ECU - replace/upgrade?
Discussion
adam quantrill said:
Hi Ian,
do we have any documentation on which resistors you have to swap around?
If we can nail this down then it will become much simpler to sort out ECU swaps.
Of course. But I and a few others on here have put a lot of time into figuring this out, not sure we want to give any more information away that may turn up on :cough: other websites... let's just say that matching two ECUs is easy, it's altering an ECU to change its response that's the tricky part. It's kept Mark Adams in cheese scones for long enough...do we have any documentation on which resistors you have to swap around?
If we can nail this down then it will become much simpler to sort out ECU swaps.
I have the 4cu guts out of an early 3.5 range rover. I tried it in the 350i before building a megasquirt into the case and stuffing a 4.6 into it. It ran OK to be honest. I also have a 4cu tweaked by John Eales for a 4.5 rangie upgrade he did around 1989. This has a vernier pot stuck on the side that "controls" the fueling across the range. I'm happy to compare them to any you come across. Knowledge is power! I have the schematics for the ecu as fitted to the 1986 twin plenum vitesse so that may be useful. There are so many old classics using old tech ECU's that are suffering from component failure that there must be a market for something like the open source megasquirt with an equivalent base map. Most of the components can be replaced but some of the custom components are obsolete.
Simon
Simon
ked-up ECUs on Ebay Adam. You probably got all this guff but if not:
For some information regarding components up for adjustment inside the typical SD1/RR 4CU ECU, you may have already seen the first two documents.
R.
For some information regarding components up for adjustment inside the typical SD1/RR 4CU ECU, you may have already seen the first two documents.
- The first is quite well known and seemingly authoritative.
- The second is an original instruction set from one of the well known sources of modified SD1 4CU ECU's.
- The third is some rather large 4CU circuit diagrams.
R.
Edited by honestjohntoo on Sunday 15th July 10:52
I'm up for a bit of the action! But, let's be honest here there's only one reason that all the information we need is not in the public domain; money! People have to make a living and if you have invested a bucket load of time developing a system, or reverse engineering it, you want to reap the benefits. I understand that, but that's a little bit like laying down the gauntlet and saying "go on then" - well it is to me anyway. However, there comes a point where the law of diminishing returns start to kick in and it's not worth perusing the goal unless you have nothing better to do with you time. Modern digital tech is far more flexible, reliable, and not necessarily more expensive, than trying to keep an old, out dated, and degrading box of mainly analogue components running. Arghh, rain my neck in, my day job is taking over!
Keeping the original kit going makes a difference, is more of a challenge, and keeps the car original. Doesn’t really make sense but neither does driving around in an 1960's V8 – other than the sound.
I’ve never tried a cheese scone, are they good?
Keeping the original kit going makes a difference, is more of a challenge, and keeps the car original. Doesn’t really make sense but neither does driving around in an 1960's V8 – other than the sound.
I’ve never tried a cheese scone, are they good?
To give anyone interested an idea of the challenge: firstly each ECU is matched to a particular AFM. The AFM has a ceramic resistor network that includes the carbon track 'swept' by the wiper attached to the flap. Leaving aside the tension on the spring which as we all know is oft-bu99ered with to ill effect, each ceramic resistor network is precision cut to create a particular response curve. This curve is different between the intended applications and even between AFMs used in the same application over the years.
The ECU, then, has a collection of resistors (you can't miss them, they're mounted on PCB standoffs) that are changed to suit the AFM and to give the desired fuel delivery curve of the injectors. This is only part of the story however, because those standoff resistors are each paralelled across another set of resistors; these vary between the application as well.
All of these resistors are 2% tolerance types and for each resistor pair there is a window of values that may be used, wider in some cases and narrower in others. If you know anything about resistors in parallel you will understand how close a value you can create using 2% tolerance components and the fact that they are used illustrates how tight the end value has to be, in each section of the circuit, to give even approximately correct fuelling. And don't forget, all this assumes you have the original AFM to which the ECU was matched AND you understand what each particular part of the circuit is contributing to the overall fuelling plan (some of the internet guesses about the circuitry are wrong).
Pick a random AFM and ECU off Ebay, then fix whatever faults they probably have to explain why they're for sale in the first place, and off you go. Shouldn't take too long... all 4CUs and all AFMs are the same, after all
The ECU, then, has a collection of resistors (you can't miss them, they're mounted on PCB standoffs) that are changed to suit the AFM and to give the desired fuel delivery curve of the injectors. This is only part of the story however, because those standoff resistors are each paralelled across another set of resistors; these vary between the application as well.
