Injector Upgrade

Last updated: January 31, 2000

Date: Wed, 12 Apr 2000 23:36:17 -0700
From: "Shiv Pathak" (

> This reference suggests you divide the cc/min rating of the injectors by 5
> to get est'd hp capacity. Seems reasonable. For the saab 16V-T, which is 160
> hp stock, and 215cc injectors at 2.5 bar figure at 171 hp potential, and
> many do get 200 hp from stock ones, with 3 bar or RR-regulators.
> The FD has 2x(550+850) = 2800 cc/min, est'd hp is 560, or mabe 475 rwhp. And
> this is rated at 2.5 bar fuel regulation.
> Seems like many jump to bigger/more injectors at lower power levels. Where's
> the gas going? Are they running that rich?
> Running an xtra 10 psi fuel (40psi at idle) gets 12% more flow, so if 560 hp
> is at 13:1 a/f for a typical boinger, then you could be at 11.6:1 a/f for a
> less efficient rotary?

For a turbo charged rotary, most would suggest that the BSFC is closer to .65, not .5 which is more appropriate for an n/a piston engine.

Calculating one injector at a time while assuming a max desirable duty cycle of 80% and a standard fuel rail pressure differential of 43psi we get...

Two 550cc/min divided by 10.5 = 52lbs/hour 2 x (52x0.80/.65) = 128 horsepower

Two 850cc/min divided by 10.5 = 81lbs/hour 2 x (81x0.80/.65) = 200 horsepower

200+128= 328HP

Raising fuel rail pressure by 10psi will increase flow by...

square root of (53/43)= 11% Now, the safe maximum horsepower output is 364HP with fuel pressures around 65psi at full boost (assuming 12psi of boost).

Since most injectors don't start acting wierd until 80psi or so, there's another 15psi that could, theoretically, be taken advantage of. Let's assume that we now want to run, oh say, 16psi of boost. 80psi-16psi=64psi

square root of (64/43)=22%

Now, maximum safe horsepower output is 328 x 1.22 or 400HP

Of course, this assumes a fuel pump of sufficient performance. And, as many have proven, it is possible to extract more or less power that these rule o' thumb numbers would suggest.


Date: Mon, 31 Jan 2000 12:12:51 -0800
From: "Ulen, Robert S" (

I was told by a fuel injector testing facility that the stock injectors go wide open ("static") at approx. 93% duty-cycle. So in other words, the stock injectors are *not* controllable between 93% and 100% - they are always full open in this range; the same as if they where at 100% duty-cycle continuously.

Also, the stock injectors (and I'm sure others), lose flow rate as the duty-cycle increases. The stock injectors are actually flowing about 83% of rating when the actual duty-cycle (as seen by the PMS) is at 90%. This is assuming they are perfectly clean of course. So maybe these guys that see a 98% duty cycle from their PMSs, but the injector is really at about 90% of its ideal flow capacity.

The stock FD injectors feed fuel directly from the side of the body, which helps alot with heat issues. But I'm sure that if they where near 100% for a very long time that the coils could heat up some. The fuel flow rate would be petty high at high duty-cycles, so if the heat transfer from the coils to the fuel is decent, then the overheating may not be a big concern with this injector design. Mazda designed the injectors for this very purpose (ref Service Highlights Manual).

The bottom line is still an accurate A/F measurement for tuning purposes. But its important to understand when upgrading the injectors where the limits are, so you can still tune accurately when the time comes. If the injectors have to run at 90% all the time to give the correct A/F, then they are *too small* for that particular application. Its better to size the injectors to run around 60 to 70% duty-cycle at WOT conditions.


Date: Fri, 18 Dec 1998 13:49:04 -0500
From: Sandy Linthicum

Note: stock injectors when controlled correctly can normally supply enought fuel at 14-15psi for short periods under hot/warm conditions. However, the injectors are functioning at their max duty cycle and if they continue to do so will get hot and flow less fuel. You are also pushing the limits of the stock fuel pump, lines etc. You are usually OK if everything is working perfectly but a minor glitch can bring disaster.

Also, the fuel delivery requirements are not just based on PSI boost pressure. Intake charge temp (directly related to ambient air temp), humidity, etc. all have a significant influence. A car that runs 14psi with no problem at 70F my be lean and detonating at 14psi 40F degrees.

Many of us use our FDs for track events where you are at WOT a large part of 20 minutes at a time, under full load with a hot motor (instead of 12 second sessions in a drag race). I have done track events where the temp were in the 90's and ones where temps were in the 30's. There are a wide range of tuning parameters which need to be addressed and they can vary quite a bit.

In EITHER case, the smart money mixes in some high octane race gas when in doubt.


Date: Tue, 22 Dec 1998 19:14:11 -0500
From: Sandy Linthicum

The stock ECU is also completely limited to controlling injectors having the exact characteristics of the stock injectors (impedience, etc). No aftermarket piggyback unit does much more than modify the stock ECU injector pulsewidth pattern.

