Long duration cam, idle and closed loop EFI
Discussion
This winter I'm planning on converting to EFI. To this point had planned to go with MassFlo however some feedback from a current owner has made me think again about a speed density system.
MassFlo tunes itself in closed loop fashion which is fine for short duration camshafts. With a long duration camshaft the valve overlap insures that some unburned fuel will escape the tailpipe at idle and likely cause the ECU to lean the mixture. The end result is a bad idle with low speed drivability problems.
I would love to install a simple system that could self tune and use off the shelf parts that are easily diagnosed and replaced when needed but I'm thinking that this is not worth the loss of power associated with a "factory" camshaft.
Any thoughts or experience with this?
MassFlo tunes itself in closed loop fashion which is fine for short duration camshafts. With a long duration camshaft the valve overlap insures that some unburned fuel will escape the tailpipe at idle and likely cause the ECU to lean the mixture. The end result is a bad idle with low speed drivability problems.
I would love to install a simple system that could self tune and use off the shelf parts that are easily diagnosed and replaced when needed but I'm thinking that this is not worth the loss of power associated with a "factory" camshaft.
Any thoughts or experience with this?
Hi
Don't know anything about these systems, but I can tell you a bit about escaped fuel and 'self learning'.
Escaped Fuel: This causes wetting of the lambda sensor and is a sure fire way to quickly destroy a lambda sensor. They are designed to measure the oxygen content of burnt gas, give them fuel molecules and they die, fast! It also obviously totally confuses them leading to wacky corrections by the ECU in a closed loop system.
Talking of closed loop systems.....hmmn! exhaustive subject, however nearly all closed loop systems have a base map, and use the lambda sensor (and other sensors) to augment the fuelling over time, rather than reacting instantaneously. This is why replacing a faulty lambda sensor does not always yield instant results (depends on how the lambda failed).
Now, a closed loop system is compromised by design, and flies in the face of having predictable throttle response, something that is critical in bikes, and also in some car applications - when you open the gas mid or post corner you NEED TO KNOW how the engine will react or you have trouble, the small throttle openings that are used to 'find/feel' grip are critical. This is why most high performance engines (and all performance bike engines) only use closed loop at the lower RPM, idle and slightly above it, the meat of the power band is all statically mapped (against some other sensors of coarse such as TPS, MAF, IAT, IAP, AP etc etc).
There are some good books out there.... there is nothing mystic about EFI, it is all rather simple (unlike a carb). I have spent many, many hours coming up with ways to treak and improve EFI systems by either hacking the ECU, or adding our own electronics.
EG: Ultra sonic cleaning obviously cleans injectors, but leave the injectors in the ultra sonic bath for extended periods of time and you start to 'open' the ports / nozzles allowing for a higher flow. This can be easily measured and is one of the 'tricks' used by some production based teams to eak out a bit more..... (you would not believe what I have been asked to do to give someone a competitive advantage by, erm, clever cheating, alas I have always refused!).
Gareth
Don't know anything about these systems, but I can tell you a bit about escaped fuel and 'self learning'.
Escaped Fuel: This causes wetting of the lambda sensor and is a sure fire way to quickly destroy a lambda sensor. They are designed to measure the oxygen content of burnt gas, give them fuel molecules and they die, fast! It also obviously totally confuses them leading to wacky corrections by the ECU in a closed loop system.
Talking of closed loop systems.....hmmn! exhaustive subject, however nearly all closed loop systems have a base map, and use the lambda sensor (and other sensors) to augment the fuelling over time, rather than reacting instantaneously. This is why replacing a faulty lambda sensor does not always yield instant results (depends on how the lambda failed).
Now, a closed loop system is compromised by design, and flies in the face of having predictable throttle response, something that is critical in bikes, and also in some car applications - when you open the gas mid or post corner you NEED TO KNOW how the engine will react or you have trouble, the small throttle openings that are used to 'find/feel' grip are critical. This is why most high performance engines (and all performance bike engines) only use closed loop at the lower RPM, idle and slightly above it, the meat of the power band is all statically mapped (against some other sensors of coarse such as TPS, MAF, IAT, IAP, AP etc etc).
