Classis fuel injection
#1 Classis fuel injection
Just thought I would show the 12 injector system I have fitted to my S3 v12 .The system comes from classic fuel injection and fits on XJS manifolds,quite a few other parts had to come from XJS .Classic fuel injection normally supply a 4 injector system for the v12 leaving carbs on as throttle bodies, I decided I wanted a full 12 injector set up so they designed me one here are some photos.The system is fully programable vac and centrifugal advance are done away with the ECU adjusts the timing and fuelling ,there is a self learn facility it takes readings from the lambda sensor and adjusts accordingly .The engine runs lovely and picks up a lot quicker ,and one very good thing it does away with the SU fuel pump in the boot no more petrol smell in boot ,ive had petrol smell in boot for 10 years could never find a leak , it now gone
Rob 1972 s3 roadster
Aston Martin DB9 Volante
Aston Martin DB9 Volante
Link: | |
BBcode: | |
HTML: | |
Hide post links |
#2 Re: Classis fuel injection
Sounds really good. Did they put the fuel pump in the tank? Did they provide any documentation, e.g. wiring diag?
Regards,
ColinL
'72 OTS manual V12
ColinL
'72 OTS manual V12
Link: | |
BBcode: | |
HTML: | |
Hide post links |
#3 Re: Classis fuel injection
Hi just bought the injection kit from classic fuel injection and i worked out what i needed to make it fit as they had not done a jag v12 12 injector set up .As i said they normally use a 4 injector set up .I dont like that set up as u still have the long inlet tract over the cam covers .I have fitted the high pressure fuel pump and filter just in front of the spare just below the boards ,i am very happy with the results .I looked at fitting the full XJS set up but it is far too complicated and using old technology and old components i think is going to be unreliable ,plus its would not be programable
Rob 1972 s3 roadster
Aston Martin DB9 Volante
Aston Martin DB9 Volante
Link: | |
BBcode: | |
HTML: | |
Hide post links |
#4 Re: Classis fuel injection
Those are all good reasons why you shouldn't go with an old XJS setup.
An inline pump inside the fuel tank from a Land Rover works well.
If using Jaguar manifolds, the ones to go for are the 6litre manifolds as these have the injectors bolted into the rails, so there are no rubber fuel hoses above every injector.
Megasquirt has come a long way and can give you fully programmable sequential fuel, coilpack ignition and control and log all of the peripherals using a laptop. You can even look at the result in real time in the car and diagnose faults or tune it that way.
kind regards
Marek
An inline pump inside the fuel tank from a Land Rover works well.
If using Jaguar manifolds, the ones to go for are the 6litre manifolds as these have the injectors bolted into the rails, so there are no rubber fuel hoses above every injector.
Megasquirt has come a long way and can give you fully programmable sequential fuel, coilpack ignition and control and log all of the peripherals using a laptop. You can even look at the result in real time in the car and diagnose faults or tune it that way.
kind regards
Marek
Link: | |
BBcode: | |
HTML: | |
Hide post links |
#5 Re: Classis fuel injection
Hi i read up on the megasquirt systems before deciding on classic fuel injection system using the maxx ECU . The thing that made me go for the CFI system was ease of fitting and backup from them ,plus it does not need a crank sensor and wheel to complicate matters .The fuel pump in the boot is very quiet i cannot hear it at all. The system came complete with all sensors ,ECU ,wiring loom, wideband lambda sensor and boss to weld to exhaust, air control valve and controller, fuel injectors, fuel pump and filter, fuel pipe .I had a pertronix ignitor fitted to my distributor already so that saved me money. It came with a start up fuel and ignition map then i simply put the system into self learn, as i drove the car the system adjusted the fueling to bring the lambda as near to 1.0 as possible. I had to tweak the ignition timing on the graph to level the tick over and advanced it a tad at various points to help acceleration. I have found the v12 likes a lot of advance to make it perform better .Its nice to have the idle adjustable on the laptop plus the cold fast idle can be set so it runs just right
Rob 1972 s3 roadster
Aston Martin DB9 Volante
Aston Martin DB9 Volante
Link: | |
BBcode: | |
HTML: | |
Hide post links |
#6 Re: Classis fuel injection
Another thing i forgot to mention i put the xjs breather oil catch and pipes on ,they are far better at making the engine breathe freely. The oil catch unit bolted to the front of the head needs to be cut down and re welded to make it clear the coolant pipes
Rob 1972 s3 roadster
Aston Martin DB9 Volante
Aston Martin DB9 Volante
Link: | |
BBcode: | |
HTML: | |
Hide post links |
#7 Re: Classis fuel injection
Funny. The PO of my XJS removed the injection completely and fitted 4 2" SUs.
