Megasquirt Injection question
Turbo Mini Forums

i must admit I am thinking about a wet manifold (electronic carb) setup for the GF's car,

what to my mind is a more easily tunable carb appeals to me immensly... especially after the 200 mile trip to and from brighton that returned 40 plus MPG (from my quick calculations) from the 7 porter including all the sitting in traffic, and a fair amount of running bdly due to wet HT leads.

definatly an efi convert now!!

one other thought for the GF's car is to put a 1.2 fiat head on (sort of stealling roberts idea) if you used the 8v cinquenceto head I think you could get away without valve cutouts, maybe not a gret flowing head, but would make a good economic drivable engine.

Schrödinger's cat - so which one am I ???

My main point was to make it clear that the wet manifold is a compromise and for which reasons it is.

Having said that, I agree that it is a major benefit over a carb from a tuning point of view and for all the side benefits mentioned. And for a daily driver with an engine that's not pushed to the limit it should make for a very nice setup.

And as mentioned before, matnrach's setup is definitely a good one for a wet manifold injection setup. And there wasn't much of a choice a year ago but to go this way. This is becoming less so with the work done by Paul and once Rod has his engine running it will give another set of data that hopefully will give a trend to what is needed get a good tune with port injection.

Jean

Edited by jbelanger on 27th May, 2009.

http://www.jbperf.com/

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

I hope maby some time next year I should be able to help with the port injection Feed back, but right now all I want to do is get away from carbs and get my 998 N/A on some thing a bit more tunable(SP),

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

matnrach, as you say, it will be a long time to prove the benefits one way or another, but look at how many people on this forum post threads and pictures of failed pistons......

I don't want to be one of them.

I personally agree with you that the benefits may not be "huge" from port injection..... until you run high levels of boost.

Only time will tell.....

And, TBH, a "dyno shoot out" would probably be the last thing on some peoples minds considering a couple of recent posts....

Your solution has to be a lot better than an SU if only because of all the other bits that come with EMS.

I would just like to think mine will be a little bit better still, and advance the process !!!

Schrödinger's cat - so which one am I ???

I do agree with you but failed pistons can come from a variety of reasons not just because you have a 5 or 10% difference in fuel distribution between cylinders.

Also I'm not sure if boost will make it that much worse. What is your reasoning? I stand to be corrected though.

Anyway I hope the port injection code works as it is certainly the optimum solution at the limit.

You mention 5 to 10% but you have absolutely no data to back this up. I cannot say that it's not that but neither can you.

And when the boost is going higher, a difference of one AFR point (or more) will make a huge difference in combustion temperature and you will see melted pistons at the leaner AFR.

Jean

http://www.jbperf.com/

You are saying that I can't back a 5-10% difference up and then everything you just said was equally as unsubstanciated.

I will hunt around on the net to find some data but I do remember seeing the sensitivity of piston crown temperature to lambda and it wasn't that high.
If the average lambda is controlled rich then unless the difference from the mean to the leanest cylinder is over 0.1 continously at full load, meltdown is extremely unlikely to occur (again in my opinion)

The part about a lean AFR melting piston is correct and based on easily available data in a lot of places. And the part about combustion temperature going up with boost is also based on published data (and physics).

The part about having 1 AFR point difference is not based on actual data but if it does happen that you have such a big difference then melting a piston is a possibility. Also if the point difference is between 12:1 and 13:1 or 13.5:1 and 14.5:1 then the results will be quite different.

Until an actual measurement is made it all remains conjecture. But what would also be important is to know how the "average" measurement relates to the actual AFR in the inner and outer cylinders.

Jean

http://www.jbperf.com/

Piston crown temperature is fundamentally proportional to cylinder pressure which in a turbo engine is proportional to boost.
The cylinder pressure is maximum at optimum lambda

http://www.sr20-forum.com/tuning/16056-why...ll-welcome.html

Either side of optimum lambda the combustion temperature reduces as less imep is produced.

Exhaust temp does go up however.