There are some good books out there.... there is nothing mystic about EFI, it is all rather simple (unlike a carb). I have spent many, many hours coming up with ways to treak and improve EFI systems by either hacking the ECU, or adding our own electronics.
EG: Ultra sonic cleaning obviously cleans injectors, but leave the injectors in the ultra sonic bath for extended periods of time and you start to 'open' the ports / nozzles allowing for a higher flow. This can be easily measured and is one of the 'tricks' used by some production based teams to eak out a bit more..... (you would not believe what I have been asked to do to give someone a competitive advantage by, erm, clever cheating, alas I have always refused!).
Gareth
http://www.youtube.com/watch?v=Pc-BDt1dj9U
Found one. The website reports that this is only good to 500 Hp.
Hopefully this can be upgraded and will allow use of a fuel pump that is outside the tank.
Found one. The website reports that this is only good to 500 Hp.
Hopefully this can be upgraded and will allow use of a fuel pump that is outside the tank.
k wright said:
The YouTube video makes this look very easy to set up.
Of course - but its never going to be that simple! Was that system using a wideband? - if not, that 13.1 afr reading on the handheld is largely a guess with a narrow band lambda.I dont really see much benefit in the TBI based systems, essentially its an electronic carb - so sure, you get the benefit of adjusting the mixture with a laptop - but you still get the wall wetting and mixture dropout problems - especially with you guys who run the big single plane open plenum manifolds.
I would say to aim for port injection (MPFI as holley calls it) - as you are squirting directly down the port then.
Not sure what you mean by Lambdas die fast with unburnt fuel.
I have wideband lambda controlling a self learning EFI system on my car which has a racey cam in it. have now covered 5000 miles and lambda still reacting quickly etc. I know 5000 miles is nothing but its a fair bit for an ultima.
I have wideband lambda controlling a self learning EFI system on my car which has a racey cam in it. have now covered 5000 miles and lambda still reacting quickly etc. I know 5000 miles is nothing but its a fair bit for an ultima.
MarkWebb said:
Not sure what you mean by Lambdas die fast with unburnt fuel.
I have wideband lambda controlling a self learning EFI system on my car which has a racey cam in it. have now covered 5000 miles and lambda still reacting quickly etc. I know 5000 miles is nothing but its a fair bit for an ultima.
Hi MarkI have wideband lambda controlling a self learning EFI system on my car which has a racey cam in it. have now covered 5000 miles and lambda still reacting quickly etc. I know 5000 miles is nothing but its a fair bit for an ultima.
Well, Lambda works by a chemical reaction, if you 'wet' them with unburned fuel then they die really quickly.
Perhaps your 'racy' cam still limits unburned fuel being pumped out the exhaust, or you lambda is on its way out.... A little bit of 'wetting' is inevitable (choke / cold starting etc) but it very bad for the sensor itself.
Gareth
I have a setup on a SBC that’s close to what your look at. It was installed by the guy that built my car but I have done quite a bit of development on the system over the past two years. The setup is; Emerald ECU running ignition and injection maps, crank sensor timing and single coil spark via distributor, throttle Position Sensor for load sensing, 8 point batched fire injection using an adapted single plane manifold, AFR input for closed loop from Innovate LC1.
If your going down the route of putting your own system together the points below are a couple of relevant things learnt whilst workin with my system;
The SBC has a combination of big cam (248 deg @5 thou) and single plane manifold. The combination results in cylinder robbing and the car kangarooing when at low engine loads between 1800 to 2100 rpm. I can Drive round this but it does get in the way a low road speed. Nothing I have tried with the ECU setup and remapping will illuminate this so finally I am going to be putting in a slightly lower duration cam with an LSA of about 113 deg for EFI type applications to hopefully address the root cause. I would be very carful selecting your cam profile and manifold design if you want both low speed drivability and high peak power.