Andrew.
881824, 1E21538. 889457. 1961 4.3l Mk2. 1975 XJS. 1962 MGB
http://www.projectetype.com/index.php/the-blog.html
Adelaide, Australia
881824, 1E21538. 889457. 1961 4.3l Mk2. 1975 XJS. 1962 MGB
http://www.projectetype.com/index.php/the-blog.html
Adelaide, Australia
Link: | |
BBcode: | |
HTML: | |
Hide post links |
#8 Re: Classis fuel injection
This appears to be a four cylinder system and a very coarse one at that.
All that is good for is to tell the ECU that the engine is turning and so it is safe to inject fuel, plus some very coarse rpm information. That's because a single sensor in the distributor tells you where in the engine cycle you are once per 720' cycle, whereas the trigger wheel (assuming a 36 minus 1 wheel) tells you that 72 times as often. That means 72x more accurate basic fueling and acceleration enrichment. You're going to want a 72 tooth wheel inside there just for starters. Even then you'll still never know where the crank is or how much the settings are drifting over time.
The classicfuelinjection setup is batch fired as per the original Lucas xjs (i.e. it's a four cylinder setup) and for £2300 list price you could have probably had an all singing all dancing sequential fuel and spark plus all ancillaries and full datalogging.
Don't get me wrong - any fuel injection will be better than a carburettored car, but this system is possibly the least desirable one possible.
Feel free to post your ignition and fuel maps. It may save you a lot of heartache if someone notices a manifest error and save your pistons when the engine is under heavy load.
kind regards
Marek
In that case, how does it know about timing for this ignition map you talk about? If it is coming from a sensor inside the distributor, or something derived from that, then this is a recipe for disaster, as this is a pretty inaccurate mechanism and a self fulfillingly correct predictor of what is correct as far as timing goes.
All that is good for is to tell the ECU that the engine is turning and so it is safe to inject fuel, plus some very coarse rpm information. That's because a single sensor in the distributor tells you where in the engine cycle you are once per 720' cycle, whereas the trigger wheel (assuming a 36 minus 1 wheel) tells you that 72 times as often. That means 72x more accurate basic fueling and acceleration enrichment. You're going to want a 72 tooth wheel inside there just for starters. Even then you'll still never know where the crank is or how much the settings are drifting over time.
The classicfuelinjection setup is batch fired as per the original Lucas xjs (i.e. it's a four cylinder setup) and for £2300 list price you could have probably had an all singing all dancing sequential fuel and spark plus all ancillaries and full datalogging.
Don't get me wrong - any fuel injection will be better than a carburettored car, but this system is possibly the least desirable one possible.
