Has anyone experimented with the injector scaling section? If so, I'm curious to hear about your scaling strategy and experiences, for better or worse.

I'm likewise curious as to why the programmed injector capacity is smaller than the rated capacity, although I've seen this same trend in the hex of other ECUs as well.

Presently I am using an Xede with SMART system. I really like the refinement the SMART system has brought to driveability, and I prefer to keep it. My idea is to dyno tune the new setup using the ECU Flash to burn the settings into the ECU, and retain the Xede (with zero'd values) strictly for SMART control. In order to do this, I will need to do a good job of rescaling my present ECU settings (set for 650cc injectors) to the new 850s.

If no one's yet played with this, I'll just have to jump in feet first. . .

 

I am by no means an expert on this and I'm not really sure how much I can/should divulge in a public forum, but my experiences with injector scaling have shown that they do not scale in a linear fashion - e.g. my 780cc injectors did not seem to run very well when simply scaled "x" amount across the board. It's my understanding that the larger injectors cannot respond as quickly at lower RPM, therefore less of an adjustment needs to be made in those areas (or am I confusing this with the AEM issues?).

l8r)

 

I would have to guess that the stock ECU tends to scale the injectors at about 10% smaller to compensate for the injector size discrepency. Each injectors has sligtly different flow rate, even though they are same model from the same manufacturer. I think it is safe to start with about 20% smaller size than your actual injectors. Check to see if your AFR look right under light load cruising and idle, with the fuel trims within a reasonable range (+/- 10% will be good).

Someone has to test it first, but as long as you don't get into boost you are pretty safe.

 

thats an issue with the AEM having the incorrect injector drivers.

I'm still working with the scaling to make any real sense out of it as soon as i find out i'll be sure to post.

 

The Xede will be nice to have for both SMART and .1 deg timing control which around peak torque can equate to some significant gains not to mention the greater resolution. It definitely has its place.

 

It costs me $100 to have Shiv reflash my ECU for the larger injectors, or, I can pay the $89 for the cable and do it myself with the ECU Flash. With all due respect and thanks to Shiv, the decision was an easy one. I'll keep the Xede for SMART and fine timing control (as mentioned above).

As far as the programmed injector size and the trend to undersize it in the software, I've found (through programming Ford EECs) that preserving that trend tends to give better results even when using larger, aftermarket injectors. This is a question that I've conveniently forgotten to ask when I had the occasional opportunity to converse with the Ford programming engineers. Nevertheless, I'll stick with that trend and start with 10% less than rated capacity (my routine for the past 12 years).

 

My experience with injector scaling is that it is mainly a short cut to get slightly larger injectors mapped faster. I think taking the number placed inside that parameter too seriously like my 720's "need" to 90% in there isn't the right way to go about it. Every car is a little bit different depending on setup. Injector scaling or injector multiplier does become more critical when moving to a much larger injector 1000cc + as it effects the resolution in your fuel map, and it relaly be near impossible to idle the big injectors otherwise. This is what I've learned from tuning motec the last couple of years... as the injector scaler in their software doesn't even refer to CC size. It is simply a matter of injector pulse which they call IJPU.

Going off CC size is really just a laymans term because its simple to understand. I mean if injectors are rated 720cc by a manufacturer... what does that really mean? what voltage was that at, what duty cycle, what fuel pressure? there really aren't established standards for these things, especially internationally...

That's why I think its not productive to take what goes on in that parameter too literally, IMHO.

 

Are you referring to the 513 in the injector size window? I didn't think that was actual flow rate (cc), but rather a unitless scaling factor.

d

 

The scaling is a rough approximation (in the calculation used to convert the hex value to a number) of flow rate.. A JDM ECU is closer to 550cc's than a USDM which is set at 513... I have to presume its to "Fatten up" the entire map slightly for the US market and the type of gas prevelant.

The maximum value afte that calculation you can use is 1000, You can go larger than that, but 1000cc's is the largest "Value" you can use

Other settings (some not published) will help with getting big injectors to idle, but its nearly impossible to get a Stoich mixture with 1000cc injectors and therefore can swing between lean and rich, never really staying at a particular level. There are settings for closed loop AFR target that should be able to be adjusted for idle which will solve the problem, but the result would be a car that ran a bit richer, and of course, be troublesome on a sniffer test.

 

The inj scaler number is most likely just a multiplier to determine final on time... and as stated the 'number' displayed is translated inside.

If it works like all other ECUs..the fuel map number (14.7, 10.0, etc) internally translates to a bit value between 0-255 (8-bit). Then that number is multiplied against the injector scalar value, which can be thought of as a multiplier to tell the ECU how many mS of on time each bit equals. Then afer applying all the % trims, it arrives at a final on-time for that particular calculation.

 

So by inside you mean that if I change the 513 to 1000 you wont see all the numbers change on the map like an AEM, but it is calculated and the ecu will know they are 1000cc?

 

That is exactly how it works..

 

It does the work for you.. if you put 1000cc injectors in (rough number remember, you may have to adjust it to 950 or 900) the 14.7 should result in roughly the same AFR (in open loop) SInce these are calculated interpretations of numbers from 0-255 (single bytes) its not 100% dead on all the time, but it lets you focus on what your desired result should be close to.

 

I almost think it would be easier if the number displayed in the fuel map was the actual bit value (0-255). Because having a 'calculated' value there, and the inj scalar not really corresponding to an actual inj size somewhat muddles things. If there were a scalar number, which was multiplied by the bit value displayed, then one could fairly quickly rough the on-time needed for various fuel setups to do base mapping very qucikly. And it would make scaling for various fuel systems very straight forward using standard math and no conversions. Maybe if there could be a toggle to between a raw value, and calculated value, it might be useful?

 

Thanks for this goldmine of a thread. Just awesome!