All of these resistors are 2% tolerance types and for each resistor pair there is a window of values that may be used, wider in some cases and narrower in others. If you know anything about resistors in parallel you will understand how close a value you can create using 2% tolerance components and the fact that they are used illustrates how tight the end value has to be, in each section of the circuit, to give even approximately correct fuelling. And don't forget, all this assumes you have the original AFM to which the ECU was matched AND you understand what each particular part of the circuit is contributing to the overall fuelling plan (some of the internet guesses about the circuitry are wrong).
Pick a random AFM and ECU off Ebay, then fix whatever faults they probably have to explain why they're for sale in the first place, and off you go. Shouldn't take too long... all 4CUs and all AFMs are the same, after all
Wedg1e said:
Shouldn't take too long... all 4CUs and all AFMs are the same, after all
It sounds like you've already got pretty extensive experience of all things 4CU and have survived to tell the tale.
Some interesting stuff here :-)
http://www.porkers.co.uk/topic.asp?t=751094&f=...
There seems to be a fair bit of info floating around on t'internet regarding the 4CU now. In 2003 when I went down the Megasquirt route things were a little different and detailed information was hard to come by.
It would be great to be able to test and evaluate the ECU independently . I know various companies have developed their own systems for doing this and you can probably buy something off the shelf for many £kkk. It wouldn't be too difficult to build up a system to simulate the resistive inputs supplied by the AFM, air and coolant temp sensors, throttle pot, and generate an ignition pulse as picked up from the coil. We could then accurately measure the pulse width / duty cycle that the ECU is supplying to the injectors. This would at least allow a baseline response of a "good" reference ECU to be established. Armed with that information the effects of each of the various calibration components could be assessed. There’s a fair bit of work involved in building the hardware and writing some software but It would make a nice winter project and could be adapted for later ECU's as well.
Edited by simon_h on Monday 16th July 09:36
Hi all,
This has been an interesting read.
I have been wondering about the ecu and as I am on a remote island it could be quite a drama should it fail.
My options would be a crash course in repairing it or fitting a much more hi tech megasquirt for example. Maybe one such ecu could be fitted while still keeping the afm in situe even if it is not being used so it at least looks original , the dizzy needs to go if you fit coilpacks but it could all be reversible.
What are your thoughts in regards to keeping our wedges original, especially the rare ones?
Is it time to bring these 1960s v8s into the 21st century?
More power?
Better fuel economy?
And of course you can get the software and see what's going on with your power plant, removing all the guess work.
Adam Q, my ecu also has a dial/knob on the side with a blob of something to stop anyone adjusting it, also are the ecu's on the sx's the same as the rest?
It looks like my ecu has been tampered with as it's held together with gaffa tape!???
Cheers
Adam
This has been an interesting read.
I have been wondering about the ecu and as I am on a remote island it could be quite a drama should it fail.
My options would be a crash course in repairing it or fitting a much more hi tech megasquirt for example. Maybe one such ecu could be fitted while still keeping the afm in situe even if it is not being used so it at least looks original , the dizzy needs to go if you fit coilpacks but it could all be reversible.
What are your thoughts in regards to keeping our wedges original, especially the rare ones?
Is it time to bring these 1960s v8s into the 21st century?
More power?
Better fuel economy?
And of course you can get the software and see what's going on with your power plant, removing all the guess work.
Adam Q, my ecu also has a dial/knob on the side with a blob of something to stop anyone adjusting it, also are the ecu's on the sx's the same as the rest?
It looks like my ecu has been tampered with as it's held together with gaffa tape!???
Cheers
Adam
Edited by Oz2 on Monday 16th July 10:35
Wedg1e said:
You're not wrong
I get the feeling you've already been there - is this yours by any chance? http://www.wedgeneering.co.uk/Lucas%20Diagnostics....
simon_h said:
I get the feeling you've already been there - is this yours by any chance? http://www.wedgeneering.co.uk/Lucas%20Diagnostics....
Certainly is.Oz2 said:
My options would be a crash course in repairing it or fitting a much more hi tech megasquirt for example. Maybe one such ecu could be fitted while still keeping the afm in situe even if it is not being used so it at least looks original , the dizzy needs to go if you fit coilpacks but it could all be reversible.
There's quite a lot of effort involved in getting a megasquirt setup installed and properly tuned. However, the end result would be better if correctly implemented. Also, depending on how much DIY you do it can work out quite expensive. As regards originality, my opinion is that provided you keep all the parts you take off and can return the car back to its original state then it's not a problem. That's provided you don't want it to look original under the bonnet while the MS is in situ. If this is the case you could retain the original dizzy with the advance locked out, fit a luminition optical setup to trigger the MS and drive the coil via the MS, and and also leave the AFM in situ so it would look completely original under the bonnet (apart from the luminition amp). You'd lose some of the benefits of using coil packs, the accuracy of a crank based ignition pickup, and have the restriction of the AFM, but it's another option. I did this originally so that I could install the MS in a phased approach. I've also retained the original wiring loom, albeit with a few tweaks, and built the MS into a 4CU case. Mine doesn't look original though!There are plenty of good megasquirt forums around where you can get help and practical advise - http://forums.lr4x4.com has some good information and a dedicated megasquirt forum.