Feel free to post your ignition and fuel maps. It may save you a lot of heartache if someone notices a manifest error and save your pistons when the engine is under heavy load.
kind regards
Marek
Link: | |
BBcode: | |
HTML: | |
Hide post links |
#9 Re: Classis fuel injection
Hi Marek
I don't understand what you are on about with the distributor sensor the pertronix ignitor gives 12 pulses per revolution of the distributor. The distributor centrifugal advance is welded solid the vac advance is disconnected, so as far as i see its no less accurate than a crank sensor and ring .The timing is set on the graph that works with revs, map sensor, lambda sensor, intake air temp sensor, plus coolant temp sensor. As far as melting pistons and such lets face it the jag v12 is a very low performance engine u would need to run it on methanol to melt anything .Ive just this morning had it on the gas analyser it now ticking over at lambda .993 and CO at 2.10 at 2000rpm again lambda .987 and CO at 1.86. When i was running on Stromberg's it would not run under 6% CO when on SU carbs it needed to be at least 4.5% the lambda was never anyway near 1.0 a vast improvement. I am very happy with these figures
Regards Rob
I don't understand what you are on about with the distributor sensor the pertronix ignitor gives 12 pulses per revolution of the distributor. The distributor centrifugal advance is welded solid the vac advance is disconnected, so as far as i see its no less accurate than a crank sensor and ring .The timing is set on the graph that works with revs, map sensor, lambda sensor, intake air temp sensor, plus coolant temp sensor. As far as melting pistons and such lets face it the jag v12 is a very low performance engine u would need to run it on methanol to melt anything .Ive just this morning had it on the gas analyser it now ticking over at lambda .993 and CO at 2.10 at 2000rpm again lambda .987 and CO at 1.86. When i was running on Stromberg's it would not run under 6% CO when on SU carbs it needed to be at least 4.5% the lambda was never anyway near 1.0 a vast improvement. I am very happy with these figures
Regards Rob
Rob 1972 s3 roadster
Aston Martin DB9 Volante
Aston Martin DB9 Volante
Link: | |
BBcode: | |
HTML: | |
Hide post links |
#10 Re: Classis fuel injection
Dear Rob,
1/ It's a big improvement on carburettors. That's great.
2/ The v12 distributor is not known for its accuracy or consistency. I would only feel comfortable if the software settings were confirmed by a timing light. Advancing the timing under low load is relatively safe but knowing how far timing can be advanced under heavy load is not straightforward.
Regarding the bit you say you don't understand:-
A 36 minus 1 wheel on the crank tells the ECU where the crank is 72 times per engine cycle.
The Pertronix tells the ECU where the jackshaft between the A and B banks is 12 times per rotation of the distributor.
The bigger the trigger wheel and the more teeth it has, the better.
The fact that a trigger wheel is located rigidly on the crank rather than somewhere else where it is subject to chain accelerating and slowing down is also "better"
So the 36-1 wheel is 6 times more accurate about crank location than a Pertronix is and it's measuring the thing you want to know, not a proxy for it which changes as the chain accelerates and whips about.
Essentially, you have less information about crank position and it's less accurate information as to whether the crank is speeding up or slowing down and the baseline information you do have is subject to the chain and sprocket wear. You need to reconfirm your timing as your chain and sprockets age as your setup ages whereas a rigidly crank mounted trigger wheel never changes with respect to where the crank is.
It's way better than a carburettor and vacuum-distributor setup, but for want of not fitting a trigger wheel....
FWIW, my answer to not wanting a trigger wheel was to CNC cut it into the pulley which sits on the damper.
kind regards
Marek
1/ It's a big improvement on carburettors. That's great.
2/ The v12 distributor is not known for its accuracy or consistency. I would only feel comfortable if the software settings were confirmed by a timing light. Advancing the timing under low load is relatively safe but knowing how far timing can be advanced under heavy load is not straightforward.
Regarding the bit you say you don't understand:-
A 36 minus 1 wheel on the crank tells the ECU where the crank is 72 times per engine cycle.
The Pertronix tells the ECU where the jackshaft between the A and B banks is 12 times per rotation of the distributor.
The bigger the trigger wheel and the more teeth it has, the better.
The fact that a trigger wheel is located rigidly on the crank rather than somewhere else where it is subject to chain accelerating and slowing down is also "better"
So the 36-1 wheel is 6 times more accurate about crank location than a Pertronix is and it's measuring the thing you want to know, not a proxy for it which changes as the chain accelerates and whips about.