Complete replacement of the stock ECU with a Haltech, Motec, etc. is necessary to use other injectors.

Carlos has the best halfway solution, using the stock injectors (stock secondaries substituted for primaries) and secondary injectors enlarged to flow 50% more fuel than stock. To the ECU, the injector ratios and other characteristics remain stock while allow a pgmable aftermarket unit like the PFS computer to do reasonable fuel maps.


Date: Tue, 22 Dec 1998 21:59:09 -0500
From: Nick Riefner

I am using 1220 cc secondaries. The ratio is screwed up, and this does cause the engine to have a rich condition at the point where the secondaries come on. Luckily, with a programmable computer, you can tune some/most of this richness away. With the Wolf 3D system that I am going to be installing soon, I can tune even more away, by using the staged injector feature. The key is having the programmable computer. The more programmability you have, the better. If you keep the ratio correct, however, you will have a much easier time setting the car up.


Date: Thu, 30 Apr 1998 18:00:19 -0700 (PDT)
From: NetBlazer

Two 850CC stock secondary injectors installed in the primarys location is typical.

Placing them before the throttle body (like the aftermarket pipe vendors do) leaves something to be desired. Places like Trust know this, but offer their Greddy elbow with the additional injectors for additional revenue, not because it's the place to put them.

To do it right you need to have an injector rail TIG welded into the upper intake manifold. Some have tried in the theoretically best spot (right above the turbos) which resulted in too much heat soak. If properly shielded that would be a better place I suppose, but either is A LOT better than before the TB.


Date: Thu, 30 Apr 1998 19:09:21 -0400
From: "Ryan W. Schlagheck" (

TWORX7S asked: "Can anyone explain to me the "added injectors" thing?"

The additional injectors will feed the additional fuel needed to run above 15psi boost, which is where I'm told the stock injectors reach their maximum duty cycle (Jim Labreck, 4/28/98).

Instead of this option, you can opt to upgrade the stock injectors to 750cc's to get the proper amount of fuel needed for running at 15psi and higher.

It is suggested that you add a fuel pressure regulator w/ stainless steel hose and a competition fuel pump just to be sure everything is safe and you don't detonate and blow up the engine. You'll also need the computer upgraded/re-upgraded with a new EPROM with fuel and ignition maps mapped out to this level of output, plus raising/eliminating the boost cut. Ari & Trey at Rotary Performance are working on a prototype fuel system that they might sell as a competition upgrade which will package the above pieces (possibly with a J&S Knock Sensor as well).


Date: Fri, 01 May 1998 00:21:43 -0500
From: Trey Cobb (

We've found that typically the fuel pump starts dropping fuel pressure first because of lack of flow. The next thing to happen is that the injectors max out so you can either push the fuel pressure up alittle more to squeeze more fuel through, add additional injectors, and/or use larger stock injectors. The GReddy elbow is an easy way to add some injectors or you can plumb them directly into your intake manifold. We've done a few different things and found a method that still gives you good gas mileage (relatively speaking) and off-boost drivability.

>Ari & Trey at Rotary Performance are working on a
>prototype fuel system that they might sell as a competition

Yeah we're still working on it so I can't give exact details but it is bolt on stuff. The only thing we're currently selling that's part of the prototype fuel system is the fuel pump which runs $264 and exchanges directly with the stock unit (and no, it's not a Cosmo unit).


Date: Wed, 13 May 1998 22:23:05 -0400

>I will soon be testing whether my fuel pressure is dropping off >in the upper gears at sustained 15 lbs of boost with my stock fuel pump.

At 15 lbs (and about .80 O2 at 6500 rpm), the critical limit is not the fuel pump, but the fact that the injectors are at 100% duty cycle. It only gets uglier from there.

>Has anyone done any testing to see how far an FD with mods can be pushed
>without the need for additional fuel pumps or injectors, using stock OEM fuel
>lines? I spoke with Eric Hsu of XS Engineering and he indicated that in their
>testing of the stock fuel system 450hp at the flywheel was the upper limit. I am
>currently at the 400 mark at the flywheel and question the need to upgrade the
>pump. What are others experiences in this area? Hard data would be great.

With a 3rd gen. dyno'ed at just under 400 HP at the wheel, the injectors are max'ed out (100 percent duty cycle) at 6800 RPM with 15 lbs. of boost and mapped to maintain **.78V** on the O2 (with 25 percent race gas for insurance). By 7500 rpms that O2 had decayed to .76V.

From my guesstimationes, the stock fuel system can maintain 85% injector duty cycle (approximate ideal long term operation max) when mapped to .82 only to about 330 - 350 horses at the wheels.

