How do you determine injector size?

[TLicense]

Like the subject says.

 

From what I understand 18 psi boost is on the limit of stock jap injector duty. If I want to increase the boost to say 2 bar, how would I determine the size of injectors required?

 

Many thanks

[Thorin]

Very roughly the injector size should correspond to max BHP I believe I read. ie. j-spec 440's would be absolute maxed for 440bhp, 720cc injectors good for 720bhp etc.

[dandan]

Try this matey, may put you in the ballpark.

 

 

http://www.rceng.com/technical.htm

 

[Adam W]

Be aware that you can't run much over 18-20psi on pump fuel without serious tuning of the ignition map, no matter how much injector you have.

[Guest Terry S]

As Adam rightly says, the octane of the fuel is your enemy. You will need some serious fuel ( not quantity) to run 2 bar.

[TLicense]

Not planning to run 2 bar, just curiuous as to how you work out injector sizing.

 

After Leon's tuned up the AEM, it should hopefully be properly mapped for 18 psi, but I'm a little concerned abot maxing out the small stock JDM injectors. (Don't want to run lean now do we? - Shouldn't be too much of a problem as the AEM will be closed loop with wideband o2). It'll also have warlbro pumps and an FSE which should also help.

 

 

Cheers

[Guest Terry S]

Be careful there. CLC won't help you on WOT, just part throttle. Seems kind of a waste having the AEM to control stock injectors.

[TLicense]

Well I know what your saying there terry, I needed a new ecu for the gearbox conversion, I didn't want to shell out twice, once for a stock ecu, and then again for something for when I start modifying. However I couldn't afford the outlay at the same time for larger injectors. Future proofing really.

 

So at WOT the ECU is still going to be blind as far as air/fuel? I didn't realise this was the case. I guess even more reason for getting some decent sized injectors in quick.

 

What size of injectors would you recommend, bearing in mind turbo wise I'm probably going to go stock hybrids and maybe later HKS twins.

With the stock hybrids I'd be aiming to get 450-500 bhp and with the HKS twins about 600. Is this reasonable or am I completely in dream land, and in need of more research. I know there's a turbo comparison guide somewhere about's but I'll be damned if I can find it.

[Wez]

I wouldn`t say the ECU is blind but your setup would depend on how the ECU reacts to certain variables for example knock etc.

 

There is also a feature in the AEM which enables you to set a target AFR, this is done on a seperate map with its own load and rpm sites, the ECU will then adjust the fueling on the fly to try and achieve your target.

 

I have not looked into this myself yet, I currently do not have my wideband wired into my ECU as I using an external display.

 

[TLicense]

Wez,

 

The target AFR feature is what I was thinking about. Isn't that basically closed loop? Or am I missing something fundamental? (I'm beginning to get the feeling I am)

 

Terry,

 

What's different about being at WOT that means the AEM doesn't work closed loop with the o2 sensor? After all is it not just effectively a range of values depending on engine load? (What I think I mean is what extra function means you can't run closed loop at WOT?)

Again I get the feeling I'm missing something fundamental and would probably learn a lot from doing some more research into engine mapping!

 

Sorry for being an ignorant gimp.

[Chris Wilson]

There are no ECU's in the price range most people on here (with respect) would consider buying that would reliably run closed loop without a tight base map to trim around. Personally I suggest you run open loop and just use the wideband to keep an eye on things. As the mapping, if done properly, will cost a LOT more than the ECU and injectors you might as well set the thing up with whatever you intend to run for the forseeable future.

 

I have always maintained, and maintain here and now, that short of mapping on an engine dyno you will never get a really good result. Just my opinion and probably contentious in these `ere parts

[TLicense]

Originally posted by Chris Wilson

I have always maintained, and maintain here and now, that short of mapping on an engine dyno you will never get a really good result. Just my opinion and probably contentious in these `ere parts

 

Careful now Chris, I think you've wandered into "street dyno" territory here

 

No your absolutely right, an Engine dyno is the way to go. Only problem is can't really afford to go the extra mile, so I'm going to do half a job. Bit annoying actually after all the blood sweat and tears, (Not my blood sweat and tears mind, just my cash), to stop short of something truey pukka. God damned budgets.

 

I'm also with you regarding doing the mapping once and once only. Again I'm currently limited by budget, which in the end is going to cost me more in the long run I know.

 

Believing that the AEM would run well in "closed loop mode", I've opted not to buy the 02 display. I've got a feeling I'm going to live to regret that. Is it the inacuracy of the AEM to run closed loop that means it isn't closed loop at WOT?

 

Sorry to be asking simple questions, but am just on the very tippy edge of engine mapping here, and trying to get an understanding of the fundamentals.

[Chris Wilson]

Closed loop, at its current level of sophistication, is only reliable if allowed to trim a pre recorded map, say by a max of 10% either way. They can and do get confused if they are expected to just "run the show". I still think you should run open loop, off a decently drafted map, and use a wide band with a display, and maybe at idle for an MOT test if you worry about such things

[Wez]

Originally posted by Chris Wilson

Closed loop, at its current level of sophistication, is only reliable if allowed to trim a pre recorded map, say by a max of 10% either way. They can and do get confused if they are expected to just "run the show". I still think you should run open loop, off a decently drafted map, and use a wide band with a display, and maybe at idle for an MOT test if you worry about such things

 

Hey Chris,

 

How you keeping?