Essentially, you have less information about crank position and it's less accurate information as to whether the crank is speeding up or slowing down and the baseline information you do have is subject to the chain and sprocket wear. You need to reconfirm your timing as your chain and sprockets age as your setup ages whereas a rigidly crank mounted trigger wheel never changes with respect to where the crank is.
It's way better than a carburettor and vacuum-distributor setup, but for want of not fitting a trigger wheel....
FWIW, my answer to not wanting a trigger wheel was to CNC cut it into the pulley which sits on the damper.
kind regards
Marek
Link: | |
BBcode: | |
HTML: | |
Hide post links |
#11 Re: Classis fuel injection
Hi marek
I understand what your saying about accuracy of timing but lets face it we are not working in formula 1, its a 51 year old car . I had to set the timing at 12deg running through the ECU after locking the advance in the distributor before removal of the carbs and fitting the rest of the system and the variation of the timing when revving was very small ,I rebuilt the engine 4 years ago and all chains and tensioners where renewed. Whatever system we fit to our cars we cannot get round the fact that it is old technology in the engine and a couple of degrees here and there don't make much difference at all to the day to day running. I could understand if it was on the race track but as a classic car its perfect for me now. Thank you very much for your input
Best Regards Rob
I understand what your saying about accuracy of timing but lets face it we are not working in formula 1, its a 51 year old car . I had to set the timing at 12deg running through the ECU after locking the advance in the distributor before removal of the carbs and fitting the rest of the system and the variation of the timing when revving was very small ,I rebuilt the engine 4 years ago and all chains and tensioners where renewed. Whatever system we fit to our cars we cannot get round the fact that it is old technology in the engine and a couple of degrees here and there don't make much difference at all to the day to day running. I could understand if it was on the race track but as a classic car its perfect for me now. Thank you very much for your input
Best Regards Rob
Rob 1972 s3 roadster
Aston Martin DB9 Volante
Aston Martin DB9 Volante
Link: | |
BBcode: | |
HTML: | |
Hide post links |
#12 Re: Classis fuel injection
FWIW isn't the distributor driven by a gear directly off the crankshaft, so chain & sprocket wear should have no influence on ignition timing, or is that not the case on a V12?
Randall Botha
'64 3.8 fhc & '51 Mk 7
'64 3.8 fhc & '51 Mk 7
Link: | |
BBcode: | |
HTML: | |
Hide post links |
#13 Re: Classis fuel injection
The distributor is driven from the jackshaft. That sits in between the A and B banks at the top of the "V" and is driven by the chain.
Rob, feel free to post up your ignition and fuel maps. Others may improve them for you for free as there are plenty of fuel injected v12s out there in cyberspace.
kind regards
Marek
Rob, feel free to post up your ignition and fuel maps. Others may improve them for you for free as there are plenty of fuel injected v12s out there in cyberspace.
kind regards
Marek
Link: | |
BBcode: | |
HTML: | |
Hide post links |
#14 Re: Classis fuel injection
"Ive just this morning had it on the gas analyser it now ticking over at lambda .993 and CO at 2.10 at 2000rpm again lambda .987 and CO at 1.86. When i was running on Stromberg's it would not run under 6% CO when on SU carbs it needed to be at least 4.5% the lambda was never anyway near 1.0 a vast improvement. I am very happy with these figures."
My Ev12 with a 6,0L motor form a 1993 X81 is running with 4 stromberg carbs with Lambda 1,1 to 1,2 at part throttle. The gas consumption is low (20mpg), the car pulls strong and the drivability is good. I am happy with these figures .
The needles are 'taylor made' by me. I have a lambda wideband system mounted in the car.
Regards Wolfgang Gatza
My Ev12 with a 6,0L motor form a 1993 X81 is running with 4 stromberg carbs with Lambda 1,1 to 1,2 at part throttle. The gas consumption is low (20mpg), the car pulls strong and the drivability is good. I am happy with these figures .
The needles are 'taylor made' by me. I have a lambda wideband system mounted in the car.
Regards Wolfgang Gatza
Link: | |
BBcode: | |
HTML: | |
Hide post links |
#15 Re: Classis fuel injection
Just to add further to what I think might be the most interesting question on here- whether to go EFI and enjoy the better reliability, smoothness and fuel consumption or stay traditional and loyal to the history and stay with Strombergs.
I have decided to rebuild my engine with EFI, and plan to use mid-eighties plenum chambers and throttle bodies from an XJS but with modern injectors with adapters, coil packs and Emerald ECU. Not done it yet but after much research this is my way forward. I’m aware the 6.0 plenums avoid the rubber injection hoses but these are much rarer to find second hand.
I would like the engine to look as close as practicable to what Jaguar envisaged for the S3 launch in 1971, albeit that a toothed crank wheel and coil packs in the vee won’t quite hit it. But the toothed wheel (as has already been said) is the best way for the ECU to know engine timing and position, and the coil packs are driven directly by the ECU and so do away with all that unreliable distributor, two coils and/or the Opus amplifier or it’s replacement.
Bob
I have decided to rebuild my engine with EFI, and plan to use mid-eighties plenum chambers and throttle bodies from an XJS but with modern injectors with adapters, coil packs and Emerald ECU. Not done it yet but after much research this is my way forward. I’m aware the 6.0 plenums avoid the rubber injection hoses but these are much rarer to find second hand.
I would like the engine to look as close as practicable to what Jaguar envisaged for the S3 launch in 1971, albeit that a toothed crank wheel and coil packs in the vee won’t quite hit it. But the toothed wheel (as has already been said) is the best way for the ECU to know engine timing and position, and the coil packs are driven directly by the ECU and so do away with all that unreliable distributor, two coils and/or the Opus amplifier or it’s replacement.
Bob
1973 E Type OTS V12
Lotus Esprit V8
Lotus Esprit V8
Link: | |
BBcode: | |
HTML: | |
Hide post links |
#16 Re: Classis fuel injection
My view if u own a trailer queen for driving short distances and a pure car then stay with strombergs .
But a car like i have that is a nice driver that i use as a virtual daily driver makes sense to convert over to fuel injection and ECU controlled ignition.
We are not looking for out and out performance but reliability and better driving experience and fuel economy that was the reason i went down the FI route
Again bob has brought up the unreliable distributor statement . on my conversion all the distributor has in it is a petronix igniter no mechanical or vac advance mechanism so all that is to go wrong in there is the ignitor i have always carried a spare under back seat just in case
When checking timing with a strobe the lash in the distributor is next to nothing .
Anyway lets face it a 5.3 v12 producing less than 300 bhp even a few degrees wont affect performance a lawn mower produces more power pound for pound
But a car like i have that is a nice driver that i use as a virtual daily driver makes sense to convert over to fuel injection and ECU controlled ignition.
We are not looking for out and out performance but reliability and better driving experience and fuel economy that was the reason i went down the FI route
Again bob has brought up the unreliable distributor statement . on my conversion all the distributor has in it is a petronix igniter no mechanical or vac advance mechanism so all that is to go wrong in there is the ignitor i have always carried a spare under back seat just in case
When checking timing with a strobe the lash in the distributor is next to nothing .
Anyway lets face it a 5.3 v12 producing less than 300 bhp even a few degrees wont affect performance a lawn mower produces more power pound for pound
Rob 1972 s3 roadster
Aston Martin DB9 Volante
Aston Martin DB9 Volante
Link: | |
BBcode: | |
HTML: | |
Hide post links |
#17 Re: Classis fuel injection
Perhaps efi or "carbs out of loyalty" may be question for S2, weren’t S3’s slated to be efi till Brico pulled out? Maybe the question should be whether to finish the job, or to struggle on with the sad, messy and temporary compromise that Jaguar was forced to make?
Imo the comments in this post are not so much about whether or not to fuel inject, they are more about which type of fuel injection; batch or sequential? Batch injection does not require a cam position sensor, sequential injection does …
Sequential is the later tech and in theory enables better across-the-range efficiencies. However, with modern adaptive ECU’s, any such benefit is probably not significant …
I’m in the middle of an upgrade to full sequential efi (and ignition) among other things …
viewtopic.php?f=9&t=18608
I’m using a 36-1 trigger wheel in place of the distributor. I had a bit of backlash between the distributor and jack-shaft that I got rid of by machining the valley cover on a microscopic angle.
I’m hoping that chain whip or flex would cause significant error, three reasons:
It’s a chain, that with Jaguar is very well guided/constrained.
2ndly there is antidotal evidence that chain drive is not an issue in practice, for very many years Jaguar HE distributors have been very successfully modified/sold/used as triggers for efi systems, E.g.: https://www.efihardware.com/products/20 ... uar-HE-V12
With this example the existing reluctor sensor serves as the speed and defacto crank position sensor (12 point wheel on the cam equivalent to 6 tooth wheel on the crank), cam position is provided by the added HAL sensor and single tooth wheel. Net result is a bolt on solution that is very flexible, can be used with single (oem appearance) or multiple coils. I actually bought one of theses before getting the idea of using a missing tooth wheel for an even more compact solution, albeit one that requires multiple coils. So its currently gathering dust, if anyone wants to trial it...
3rd reason depends on how the crank pos’n sensor signal is used, modern ecu’s calculate it, effectively interpolating and extrapolating so that any jitter is effectively smoothed out?
I'm imagining that, if chain drive does turn out to be an issue, it will be with sequential injection (mine), for batch injection (Rob's) there should be no issue whatsoever?
Imo the comments in this post are not so much about whether or not to fuel inject, they are more about which type of fuel injection; batch or sequential? Batch injection does not require a cam position sensor, sequential injection does …
Sequential is the later tech and in theory enables better across-the-range efficiencies. However, with modern adaptive ECU’s, any such benefit is probably not significant …
I’m in the middle of an upgrade to full sequential efi (and ignition) among other things …
viewtopic.php?f=9&t=18608
I’m using a 36-1 trigger wheel in place of the distributor. I had a bit of backlash between the distributor and jack-shaft that I got rid of by machining the valley cover on a microscopic angle.
I’m hoping that chain whip or flex would cause significant error, three reasons:
It’s a chain, that with Jaguar is very well guided/constrained.
2ndly there is antidotal evidence that chain drive is not an issue in practice, for very many years Jaguar HE distributors have been very successfully modified/sold/used as triggers for efi systems, E.g.: https://www.efihardware.com/products/20 ... uar-HE-V12
With this example the existing reluctor sensor serves as the speed and defacto crank position sensor (12 point wheel on the cam equivalent to 6 tooth wheel on the crank), cam position is provided by the added HAL sensor and single tooth wheel. Net result is a bolt on solution that is very flexible, can be used with single (oem appearance) or multiple coils. I actually bought one of theses before getting the idea of using a missing tooth wheel for an even more compact solution, albeit one that requires multiple coils. So its currently gathering dust, if anyone wants to trial it...
3rd reason depends on how the crank pos’n sensor signal is used, modern ecu’s calculate it, effectively interpolating and extrapolating so that any jitter is effectively smoothed out?
I'm imagining that, if chain drive does turn out to be an issue, it will be with sequential injection (mine), for batch injection (Rob's) there should be no issue whatsoever?
Regards,
ColinL
'72 OTS manual V12
ColinL
'72 OTS manual V12
Link: | |
BBcode: | |
HTML: | |
Hide post links |