Unless I am reading this wrong I can't see how a relatively small variation in lambda from a mean of 0.82-0.85 would melt a piston assuming the ignition was not too advanced or charge temps did not wildly increase.
Failure under boost is most likley to occur from an increase in crown temperature due to the increase in pressure alone , not lambda variation.
However the variation in lambda to the lean side will affect the knock limit and it is this which will obviously cause piston failure not the fact it runs lean itself.

So assuming the difference between mean and leanest cylinders are kept sensible (this is where I can't produce actual data) and ignition timing is such that knock is mitigated to the same extent on both, there can't be such a big difference in the piston limit between PFI and TBI.

The only advantage of PFI, as I said before will be a small (again I can't quantify) improvement in the knock limit on the leanest cylinders and a small loss in combustion efficiency from running the richest cylinders too rich.
The knock limit on both PFI and TBI, unless done correctly on a dyno with individual pressure sensing will be as equally difficult to do on both setups.
Of course I forgot the two obvious advantages of PFI are
a) emmisions (lower Nox and HC).
b) fuel consumption as you need to make the average AFR slightly richer for the same level of driveability/knock limit
Not sure of the real life benefit of these two but I guess it is worth developing to get the absolute best solution if you are able to exploit it.

Sorry for the rambling but these are my thoughts

Edited by matnrach on 29th May, 2009.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

It is not the difference in AFR alone that will damage the pistons, it is the ignition timing.
Sure you will be better off as long as your ignition timing is adequately retarded.

Piston temp reduces as you go either way from optimally rich. It only goes up if you advance to compensate for the reduction in flame speed from the optimum point.

So you will still have to run retarded from knock the same way a TBI sytem would , but you may be able to run a small amount more ignition.

This is the main advantage with PFI.
Yes you could destroy an engine at 20psi with TBI but so you can just as easily with PFI.
It depends on the person entering the numbers in the map.

Edited by matnrach on 29th May, 2009.

Matnrach, you obviously have a good working system that you are happy with. You also have a handle of what you need to do ignition timing wise to get the best out of it. I wish you well.

But, there is little point in promoting your system for high boost use without proving the fuel distribution.

There is every chance that your outer cylinders are running lean, due to charge robbing and your ignition timing will have to be retarded to counter for that.

With port injection on the new MS code and timed injection, I have equal AFRs at idle all the way up to 6000rpm at WOT. I can now optimise my ignition timing knowing that all cylinders are working the same.

A TBI system with no control over the timing of the slugs of fuel cannot achieve that.

My goal of maximum mpg on the 998 and 200+hp on the 1293 mean that I have to start with good fuel distribution. Everything else then follows.

Surely you can understand that?

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

One other point.....

I can't see anywhere in Coupe's build thread, but in matnrach's pictures it's clearly an MS1.

As the siamese code is based on MS2-Extra, presumably it will only run on an MS2 ???

But, at the end of the day, matnrach has it running, coupe soon will and it TBI/SPI has to be at least as good as an SU, better for everything else on the engines and a damned site easier to tune than filing needles once you've got the hang of it.

Schrödinger's cat - so which one am I ???

Edited by matnrach on 30th May, 2009.

Now, Without opening a very large can of worms.

Lets say i want to run an easily tuneable carb.

Im thinking an SPI/TBI set-up for the fact thats it will be essentially a carb that i could easily remap it for a differnent set-up with less restrictions than a carb set-up.

So the next step up from a carb in my mind would be a TBI set-up.

Now to the question! From reading through this thread and im hearing that a TBI set-up is dangerous to your engine on boost. Im readin that and thinking that surely it cant be worse than a carb?

And i was prepared to run my future engine on a carb so if i decide to run TBI then it should really be better than a carb. no?

I appreciate that the AFR difference between the inner and outer cylinders is not a good thing and its going to be there with a TBI set-up just as much as it is going to be with a carb. I also appreciate that a port injection/MPI set-up is going to be better still, and ideally will be the way to go!

However i am basically a noob to fuel injection so i would much preffer to wet my feet with a TBI set-up and just use it to better the carb and hopefully gain abit more power/tuneability.