 

I am pretty sure that this is how the AEM works, you setup your base map as closely as possible and you then setup another map which contains your target AFR and the ECU from my memory doesn`t change the value in the base map but has the ability to tweak on the fly to acheive its AFR target.

 

Of course the wideband would have to wired into the ECU which I do not so I am not able to test this but I believe there are people in US using this method.

 

There is also another feature which is automapping which will update the base map as well but this has been known to create holes in the map and should only be used during initial setup.

 

I personally dont believe there is a better substitute for a correctly setup base map.

 

[Guest Terry S]

As Chris said it is all down to how good the base map is. CLC will not work at all if the base map is not close enough to the target AFR's. There is an awful lot of work here. I use CLC on part throttle and in the switches to Open loop on higher TPS settings. These ECU's are good but not perfect. Just because you put in a target AFR and have a wide band connected to the ecu ( I have both) doesn't mean the car will achieve the AFR. That is where your mapper earns his money.

[Wez]

Originally posted by Terry S

As Chris said it is all down to how good the base map is. CLC will not work at all if the base map is not close enough to the target AFR's. There is an awful lot of work here. I use CLC on part throttle and in the switches to Open loop on higher TPS settings. These ECU's are good but not perfect. Just because you put in a target AFR and have a wide band connected to the ecu ( I have both) doesn't mean the car will achieve the AFR. That is where your mapper earns his money.

 

I agree with everything you are saying here Terry, as I said before I don`t have any O2 feedback into my ECU its all running from the maps which have been setup.

 

[Chris Wilson]

Self learning ECU's are a million miles away at the moment. For example, and ecu needs to map ignition timing. Assuming you let the ecu have control and advance until det is heard, which is bad anyway, the current technology of knock sensors makes them unreliable at higher rpms. Production knock sensors are only read at lower RPMs, up to arounf peak torque. So how the hell does it self map at peak power. As has been said here, this is where an experienced mapper earns his crust. And an experienced mapper will want to work on an engine dyno, not being driven by some maniac down the A1 all day, or on the rollers trying to listen and map amid a hubbub of fans and roller thrum.

[Guest Terry S]

FWIW Chris, Autronic were looking at an Autotune for the ign timing as well, using the spark pklug to determine det, but AFAIK gave up as it seems unworkable, so as you say, complete ECU control seems years away.

[TLicense]

Ahhh, I think I'm beginning to understand a little now.

 

OK, So the way forward with current technology and my stock twin set-up would be to go CLC up to a certain %age of throttle and then to go into open loop as per Terry's set-up.

 

The danger is with my piddley 440 cc injectors and at 18psi, that it'll start to run lean at WOT (Best place to do it as well - NOT!), Also bearing in mind I don't have a display for the wideband or a EGT gauge and the fact that the ecu will happily let the engine blow itself to bit's in open loop mode, it means making sure the fuel system and map is spot on. As you can all probably tell, as far as mapping, I'm completely in the hands of whoever is doing the mapping, in this case Leon. However the fuel system is something I can make sure is pukka myself.

However this brings me back to my original question, if I'm intending to be able to run at absolutely peak 1.5 bar (Not right now but in the future), what size of injectors do you guys think will be up to the job for then, but not too much of a pain in the ass to set up for the shorter duration's that they'll be using in the mean time? Obviously the UK spec injectors would tidy up the problem at 18 psi, but are they likely to run out of steam if I cranked the boost up to 1.5 bar with a pair of hybrid's?

 

Regards

[Ian C]

550cc injectors would be fine. I found that 1.5bar slowed the car down (overheating the airstream, maybe even some ignition retarding due to knock, and the EGTs went above 900degC), 1.3 to 1.4 is peak power and the 550's do the job.

 

An Apexi AFC could handle this job for £200 and the mapping costs. An AEM is a bit sledgehammer/peanut like

 

Yes, you are in the hands of the mapper. Scary, isn't it? CW set up mine and it hasn't blown up for two years so I'm just about relaxing

 

-Ian

[TLicense]

Ian,

 

I take it you mean 1.5 bar with stock twins and SMIC?

 

I know what you mean as far as overkill but it's not as if I'm going to stop modifying, the AEM will come into it's own when I start to look at hybrid/larger turbo's, and I didn't want to shell out for an AFC and then later for another ECU mod. Figured I'd hit it once.

[Ian C]

Nope, hybrids, FMIC, and 550's

 

Fair enough on the forward planning aspect of the AEM

 

-Ian

[TLicense]

Hmmmm, I would have thought 1.5 bar would have been OK with the hybrid's and a FMIC? What have you done as far as the hybrid's as in trims and things (I hate all this "stage 1/2/3" stuff. Who defines what's stage 1 or what's stage 3?).

What kind of inlet temps were you seeing?

[Guest Terry S]

Tony, listen to Ian. The cars make no significant power gains above 1.3 bar on stock based turbo's. I also agree that the 550's would be ideal, but if single is your long term goal then a set of bigger injectors would be a good idea now, as long as the AEM can control them well ( I assume it will, but don't know enough about them to say for sure).