Simon.
Every modification I've seen to a 4CU to make it give more power has usually involved some means of deceiving the thing into thinking the engine is cooler than it really is so it runs richer. Even the Sprintex supercharged Rangey ECU I have does it in a similar way, albeit using engine RPM as a modifier (it has a fequency-voltage converter IC onboard).
I don't think there would be much to be gained in trying to squeeze a couple of extra horses out of your 350i/ 390SE by seeking one closer resistor value to those already fitted: it's clear that Lucas did optimise each AFM/ECU combo rather than just banging in random resistors and hoping for the best. It's proven that altering the fuel rail pressure gives some gains which may be of more benefit at full throttle. From my experiments I can say that upping the pressure on my car, whilst giving you the sort of surge normally reserved for being thrown off a roof, makes the damn thing most unpleasant to drive elsewhere than flat-out. I prefer my cars more tractable than that.
I think the best that could be reasonably achieved by the DIYer is to match one 4CU to another in order to have a serviceable spare (as in the case of our Antipodean mate here)... or maybe retuning a 350i ECU to 390SE spec to go with the necessary engine mods. - though arguably if you're going to that much trouble with the engine you'd probably opt for a modern injection/ management system.
One might postulate that by creating a spreadsheet of resistor values taken from a variety of 4CUs, you could equate certain ranges of values with a certain spec. of ECU. That's what I thought (as did others, both on PH and elsewhere). Alas no: again the vagaries of the AFM resistor network come into play making it impossible to see a pattern - at least, not from the more than three dozen ECUs I have data on, ranging from humble RangeRover through SD1 (inc. the twin-plenum) to 390SE and several 'tuned' variants with twiddly pots mounted on the side or remotely.
Merely putting a 390SE-spec ECU on a lower-spec engine does not give gains like you may expect, as Hotty and I proved at last year's 'Fest - we tried three different ECUs on his 400SE and whilst one or two gave better response at certain revs, the overall effect was not to improve the performance of the car. All of which goes back to the necessity of matching AFM and ECU. After all, my car made 256bhp on the rollers, you might have thought his would so something similar using my ECU... er, no.
I don't think there would be much to be gained in trying to squeeze a couple of extra horses out of your 350i/ 390SE by seeking one closer resistor value to those already fitted: it's clear that Lucas did optimise each AFM/ECU combo rather than just banging in random resistors and hoping for the best. It's proven that altering the fuel rail pressure gives some gains which may be of more benefit at full throttle. From my experiments I can say that upping the pressure on my car, whilst giving you the sort of surge normally reserved for being thrown off a roof, makes the damn thing most unpleasant to drive elsewhere than flat-out. I prefer my cars more tractable than that.
I think the best that could be reasonably achieved by the DIYer is to match one 4CU to another in order to have a serviceable spare (as in the case of our Antipodean mate here)... or maybe retuning a 350i ECU to 390SE spec to go with the necessary engine mods. - though arguably if you're going to that much trouble with the engine you'd probably opt for a modern injection/ management system.
One might postulate that by creating a spreadsheet of resistor values taken from a variety of 4CUs, you could equate certain ranges of values with a certain spec. of ECU. That's what I thought (as did others, both on PH and elsewhere). Alas no: again the vagaries of the AFM resistor network come into play making it impossible to see a pattern - at least, not from the more than three dozen ECUs I have data on, ranging from humble RangeRover through SD1 (inc. the twin-plenum) to 390SE and several 'tuned' variants with twiddly pots mounted on the side or remotely.
Merely putting a 390SE-spec ECU on a lower-spec engine does not give gains like you may expect, as Hotty and I proved at last year's 'Fest - we tried three different ECUs on his 400SE and whilst one or two gave better response at certain revs, the overall effect was not to improve the performance of the car. All of which goes back to the necessity of matching AFM and ECU. After all, my car made 256bhp on the rollers, you might have thought his would so something similar using my ECU... er, no.
You’ve obviously invested a great deal of time in optimizing the 4CU with mixed results. I guess this is one of the problems of what is really a mixed analogue / Digital system. Do you think there would be any mileage in a slightly different approach. Based on the excellent document by Gene Thompson U102 is generating a 7-bit representation of load taking the AFM and RPM inputs into account. If we were to take those 7-bit values and manipulate them could we not alter the fuelling curve? There are a couple of small inexpensive single board dev systems with plenty of I/O which could accomplish this. They also have USB interfaces so you could simply plug a laptop in and tweak away. I know what you’re going to say – you’ve tried this and it doesn’t work either 
You could also add additional features if required since once you've got a microprocessor in there the sky is the limit.
