Cam sensor disc on a TVR Griffith Rover V8
Discussion
I think I need to install a camsync sensor on the TVR so I can understand some of the strange behaviour at idle. The idle problems are interesting enough for a thread of their own, so I will do that later...
I have drilled the boss out in the "intermediate" Serpentine cover so that it will take a standard Hall sensor, but since I don't want to part with the duplex timing chain, I haven't got anything from the Rover parts bin to reuse.
I have drawn up what I think I need to make, but wondered if anyone had already got a solution on the market. I need one single pulse (+ve or -ve doesn't matter) from the cam sensor between 70 and 80 crank degrees before TDC of cylinder 1.
I would prefer to not take the car apart twice, so the loan of an appropriate timing cover for measurement purposes would be useful if I do need to get something made. Using a modified dizzy isn't an option any more - there's no room.
All help greatly appreciated!
I have drilled the boss out in the "intermediate" Serpentine cover so that it will take a standard Hall sensor, but since I don't want to part with the duplex timing chain, I haven't got anything from the Rover parts bin to reuse.
I have drawn up what I think I need to make, but wondered if anyone had already got a solution on the market. I need one single pulse (+ve or -ve doesn't matter) from the cam sensor between 70 and 80 crank degrees before TDC of cylinder 1.
I would prefer to not take the car apart twice, so the loan of an appropriate timing cover for measurement purposes would be useful if I do need to get something made. Using a modified dizzy isn't an option any more - there's no room.
All help greatly appreciated!
dnb said:
I think I need to install a camsync sensor on the TVR so I can understand some of the strange behaviour at idle. The idle problems are interesting enough for a thread of their own, so I will do that later...
I'm intrigued. If you were running sequential you must already have a cam phase sensor - if not, what sort of problem is helped by knowing the cam phase?It does use sequential injection, but it's not phased to the cam at the moment. (Meaning there are 2 opportunities per cycle to inject, but a single random one is "selected") Ignition is fairly obviously wasted spark.
If I have cam phase information, then the engine always syncs up to the correct way, and I can investigate when in the cycle the best time to inject fuel actually is, and knowing this should help stop the random idle misfire problem that seems to exist.
If I have cam phase information, then the engine always syncs up to the correct way, and I can investigate when in the cycle the best time to inject fuel actually is, and knowing this should help stop the random idle misfire problem that seems to exist.
A random idle missfire could be anything. I'd say you're barking up the wrong tree by insisting on going sequential.
Any pics or suggestions of what you had in mind for that phase sensor ?
Even if the dizzy isnt in the hole...Im sure you could knock something up to stick a sensor in there, and read off a tooth affixed to the old dizzy drive ??
Any pics or suggestions of what you had in mind for that phase sensor ?
Even if the dizzy isnt in the hole...Im sure you could knock something up to stick a sensor in there, and read off a tooth affixed to the old dizzy drive ??
dnb said:
It does use sequential injection, but it's not phased to the cam at the moment. (Meaning there are 2 opportunities per cycle to inject, but a single random one is "selected") Ignition is fairly obviously wasted spark.
If I have cam phase information, then the engine always syncs up to the correct way, and I can investigate when in the cycle the best time to inject fuel actually is, and knowing this should help stop the random idle misfire problem that seems to exist.
I'm puzzled as to how any system can be sequential or even pseudo sequential without a cam sensor and I very much doubt if opportunities to inject are selected at random however it's set up. Why not just set it to simultaneous which is how it ought to be working without a cam sensor anyway?If I have cam phase information, then the engine always syncs up to the correct way, and I can investigate when in the cycle the best time to inject fuel actually is, and knowing this should help stop the random idle misfire problem that seems to exist.
As Stevie says I doubt this is the cause of your misfire problem. I've not found engines to be very much bothered how you inject fuel into them as long as the right amount is there when the inlet valve finally opens.
Pumaracing said:
dnb said:
It does use sequential injection, but it's not phased to the cam at the moment. (Meaning there are 2 opportunities per cycle to inject, but a single random one is "selected") Ignition is fairly obviously wasted spark.
If I have cam phase information, then the engine always syncs up to the correct way, and I can investigate when in the cycle the best time to inject fuel actually is, and knowing this should help stop the random idle misfire problem that seems to exist.
I'm puzzled as to how any system can be sequential or even pseudo sequential without a cam sensor and I very much doubt if opportunities to inject are selected at random however it's set up. Why not just set it to simultaneous which is how it ought to be working without a cam sensor anyway?If I have cam phase information, then the engine always syncs up to the correct way, and I can investigate when in the cycle the best time to inject fuel actually is, and knowing this should help stop the random idle misfire problem that seems to exist.