After all how many problems have the likes of Nic and others had with destroying pistons with a carb set-up? Plenty of people on the forum have properly set-up a carb and ran large power through the set-up without destroying pistons every time the put thier foot down (not you Shane!).

So how can it be anyworse than a carb? Obviously set-up comes into play but if both properly set-up then my little brain just cant see it being worse?

Please inform my newbie mind.

My setup is actually all up and running now, running 2 injectors in alternating mode on a wet manifold (so basically an electronic carb) and its doing well so far. More work to be done on the map, but the car drives well, cruises nicely, and seems to be running well on boost (apart from my charge temps, but then thats a different story).

A TBI is worse than a carb in that the air/fuel mixture is not constant. A carb will have pretty much the same AFR downstream of the carb so the mixture going into the cylinders will be the same from start to finish of the intake stroke (I know this is not entirely true but close enough to illustrate my point). With a TBI, the AFR will be very rich when the injector is on and very lean when the injector is off.

Having said that, if the TBI has 4 injection events per engine cycle everything should even out. And the advantage of the TBI is that it is relatively easy to have whatever fueling curve the engine wants by simply entering the correct numbers on a PC after analysing some data logs. With a carb, it is almost impossible to optimise all load and RPM fueling points (especially if you want max power at the top end and max economy at the lower end).

So I think you can take this first step and install a TBI setup. However, it would not be a bad idea to plan for any future upgrade you might want to do. One thing that would be very good to really know what is going on would be to get 2 WBO2 sensors. I know this is quite expensive but there isn't any other way to know how good or bad your setup is and to maybe take some corrective steps. And you'll be able to get the best compromise on the AFR for all cylinders (and know what it actually is instead of relying on some average from all cylinders).

Jean

http://www.jbperf.com/

right so if i do it then i will need (sort of already realised from previous posts but the wallet didnt want to accept it) to geta twin wideband set-up on the go. So that means sample chambers and all the rest.

I would be expecting to go to port injection in the future but i want to get a grasp of things first and try and optimise this system even if its just from a comparison point of view.

Looks like i have some designing and planning to do!

Coupe what boost are you running at the moment? Im going to have a detailed look through your build thread later. See if i can get some ideas.

As Jean says, I'm running 4 injections events per cycle and it seems to work fine at the moment.

James, I'm currently running 18 PSi - car is currently running on the safe side - quite rich and not alot of advance at the moment. I'm still tweaking and improving it all the time.

What size injectors are you using coupe? Im thinking this might be quite a limiting factor in power output with just two injectors...

Im going to be running an emerald ECU so (from what ive been reading) i could use 4 smaller injectors at the throttle body, using two to idle and then bringing the other two in when needed.

I want to have injectors capeable of running up to 250hp (dont want to run out of puff on them so ive gone OTT on the power) I want to be putting out around 200hp+

From reading through emeralds site it says this:

"As a rough guide to power output, take the injector flow in cc/min and divide by 5 then multiply by the number of injectors."

soooo in order to be rated at 250hp:

250 / 2 = 125

125 x 5 = 625

so if using two injectors i would need circa 625cc/min injectors to cover my power output.

Would i be right in thinking that this is going to be detremental to the idle and cruise conditions? or would you have to go alot bigger for this to come into play?

Or have i got this completely wrong and because of the TBI type set-up these rules dont apply/are different?

Cheers, James. Off to research more. expect more questions people.

I'm running a pair of 650's in alternating mode. They're fine at cruise, although I am running the car a bit rich for the moment to be safe. I've got problems with idle at the moment, but I'm still working on that, and have an idle control valve to go on yet for cold starts, but there's no reason why it can't run and idle fine with injectors of that size.

The A-Series definitely does need a squirt for every ignition event - I experimented early on with a squirt on every other ignition event and it didn't like it in any way shape or form.

My injectors are both after the throttle body, but approximately 8" along the inlet tract, so a good way from the inlet manifold itself. There are some pictures of my banana shaped home made inlet tract in my build thread. To be honest, for something that involved no science, or design, and was just knocked up in my garage (my first piece of welding at that), it works pretty well so far!