progression, visually tight throttle discs and a convincing reading of ECU input (see above).
And a really gross flooding problem!
I have rechecked using the old throttle pot, and the problem doesn't go away. And I sent the injectors away for cleaning: they came back with a good report, but haven't solved the problem.
Reading all the above convinces me that the problem must be at the throttle pot or its connections (including earth) to the ECU. Nothing else but a misinterpreted TPS value could convince the ECU to deliver such gobs of fuel at the wrong time. And it seems such an unlikely mode of failure of the ECU itself. I think I know where to go next.
Thank you for reading so far: you'll understand that I am having this enjoyable discussion with myself at the same time as with you all!
And I'll never resent kindly-meant suggestions: I'm more aware of my ability to c@ck things up than anybody else could ever be!
Thanks again
John
If you have changed the throttle pot, I wouldn't rule out the possibility that the replacement is connected wrongly. I never had to change one of these myself on that engine but I seem to recall stories about having to switch leads over to work correctly depending which throttle pot variant was fitted. It doesn't seem very likely, because you say the voltage at the ECU is changing reasonably, but I'm looking for straws for you to clutch at.
Given how long this has been down now, if that was me I'd think about putting some clean, dry, warm plugs in and crank it over with the fuel pump disconnected. If you can crank it over enough to get some manifold depression you may well get slight signs of life just on fuel being sucked out of the injectors, and adding a few blasts of damp start while cranking should see it fire and run for a second or so. That would just give you confidence that there's nothing fundamentally wrong with the oily bits.
Given how long this has been down now, if that was me I'd think about putting some clean, dry, warm plugs in and crank it over with the fuel pump disconnected. If you can crank it over enough to get some manifold depression you may well get slight signs of life just on fuel being sucked out of the injectors, and adding a few blasts of damp start while cranking should see it fire and run for a second or so. That would just give you confidence that there's nothing fundamentally wrong with the oily bits.
Gassing Station | S Series | Top of Page | What's New | My Stuff