[R4ND0M_AX3]

As part of our ongoing effort to understand the proper numbers to use in StreetTUNER (and other EM) I sent out my injectors to WitchHunter Performance for flow and latency tests. These were a brand new set of four PE 850s SH850FC.


Here are the results:
==============================================
Resistance = 13.6 ohms
Fluid Temp = 72DegF
Driver type = Saturated
Balance before cleaning was 2.3%
Balance after cleaning was 1.4%

Flow CC/Min @ 43.5psi (temp corrected) = 848 838 854 860

==============================================

Latency
6v = 9.8ms
9v = 2.7ms
12v = 1.8ms
14v = 1.5ms
16v = 1.3ms

Latency test comments from WitchHunter:
The injector is barely working at 6 volts. At low voltages, times will vary greatly with pressure and coil temp.

==============================================


I'm not too thrilled with the difference in flow (838 vs. 860). I will be putting the highest flowing injectors on the rear cyclinders.

Related threads:
Tuning JECS 720cc injectors and more
Cobb Tuning would like to ask for your feedback...

[2fas4u]

very interesting! Thanks for the data, those are the same units I run...

[b00st3d]

was there any tests done on the helix injectors?

[2fas4u]

yeah has me thinking myself.. Im wondering what mine are doing as far as flow...

[jays05]

Thank you so much for this, I'll be installing my PE 850's any day now and really needed latency numbers.

[WolfPlayer]

For those with the Nismo/JECS 740/720 injectors, here is the WitchHunter data for those injectors ...

http://www.bescaredracing.com/sti/fu...o/witchhunter/

I also had the stock injectors fully tested by WitchHunter.

View the attached Zip file. It has the results from the Nismo/JECS test and the results from the stock injectors ... and calculations needed for StreetTUNER. I'm currently running a 3555 scalar along with the latency values listed in the attached XLS.

t

[04stiguy]

Great info. guys. Appreciate the transfer of info. Thanks

[R4ND0M_AX3]

***EDIT: this post has info that was the result of a boost/vacuum leak I found later***

So I finished up my install (injectors, Fuel pump, APS inlet, APS TGV deletes) and used wolfplayers excel spreadsheet to calculate the latency and scale in StreetTUNER.

I came up with 3122 for the new scale.
The latency worked out to;
6v=9.75
9v=2.79
12v=1.37
14v=0.98
16v=0.70

The car started and ran pretty rough. It was idling lean (17:1). The fuel trim eventually maxed out at 25%. I took it for a short drive and found that higher rpms were better. I messed with a couple things until I figured out the the latency was what was off. So I bumped up the values until I got a smooth idle with near zero fuel trim.

This is what I ended up with. ***EDIT: see note at top of this post***

6=7.66
9=2.69
12=1.67*
14=1.55*
16=0.75

Scale=3122

I only changed the 12 and 14 volt values because there wasn't any way to test for the other voltages.

I still have some fine tuning to do with this. I might just scale all of the latency values by the same amount.

[WolfPlayer]

Another thought would be to try a different approach just for the heck of it. Try using those latency values but then a different scalar and see what happens. If I would try to use the exact values in the spreadsheet, I too would have been lean. However, I raised the scalar to compensate (mostly because the new latency values I was using were already higher than the previous values I was using).

What I am really curious about is what is better: higher latency and lower scalar or lower latency and higher scalar. I think the idea is tweak them until your overall LTFTs are equivalent across the board ... and not 0 at idle with high trims at higher RPMs.

t

[R4ND0M_AX3]

I knew from my Hondata days that if the latency was off then you would be rich down low nad lean up top or the other way around. The scaler would move everything in one direction.

[jays05]

Any updates? I'm wondering if anybody else has installed/tuned for these injectors and could add anything?

I installed my PE 850's yesterday. I started the car using the 3122 scaler and the latency values that R4ND0M_AX3 used... lots of smoke and it was idling at 11:1 when the car first started. I switched to the translated WitchHunter latency values and the car was running lean. I tried adjusting the fuel injector scaler but the car didn't seem like it liked that, so I scaled the MAF (+22% low end, +6% high end). It probably isn't the correct way to tune it, but it seems to have worked. The car idles nicely now and WOT AFR's are just about perfect, but when I lean into the gas it goes very rich for a moment, I see it hitting high 9's. Also in the 2-3k range I can back out of the throttle slowly and get it to a point where the car shudders. Anybody dealt with these issues before?

With mid 11 AFR's I was seeing 96-98% IDC's earlier, now I see 58-60%

[R4ND0M_AX3]

For the last week I've been driving with these settings on the PE850's.

7.66 2.69 1.70 1.15 1.15
2997

I haven't been able to work on fine tuning this because of the vacuum leak I had but now I can start again. The above values aren't that far off.

[STirish]

I would try getting your fuel trims at idle correct (+-2-5%) with the latency and scalar before anything else. Once you're close, you can tidy the trims up w/ intake scaling. You should be able to get your trims close with out even putting the car in gear.

[WolfPlayer]

Jay,

Scaling the MAF the way you did is totally wrong. I think you know this but am hoping that by me poking at you ... maybe it will give you the motivation to go back and redo it Simply put, the easiest thing to do is to start with a scalar in the neighborhood (you did). However, I have found the slightly lower scalars seem to work a little better (but require you to retune your WOT a tad bit). You might want to consider trying a 3050 scalar and then slowly dialing your latency numbers to get you close to 0 fuel trims at idle. Latency values make a huge difference. For example, the original values posted by Random were 1.5 @ 14V and then new values are 1.15 @ 14V. 0.35ms at idle is a lot of fuel!!!! Basically, start the car with 3100 scalar and a 14V value of around 1.0 (my calculations show you actually need a 0.98 value for the 14v point). Tune the latency up/down from there to get you to where you need to be for fuel trims at idle. Do extremely short blasts of WOT slowly crawling up the RPMs. Tune your fuel curve as needed. Furthermore, Daemon and I have seen that as the RPMs increase at WOT the battery voltage goes down. Thus, the 12V latency value is really important. If this 12V value is off then you'll have some issues at part throttle. Try to dial in this 12V value through part throttle fuel corrections. It will also affect your WOT fuel curve too.

Good luck. Just approach it as scientific as possible and take your time. Download the Excel sheet I posted above and plug in Random's latency values. It will give you the new values you should be using in StreetTUNER (that is where I got the 0.98 from).

FWIW, I am running the Nismos but will be installing the Helix 860s for testing purposes.

t

[jays05]

Hah, to throw another wrench into all of this, I discovered a huge boost leak yesterday. That explains a lot. The 2.5->2.75" silicone coupler (ebay special) that I was using to connect the IC piping to the TB ripped. It tore right along the edge of a t-bolt clamp, so it was not noticable until everything was apart. I cobbed something together to make it drivable, but didn't get a chance to datalog since I fixed it last night. After 150 or so miles, my oil stinks like gas.

I agree that scaling the MAF was the wrong way to go about everything - the boost leak explains it all! I'll be traveling for work the rest of the week, but plan on spending a couple hours on friday working on injector latencies. I am tentatively planning on installing my perrin gt35r kit next weekend - in this case, I only need to get the car safe and drivable enough to bring it to get tuned.