You could also add additional features if required since once you've got a microprocessor in there the sky is the limit.simon_h said:
You’ve obviously invested a great deal of time in optimizing the 4CU with mixed results. I guess this is one of the problems of what is really a mixed analogue / Digital system. Do you think there would be any mileage in a slightly different approach. Based on the excellent document by Gene Thompson U102 is generating a 7-bit representation of load taking the AFM and RPM inputs into account. If we were to take those 7-bit values and manipulate them could we not alter the fuelling curve? There are a couple of small inexpensive single board dev systems with plenty of I/O which could accomplish this. They also have USB interfaces so you could simply plug a laptop in and tweak away. I know what you’re going to say – you’ve tried this and it doesn’t work either You could also add additional features if required since once you've got a microprocessor in there the sky is the limit.
No I haven't tried it but of course it's not improbable that it could work, it's just a case of whether the results are worth the effort: why go to those lengths when Megasquirt et al make it easier? In fact easier still for some people is just to throw carbs in there. You could also add additional features if required since once you've got a microprocessor in there the sky is the limit.I don't know whether Gene recorded the digital data in a fashion that relates to the actual analogue inputs and even if he did, without unlimited access to a rolling road how would you work out what to tweak and when? Same problem as the trial-and-error method really: my car was set-up by the good old seat of pants technique and quite by chance the dyno graph shows a maximally-flat air:fuel ratio curve (compared to some others that have been posted on here and elsewhere). If the hypothetical ECU under test was not creating those results then the record of digital vs. analogue input wouldn't help much without a plot of the engine's power - but then a few hours of dyno twiddling would soon outweigh the cost of Megasquirt etc.
I'm not saying it can't be done, nor that it shouldn't be played with but from a DIY point of view it's a steep mountain to climb compared to plugging in a laptop and downloading some preset maps... and anyone who was prepared to spend the time tweaking and twiddling for other people would, as you sad earlier, want to be making money out of it.
I suspect a properly set-up hotwire system on a hot-spec. engine would be a lot of fun; sadly I think most 400SEs are not hot-spec. engines so the potential is missed by most owners.
You have some valid points there. Commercially it doesn’t make any sense as due to the small volumes the development costs would probably make it uncompetitive when compared to a Megasquirt or similar. Carbs! I had a nice set of quad downdraught Dellorto’s on an SD1 Vitesse I once owned, ah what a sound. You couldn’t hear the exhaust for the induction roar and other drivers would swerve out of your way when you started an overtake because they thought you would be joining them in their car soon. Again, carbs would diminish the originality of the vehicle and the economy would not be improved.
For the average DIY’er the Megasquirt route is not a complete bed of roses and demands a certain degree of knowledge, time, and determination. I don’t see this project as a like for like comparison to ripping out the 4CU and putting in a MS or other aftermarket ECU, or even a 14CUX. If you want a fully mapped system with Ignition control and a whole host of other features than there’s no point in messing around with the old 4CU.
I’m just wondering how many people are out there who want to keep everything pretty much original but be able to optimise what they already have in order to get good economy and decent top end power. If this design works, I’d look at having a remote control unit or a laptop running some simple software that would allow you to adjust the fueling in realtime at any RPM / AFM input level while on the rollers or driving down the road. Of course, you’d have to have an assistant with you to operate the controls while driving down the road. This is far better than swapping resistors around, adjusting springs in the AFM, or fooling the ECU into thinking it’s a cold December’s morning when it’s the middle of summer. Having said that, with the summer we’re having the temperature differential would not be that great.
I'll put something together and see if it works anyway just out of interest.
Again, carbs would diminish the originality of the vehicle and the economy would not be improved.For the average DIY’er the Megasquirt route is not a complete bed of roses and demands a certain degree of knowledge, time, and determination. I don’t see this project as a like for like comparison to ripping out the 4CU and putting in a MS or other aftermarket ECU, or even a 14CUX. If you want a fully mapped system with Ignition control and a whole host of other features than there’s no point in messing around with the old 4CU.
I’m just wondering how many people are out there who want to keep everything pretty much original but be able to optimise what they already have in order to get good economy and decent top end power. If this design works, I’d look at having a remote control unit or a laptop running some simple software that would allow you to adjust the fueling in realtime at any RPM / AFM input level while on the rollers or driving down the road. Of course, you’d have to have an assistant with you to operate the controls while driving down the road.
This is far better than swapping resistors around, adjusting springs in the AFM, or fooling the ECU into thinking it’s a cold December’s morning when it’s the middle of summer. Having said that, with the summer we’re having the temperature differential would not be that great.I'll put something together and see if it works anyway just out of interest.
Edited by simon_h on Wednesday 18th July 00:32
Gassing Station | Wedges | Top of Page | What's New | My Stuff