canarys2000

I bought these injectors from Kraken as well, because they are apparently a similar stopper valve/guts as some more expensive injectors. I had the exact same problem as those above. Adjusting dead time did not help me. If I ran with the calculated req. fuel in tunerstudio, my VE table was around 70 at idle. I'm confident it is not tunerstudio's fault. Here is how I made mine idle better than stock with a normal looking (i.e., ~40 at idle - 34 now [7/5/23]) VE table:

1. Used 2.5-3mm spacers (i.e., washers) that sit around the top of the existing plastic spacers. I used some thick lock washers I had sitting around; pay careful attention when tightening everything and make sure the injectors are being slightly compressed (~15-19 ft-lbs)
2. Used required fuel multiplier of 7.0 (updated 7/5/23) (I got this by doing ratios of my new to old VE tables after getting these running). The car wouldn't even stay running until I nearly doubled my VE table originally - it was not cool
3. Used dead time of 1.250 ms (updated 7/5/23), determined by analyzing lots of data
4. Used the voltage curve and the non linear pulsewidths as follows (I got these from digging around online, the voltage curves are from other EV14 700cc bosch injectors, and the nonlinearity is from some of the higher-end similar sized injectors):

I have not figured out why these are flowing so much less than they should. I assumed my 20 year fuel pump was bad, but now I'm not sure. The poor flow numbers suggest they might only be good for ~200 WHP at stock fuel pressure
Injectors are working well now after the settings above and getting my tuning (acceleration enrichment, VE, etc.) dialed in. The max duty cycle I see is 75% - these are good at around 220 WHP for now (Updated 7/5/23)

der_vierte

What's your duty cycle?

canarys2000

Duty cycle is 1.6 while idling and 36 at WOT. I attached a log file. This is an all stock 1.8 VVT engine with the Bosch EV14 700ccs

Kyle Konz

Got these values from Kraken today in regards to these injectors in case anyone stumbles on this thread in the future. Values are for 43.5 psi/3 bar, this is stock pressure for an NA. I don't know if these can be ratiod up or not for an nb (which runs higher fuel pressure). Note that these are the same values listed in the picture above (written out because sometimes pictures die after a while).
8v=2.979ms
9v=2.186ms
10v=1.700ms
11v=1.448ms
12v=1.268ms
13v=1.042ms
14v=0.917ms
15v=0.799ms

Random side note for those on speeduino. Default settings is set to "Open only" for injector settings. For these values you'll need to set it to "whole pulsewidth" (basically open+close time).

canarys2000

These settings are interesting. They are basically saying that at 13.2V the dead time is around 1 ms. I spent a lot more time on mine after my post above, and found that something closer to 1.25 ms at 13.2V worked better. I followed the procedure from this thread. Since my original post above, I turbocharged my 2001 NB and it is running great after a few weeks of troubleshooting. I made this megasquirt video: before I turbo'd it. I'll be making a more advanced video in a few weeks showing the injector dead time determination and many other important things I had to learn

Kyle Konz

I'll look into that. Also something else seems way off (could be the dead time, could be I set the req fuel incorrect). I ended up with like 4.3. Had to set the ve to like 60 in the idle region, my old ve was like 37 before installing these.

Kyle Konz

I think I may have figured out a disconnect between what canary and I are seeing. Canary is on an NB and I'm on an NA. NB's run a higher fuel pressure (60ish if I remember correctly) whereas NA's run a more "normal" fuel pressure of 43. That might explain some of the differences that are being shown and why Canary's settings very much did not work for my setup. I'm finally getting back around to getting mine started after some down time, and I'll post results as I have them available to kind of close up this thread.

reynolds716

Hello I'm new to tunerstudio for MS3 Pro and I'm not sure where I can make these changes for dead time. I see dead time under the fuel settings but when I open it I don't see where I enter voltage and ms nujmbers. Help would be appreciated. Thanks

curly

iTrader: (25)

Press the little button with the three dots in the upper right corner, you can see it in the above photos. It hides/shows the table.

reynolds716

When you do that it shows in % and voltage not in millisecond and voltage

Kyle Konz

Right, so you'll need to do some ratios for yours instead of straight values. I'm not super familiar with MS3, but the way I believe it works is there should be a base dead time at 13.2V. Use that as your 100% and then and then you ratio from there based on the dead time values given with your injectors. For example, if your dead time at 13.2V is 1.1ms and at 14.5V it's 0.85ms then you would do 0.85/1.1=77.3% at 14.5V in the chart.

ianp0817

I did a little Excel data analysis for ***** and grins. It probably has little to no impact on performance, but it makes me feel better.
Kraken gave me 8 data points, but there are only 6 cells. From looking at the graph of Kraken's data, it looks like the best points to exclude are at 11V and 14V, as they have the least impact on the curve.
This gives these results:



The Values are pretty similar to Canary's, but these are more accurate to what Kraken has reported for 3 Bar.
I am also going to use Canary's small pulse-width values, but I am interested in how he came up with them.

I will report my results

Values in case of image death:
8V => 234.9
9V => 172.4
10V => 134.1
12V => 100.0
13V => 82.2
15V => 63.0

canarys2000

Woah, I need to look at these posts more often. Thank you everyone for the contributions to this thread!