As Stevie says I doubt this is the cause of your misfire problem. I've not found engines to be very much bothered how you inject fuel into them as long as the right amount is there when the inlet valve finally opens.
Big cams can also have a detrimental effect for similar reasons ( combined with big injectors....as I found out myself with 750cc RC injectors once on a build. But that engine wouldnt even start ffs until I put the 600cc Siemens back in )
StevieTurbo:
No matter what I do, there is no room for any sensors where the dizzy used to be. That area is full of coils. I have eliminated all the other causes of misfire that I can think of - the plugs and leads are good, trigger is good, driver transistors are good, injectors all work etc... The problem only occurs at idle - it does not seem to occur when there is any load on the engine, even if the RPM is similar (within reason) to idle.
I suspect camsync will not solve this problem completely, but it will help me eliminate another cause and it should let me investigate other things too. The injection is already sequential, and moving the injection events with respect to the inlet valve openings does seem to lessen the misfire problem, but never eliminate it. With camsync, I will be able to get better control of injector timing (there seems to be an ECU limitation about not overlapping injection and and ignition events on the same cylinder - don't ask me why - wasted spark gets in the way of this) and it will let me start to investigate individual cylinder fuelling trims - I suspect this is more important for making idle smoother than the timing.
I was thinking of a simple 1mm thick disc mounted on the cam with a single hole cut in it, lining up with the standard cam sensor position used in the "Thor" style RV8 engines. The ECU would see a single pulse every 2 crank rotations which defines the phase of the engine, leaving the crank sensor to give accurate position data.
Puma Racing:
The "random" phase is simply due to having a 36-1 trigger on the crank and deciding crank position from the first time a missing tooth event is encountered. The ECU knows where the crank is, but not the cam, but this doen't really matter because as you yourself say engines aren't usually all that bothered when the fuel is injected... It does not mean that the injection event happens at an entirely random and unpredictable time.
You don't need cam trigger to be sequential - all it means is that it fires one injector at a time. There is a 50/50 chance as to where the cam is when the engine starts, which can be considered as essentially a random variable. In position A, then the injection events are all 10 degrees before the inlet valve starts to open. In position B, the injection events are all 20 degrees after the inlet valve closes (so it's ready for the next cycle).
By "Simultanious", I assume you mean batch fire? I tried use paired injectors, I get the same misfire result as unphased sequential. This is a "good" result since in sequential mode, the car does not behave noticably better 50% of the times it is started. Grouping the injectors in batches of more than 2 is not helpful or desirable because it starts to make the pulsewidths at low load/RPM too small for reliable operation. (In fact batches of 2 is pushing it with these injectors.)
No matter what I do, there is no room for any sensors where the dizzy used to be. That area is full of coils. I have eliminated all the other causes of misfire that I can think of - the plugs and leads are good, trigger is good, driver transistors are good, injectors all work etc... The problem only occurs at idle - it does not seem to occur when there is any load on the engine, even if the RPM is similar (within reason) to idle.
I suspect camsync will not solve this problem completely, but it will help me eliminate another cause and it should let me investigate other things too. The injection is already sequential, and moving the injection events with respect to the inlet valve openings does seem to lessen the misfire problem, but never eliminate it. With camsync, I will be able to get better control of injector timing (there seems to be an ECU limitation about not overlapping injection and and ignition events on the same cylinder - don't ask me why - wasted spark gets in the way of this) and it will let me start to investigate individual cylinder fuelling trims - I suspect this is more important for making idle smoother than the timing.
I was thinking of a simple 1mm thick disc mounted on the cam with a single hole cut in it, lining up with the standard cam sensor position used in the "Thor" style RV8 engines. The ECU would see a single pulse every 2 crank rotations which defines the phase of the engine, leaving the crank sensor to give accurate position data.
Puma Racing:
The "random" phase is simply due to having a 36-1 trigger on the crank and deciding crank position from the first time a missing tooth event is encountered. The ECU knows where the crank is, but not the cam, but this doen't really matter because as you yourself say engines aren't usually all that bothered when the fuel is injected... It does not mean that the injection event happens at an entirely random and unpredictable time.
You don't need cam trigger to be sequential - all it means is that it fires one injector at a time. There is a 50/50 chance as to where the cam is when the engine starts, which can be considered as essentially a random variable. In position A, then the injection events are all 10 degrees before the inlet valve starts to open. In position B, the injection events are all 20 degrees after the inlet valve closes (so it's ready for the next cycle).