BTW, in my rotary narcissitic, self-promotion, check out my nearly over-engineered fuel system design:


Date: Sun, 19 Jul 1998 12:09:47 -0400

Injector Pulsewidth Staging Algorithm

The injectors on the stock [3] fuel injection system are staged (i.e. only primary. injectors followed by primary + secondary injectors). The system includes two primary (pri) 550cc injectors and two secondary (sec) 850cc injectors, with a switchover point from pri only to pri + sec which is a f(x) of coolant H20, rpm and load (to name the main ones). At switchover, the computer notes the pulsewidth (pw) of the pri injectors and reduces the pw to 39% of the pri pw, for the switchover pw of the pri + sec.

For illustration, let us assume a pri pw of 10ms when switchover occurs. Also all calculations will be for fuel flow PER ROTOR bank of injectors. The pulsewidth of 10ms * 550cc(per minute) injectors = .010(sec.) / 60(sec/min) * 550(cc/min) = ** .09166.. cc's per pulse. **

When the switchover occurs, the system stages from the pri to pri PLUS sec injectors, fuel capacity increases to 1400cc (550 + 850) per rotor. Now the fuel flow math looks something like this: 1400cc * 10ms * .39 (switchover correction factor) = .010(sec.) / 60(sec/min) * 1400(cc/min) * .3928 (switchover correction factor) = ** .09166.. cc's per pulse. **

Notice that both fuel flows are equal, and therefore the switchover does not interrupt the total fuel delivery.

So why is this important? Both the Techtom-type and the "piggy-back" computer upgrades use this algorithm at switchover, and are therefore limited by it. None of them, as far as I can tell, have the capability to work around this mapping logic. This becomes a problem if the injector sizes are changed, because the stock system is designed around this critical correction factor. Make the ratio of the pri to pri + sec too large, and the switchover results in momentary lean condition. Make the ratio to small, and the car runs rich at staging. So the trick to increasing injector size is to maintain the ratio of the pri (550cc) to pri + sec (1400cc) as close to .39 as possible.

And yes, this knowledge did come about through much mechanical self-flagellation, until I installed a set of 1220cc sec to match my 850cc pri. The ratio isn't perfect (.41) but it's the max flow within a margin of the desired ratio that I could design given the stock injectors and fuel management system. And I DEFINITELY CAN GO THROUGH SOME GAS.. Viva Gasolina!


Date: Fri, 30 Apr 1999 10:39:58 -1000 (HST)
From: (Richard H Thomason)

While the 850s will fit in place of the 550 primaries, you could not fit non OEM, non side feed "custom" injectors in place of the secondaries. What you would need to do is get ahold of an extra set of 850s and have them enlarged. Or you could enlarge your 550s to 850cc, then enlarge your 850s to 1220s (the 550s cannot be enlarged to 1220 cc). Or, you could do what I did and use your 850s in place of the primaries and fabricate a custom fuel rail to hold non OEM injectors for secondaries (1,400cc in my case). I'm not running the stock ECU however, and only went this rout for reasons unrelated to this question. Figure to spend about $175 a pop to mod the OEM injectors and expect to pay a pretty price for two extra 850s.


Date: Wed, 5 Apr 2000 21:27:52 GMT

It'll lose atomization with bigger injectors as you can see from my dyno chart before and after the injectors upgrade. It takes about 1.5psi more to get the same rwhp from my car. I was able to get 403rwhp@14.5psi on stock injectors but I can only get 400rwhp@16psi with the big injectors. This is just my theory but I could be wrong. Maybe somebody with better understanding on fluid dynamics can help explain this for us.


Re: Injector Pulse Width vs. Duty Cycle

Date: January 30, 2001

Check out at RC Engineering's website they have a program to calculate that sorta stuff.


From: e s (
Date: January 31, 2001

Duty cycle is the duration that a signal is at it's peak divided by the the duration of one cycle. (Time)

So if our friend has a Pulse width of 13.3ms then we need to know what frequency the signal is to derive the duty cycle. (Time is the reciprocal of Freq.) I suspect the frequency varies with engine speed but that is probably not the case since I have never attemted to view an injector signal. (which would make my calculations WRONG!)

But if so, 4025 RPM divided by 60 equals 67.0833333. The reciprocal of that is 14.9068 ms. (this is the total signal duration at 4025 RPM) 13.3 over 14.9068 equals 89.2208333 percent.

So if the injectors are fired once per rev. then our friend is running an injector duty cycle of 89.22%

How many times is one injector fired in a rotary? Is our friend even talking about a rotary engine? Have I lost my mind?


From: Sandy Linthicum (
Date: January 31, 2001

100% duty cycle is applying constant voltage to the injector to hold it open and have it spray constantly

85% duty is the max an injector will operate properly. At this rate the 15% it is off is bare enought to allow it to close completely. At higher rates, the injector doesn't have time to close completely before the next open pulse.


Also some good info on the Injector Cleaning and Injector Install pages. --Steve

[ Mail me ] [ To Lightning home page ] [ To my home page ] [ Copyright Notice ]