Ian P's data probably will work fine. But, I think it is customary to make 13.2 volts = 100% (not sure why). I scale my data to conform to this to make it easier to compare with other injector data sets I find online. In Ian P's case, the same voltage curve he shows above could be shown as:

(Dead time: 1.018 ms at 100%)
8 V => 292.6 %
9 V => 214.8 %
10 V => 167.0 %
12 V => 124.6 %
13 V => 102.4 %
15 V => 78.5 %

When you multiply this dead time by the percentage, it will give the same pulsewidth curve as Ian P's. But note how 100% is exactly at 13.2 V.

Regarding my system, it is an NB2, so my fuel pressure is 60 psi (4 bar) and its a return-less fuel system, therefore my posted numbers would all change if this were used on an NA, which operates at ~40 psi (2.75 bar). The deadtime would likely be lower on the NA. The Injector Dynamics website has a downloadable Excel sheet for larger injectors that has fuel pressure as an input; this can give you an idea of what to expect with different fuel pressure. My small pulse data and voltage data I obtained from doing an average of some other bosch 700 CC injectors I found on the internet.

Ian P, one recommendation I have. If your megasquirt controls your battery voltage at 14.0 (which I think is typical), you might want to purposely use this point, because your voltage is being controlled at 14.0 the vast majority of the time you are running and especially at wide open throttle. If you have data for that point, it might be the most valuable data point.

An update:

1. I broke a piston skirt in my main engine (previous overheat), so I had a temporary NB2 engine in my car recently with the same Kraken supplied injectors, and for some reason I no longer needed washers for additional spacers on the fuel rail. The OEM black plastic spacers worked fine this time. Either the aluminum spacers on the Kraken EV14 injectors settled lower, or there is some minor difference in fuel rails - still not sure!
2. Also - I gave into peer pressure and purchased Injector Dynamics 1050x, thinking it might make my car run better (even though it runs fine). I just thought that having the right small pulsewidth and dead time might help with cold/hot starts. The difference is negligible between the IDs and the Kraken EV14 700C. Don't waste your money unless you are going for 250+ horsepower.

ianp0817

I will go ahead and switch to using the values that are centered at 13.2. I'm not sure of the reason either.

As for the data point at 14, I more or less understand your reasoning, but I'm not sure what difference it would make internally. From what I understand, the MegaSquirt simply performs a linear interpolation to get the value to use. Since the actual voltage will never be exactly 14(I don't think), it will always be interpolating so not having a value there makes no difference as long as the interpolated value at 14 is the same as the one you would put in. If that is the case, you would want to get as close to the recommended curve as possible with your limited number of points.
Of course, I could be totally misunderstanding exactly what is going on. Admittedly I am quite new to this and I'm going off of assumptions.

Also, if you aren't too happy with your ID's, I will gladly take them off your hands

canarys2000

My reasoning was simply that if you have data points at 13, 14, and 15, and you eliminate 14, you are losing data fidelity at your most important point. Eliminating 14 makes an inherent assumption about linearity between 13 and 15, even though 14 is probably not linear between 13 and 15 - the data probably has a slightly different slope between 13 and 14 versus 14 and 15. As long as the data isn't crazily different, this almost certainly does not matter. It's just an exercise in obsessiveness lol. You are probably fine with your points

I'm not sure how accurate the voltage is on the aftermarket ECU - but yeah, probably never exactly at 14. It seems pretty stable though on my MS3. I am definitely no expert either; I learn something new every day! I'll keep the ID's for now. Since I have a spare set of injectors... I might as well find another miata to turbo

ianp0817

not an inherent assumption, we can check it. The maximum difference is 0.9205 - 0.917 = 0.0035ms. I'm pretty happy with this but at this point, in for a penny in for a pound.

Linear spacing so just average and subtract. Oh boy are we shaving hairs...
I've concluded that it does not really matter. I'm pretty happy with my excluded values, and how well the car runs for that matter. My VE's are looking much more stable and realistic at idle now.

You should supercharge the second one.

canarys2000

If she's running well, it didn't happen on accident. Good to hear!! Supercharger would be cool - less flow restriction in the exhaust!

sharp_cheddar

Adding some more information to this discussion, take it for what it's worth...

The injectors that I received with my Kraken kit were Bosch part 0280158235. I reached out to Kraken for injector data and was provided with the same table that @kricsesz received from Finjector. I'm installing on an NB, so I needed data for 4 bar. Kraken didn't come through with this information, so I reached out to Finjector a few days ago and they provided the attached data from testing that was done this week. Note that there are differences from the data that was provided previously. And shout out to @canarys2000 for taking the lead on a lot of this.



Finjector product link for reference: https://www.finjector.com/eng/ev14_1...mo_nd-p-722658

oreo

Interesting. I never realized that the dead time varies with fuel pressure.

It doesn't vary too much if you have decent battery voltage though.

Ukturbomx

Hi,

i tried using the data in that picture from finjector and something doesn’t seem right.

my VE at idle is around 82.. which seems double what it probably should be.

has anyone else used this data and it worked?