By "Simultanious", I assume you mean batch fire? I tried use paired injectors, I get the same misfire result as unphased sequential. This is a "good" result since in sequential mode, the car does not behave noticably better 50% of the times it is started. Grouping the injectors in batches of more than 2 is not helpful or desirable because it starts to make the pulsewidths at low load/RPM too small for reliable operation. (In fact batches of 2 is pushing it with these injectors.)
Ok, if I understand you are you saying the ecu has a firing order programmed into it and it triggers each injector sequentially but only knows from the crank sensor when each cylinder is at TDC but not if that's an inlet or exhaust stroke?
Well I guess the obvious thing is to double check that firing order however my rule of thumb in tracking down engine problems is basics first. Compression test, leak down test if necessary, valve clearances, cam timing, check cam lift on each lobe to make sure a lobe hasn't been wiped out.
Well I guess the obvious thing is to double check that firing order however my rule of thumb in tracking down engine problems is basics first. Compression test, leak down test if necessary, valve clearances, cam timing, check cam lift on each lobe to make sure a lobe hasn't been wiped out.
dnb said:
StevieTurbo:
No matter what I do, there is no room for any sensors where the dizzy used to be. That area is full of coils. I have eliminated all the other causes of misfire that I can think of - the plugs and leads are good, trigger is good, driver transistors are good, injectors all work etc... The problem only occurs at idle - it does not seem to occur when there is any load on the engine, even if the RPM is similar (within reason) to idle.
What inlet manifold do you have? Your symptoms seem rather similar to mine, which I have attributed to fuel stand-off caused by exhaust gas reversal under low rev/low load conditions. But with a conventional injection position in a standard manifold, I wouldn't have thought you would have the same problems.No matter what I do, there is no room for any sensors where the dizzy used to be. That area is full of coils. I have eliminated all the other causes of misfire that I can think of - the plugs and leads are good, trigger is good, driver transistors are good, injectors all work etc... The problem only occurs at idle - it does not seem to occur when there is any load on the engine, even if the RPM is similar (within reason) to idle.
dnb said:
StevieTurbo:
No matter what I do, there is no room for any sensors where the dizzy used to be.
Does the hole still exist for the dizzy ?No matter what I do, there is no room for any sensors where the dizzy used to be.
If so, how can there not be room ? Its a hole...fill it with a sensor. Lathe up a suitable adaptor to slot into the hole, which will hold a sensor. Nothing will protrude or extend externally apart from the wires.
Pumaracing said:
Well I guess the obvious thing is to double check that firing order however my rule of thumb in tracking down engine problems is basics first. Compression test, leak down test if necessary, valve clearances, cam timing, check cam lift on each lobe to make sure a lobe hasn't been wiped out.
Yes, I've done all of those...several times. I'm not totally daft 
The misfire is not a consistant cylinder either, just to make matters interesting.Pete - the inlet manifold is my own design of twin plenum. The injectors are in a conventional position, but nothing much else about it is normal. I have suspected it's something to do with exhaust gas reversion interfering with injection for a while, but thought I was just being silly.
Stevie - the hole for the dizzy is permanantly bunged up. To unbung it, I would have to take the timing cover off and would have to refit the dizzy drive gear on the cam too. If I go to these lengths I may as well do something like what's on the last generation of Rover v8 Range Rovers.
dnb said:
Pumaracing said:
Well I guess the obvious thing is to double check that firing order however my rule of thumb in tracking down engine problems is basics first. Compression test, leak down test if necessary, valve clearances, cam timing, check cam lift on each lobe to make sure a lobe hasn't been wiped out.
Yes, I've done all of those...several times. I'm not totally daft The misfire is not a consistant cylinder either, just to make matters interesting.Pete - the inlet manifold is my own design of twin plenum. The injectors are in a conventional position, but nothing much else about it is normal. I have suspected it's something to do with exhaust gas reversion interfering with injection for a while, but thought I was just being silly.
Stevie - the hole for the dizzy is permanantly bunged up. To unbung it, I would have to take the timing cover off and would have to refit the dizzy drive gear on the cam too. If I go to these lengths I may as well do something like what's on the last generation of Rover v8 Range Rovers.
Do you have any pics of your setup showing where the injectors are?
On mine I have very short runners into a plenum right down level with the heads and the injectors point *away* from the heads. It would idle consistently on 5-6 cylinders and then as the revs or load increased the other cylinders would kick in. Even in fifth it was very difficult to keep enough load on it under cruise conditions to keep them all firing cleanly.
Running it very very rich off load helped a bit but was not a viable long term solution.
I spent a long time looking for mechanical problems, tried fixing it with different cam, sequential injection all helped a little but didn't fix it. In the end I fixed it by adding upstream fuelling and using this for part load running.