I installed an Innovate LC1 wideband AFR and can run closed loop in real time or target map AFR modes. My experience with closed loop is that in target map mode it is very useful helping lean map corrections at engine speeds above about 1.7k rpm. Below this figure the AFR reading is spiky, even with data smoothing set (due to fuel in the exhaust and the fuel robbing mentioned above. For for this reason I don’t run real time closed loop. Emerald also said real time closed loop isn’t the way to go.
The emerald is a really good and flexible system though and its been fun to work with to tune the base map that came with the car. If you are up for an alternative to buying a turn key system, I would recommend looking at an Emerald setup. The ECU has a lot of flexibility and will run load input via manifold air pressure or mass air sensors if you prefer.
Nick
If your going down the route of putting your own system together the points below are a couple of relevant things learnt whilst workin with my system;
The SBC has a combination of big cam (248 deg @5 thou) and single plane manifold. The combination results in cylinder robbing and the car kangarooing when at low engine loads between 1800 to 2100 rpm. I can Drive round this but it does get in the way a low road speed. Nothing I have tried with the ECU setup and remapping will illuminate this so finally I am going to be putting in a slightly lower duration cam with an LSA of about 113 deg for EFI type applications to hopefully address the root cause. I would be very carful selecting your cam profile and manifold design if you want both low speed drivability and high peak power.
I installed an Innovate LC1 wideband AFR and can run closed loop in real time or target map AFR modes. My experience with closed loop is that in target map mode it is very useful helping lean map corrections at engine speeds above about 1.7k rpm. Below this figure the AFR reading is spiky, even with data smoothing set (due to fuel in the exhaust and the fuel robbing mentioned above. For for this reason I don’t run real time closed loop. Emerald also said real time closed loop isn’t the way to go.
The emerald is a really good and flexible system though and its been fun to work with to tune the base map that came with the car. If you are up for an alternative to buying a turn key system, I would recommend looking at an Emerald setup. The ECU has a lot of flexibility and will run load input via manifold air pressure or mass air sensors if you prefer.
Nick
eliot said:
NickQuick said:
single coil spark via distributor
This is something I would avoid if possible. On a high revving V8 you start to run out of coil dwell time resulting in a weaker spark. A simple wasted spark,either directly driven or ford edis eliminates this problem.My ECU has been back at Emerald for 3 weeks now to get an additoinal coil driver fitted. The rest of the wasted coil install is ready to go so i'll be on wasted spark when the ECU gets back.
NickQuick said:
eliot said:
My ECU has been back at Emerald for 3 weeks now to get an additoinal coil driver fitted. The rest of the wasted coil install is ready to go so i'll be on wasted spark when the ECU gets back.
Funny how people critisise megasquirt ECU's for their diy add-on nature, yet think nothing of sending their ECU back to the manufacturer to do exactly the same thing. (not Critising Nick, just an observation)Eliot,
I sense some sensitivity from somewhere about Megasquirt
, never been any criticism from here. The megasquirt web handbook has been a great help on many topics including planning this wasted spark fit. The module based system looks interesting but Emerald was already fitted. It’s worth noting the low costs of Emerald upgrades, the coil driver is costing £25 and a major ECU firmware version upgrade was only £35 which given what it provided was great value. If you can cope with the turn around to get the ECU back, the Emerald upgrades are very reasonable. I can’t comment on other vendors approach to upgrades.
Nick.
Ps really nice informative web site.
I sense some sensitivity from somewhere about Megasquirt
, never been any criticism from here. The megasquirt web handbook has been a great help on many topics including planning this wasted spark fit. The module based system looks interesting but Emerald was already fitted. It’s worth noting the low costs of Emerald upgrades, the coil driver is costing £25 and a major ECU firmware version upgrade was only £35 which given what it provided was great value. If you can cope with the turn around to get the ECU back, the Emerald upgrades are very reasonable. I can’t comment on other vendors approach to upgrades.Nick.
Ps really nice informative web site.
Gassing Station | Ultima | Top of Page | What's New | My Stuff