Here's a picture of the plenums. Ignore the air filters, these were only temporary until the cold air induction was finished.
The manifold base is a modified TVR manifold base with injectors in the usual position.
Mine idles on 5 to 6 cylinders too. As soon as any load at all is applied, the misfires go away.
Here's a plot of some datalogging. The misfire indicator is called "MAD_spark" and is based on the standard deviation of the time taken for the crank to travel between trigger teeth during a cycle. If the std deviation is big, then the "outliers" can be found in the raw data and thus I know which cylinders are misbehaving. It is not constant or predictable.
The manifold base is a modified TVR manifold base with injectors in the usual position.
Mine idles on 5 to 6 cylinders too. As soon as any load at all is applied, the misfires go away.
Here's a plot of some datalogging. The misfire indicator is called "MAD_spark" and is based on the standard deviation of the time taken for the crank to travel between trigger teeth during a cycle. If the std deviation is big, then the "outliers" can be found in the raw data and thus I know which cylinders are misbehaving. It is not constant or predictable.
Looks good, I'd be interested to know more about it.
How are you doing the data logging - what sample frequency are you getting? The tooth rate obviously is very high even at tickover.
The idle speed is very low which will make EGR problems more pronounced. Does it help to pick the revs up to 1000, 1200 rpm?
Have you tried retarding the ignition around idle? If you pull it right back you will reduce the manifold depression which is directly linked to EGR.
PS Are those blue silicon hoses downstream of the throttle? Have you got solid inserts to stop them collapsing under vacuum?
How are you doing the data logging - what sample frequency are you getting? The tooth rate obviously is very high even at tickover.
The idle speed is very low which will make EGR problems more pronounced. Does it help to pick the revs up to 1000, 1200 rpm?
Have you tried retarding the ignition around idle? If you pull it right back you will reduce the manifold depression which is directly linked to EGR.
PS Are those blue silicon hoses downstream of the throttle? Have you got solid inserts to stop them collapsing under vacuum?
Edited by GreenV8S on Sunday 21st June 16:11
The datalogging is straight from the ECU, logging rate is approximately 20Hz. Not fast enough really, but the tooth speed data is internally recorded and downsampled for output. At idle, I think I get data for at least every 3rd cycle.
The idle speed is supposed to be 1000 RPM, but the misfires tend to drag it down well below this, and the ECU advances the timing to prevent stall, which causes an overspeed, which causes the ECU to retard the timing heavily which overshoots the RPM target which makes more misfires... It only takes a couple of misfires to be there when not expected or not be there when they are expected to upset the idle controller. I'm sure you've had similar nasty +ve feedback systems before.
The revs have to be 1500 at idle before the problem gets better, but it's still not exactly good.
I've not exhaustively tried retarding the timing at idle since it doesn't seem to like this and it stalls. The manifold vacuum at idle isn't particularly low - 55 to 60 kPa absolute - at the moment when it's not misfiring like mad. (It's silly on the chart above)
They are indeed blue silicone pipes. They have a spiral of wire inside them to stop them collapsing. It was done that way because of the lack of aluminium bends of the right radius.
The idle speed is supposed to be 1000 RPM, but the misfires tend to drag it down well below this, and the ECU advances the timing to prevent stall, which causes an overspeed, which causes the ECU to retard the timing heavily which overshoots the RPM target which makes more misfires... It only takes a couple of misfires to be there when not expected or not be there when they are expected to upset the idle controller. I'm sure you've had similar nasty +ve feedback systems before.
The revs have to be 1500 at idle before the problem gets better, but it's still not exactly good.
I've not exhaustively tried retarding the timing at idle since it doesn't seem to like this and it stalls. The manifold vacuum at idle isn't particularly low - 55 to 60 kPa absolute - at the moment when it's not misfiring like mad. (It's silly on the chart above)
They are indeed blue silicone pipes. They have a spiral of wire inside them to stop them collapsing. It was done that way because of the lack of aluminium bends of the right radius.
dnb said:
I'm sure you've had similar nasty +ve feedback systems before.
I'm injecting fuel about eight feet from the engine, what do you think!I would suggest you turn off the scattered spark and try to establish a fixed idle. You should only need active idle control if the engine parameters are changing due to temperature changes, electrical load etc and it's a distraction you don't need here.
If it is a fuel distribution problem then retarding the spark should improve it (you will need a lot more air in to maintain the same idle speed) since it will reduce manifold depression and increase the air flow. If this seems to make things worse perhaps the theory that it's a fuel distribution problem is wrong.
Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff