Kraken EFR 6758 with some FPR testing
 

About a year ago, I didn't quite make the power I was hoping for: https://www.miataturbo.net/dynos-tim...jector-101222/

That setup was great, pulled hard to reline, but wastegate was 265hp. I built the engine, added ID1050x's, but the turbo ran out of steam just under 300.

Enter Kraken's awesome manifold and EFR's 6758. My coworker built a 3" exhaust, I did these dyno's with no filter, haven't figured that out yet. I will say, Aeromotive's compact EFI regulator is complete junk, never buy it. Get Aeromotive part # 13109, it's a bigger regulator, but worked flawlessly. Here's a log of fuel pressure vs. MAP on the old regulator; after it initially matched the increase in MAP, it would completely nose dive:


https://cimg4.ibsrv.net/gimg/www.mia...c73b5c949f.png

And the new FPR:

https://cimg4.ibsrv.net/gimg/www.mia...c6a517e13c.png

I would chaulk that up to some how a bad brand new FPR, but Kris at KO Racing has experienced this exact same symptom on a number of other Aeromotive compact EFI regulators.

Moving on to the results, here's 3 runs right around the ~335 max:


https://cimg4.ibsrv.net/gimg/www.mia...dc8d5dc77d.png

Throughout the 30+ dyno pulls, I fine tuned fuel and ignition per cylinder, basically based off visually looking at the average knock per cylinder. It really helped to scale the graphs to the same 50% max so they were even compared to each other. This is done under "calculated fields", "field min and max editor", then set max for knock 1-4 to ~50. Should show something like below. After this run, we added a little fuel to cylinder 4 and 3, and they smoothed out a lot. Hope that helped someone.


https://cimg6.ibsrv.net/gimg/www.mia...d5965b2aaa.png

By the end of the runs, we had tried more timing, more boost, less timing and more boost, but it simply wouldn't make more power without concerning increases in knock, essentially this is the max for this setup on pump gas. I may return with a tank of race gas to shoot for 400, we'll see.

Very nice results! I can appreciate the testing and tuning that went into this. Boost looks really flat too.

If you want to try for any more power on pump gas, I would recommend dropping to an even colder plug one day and see if your knock counts drop. That's the one thing I've found on miatas with high boost, they need a really cold plug to prevent knock.

Ohhhh this is relevant to my interests!! Is this on pump 91 or 93?

These are BKR7Es, looks like 8Es are popular in the high power 2jz club, might be worth a shot, although I probably won't go back without different fuel as well. Worth saying, this is also on stock VVT coils with NGK wires.

It's 92 in Oregon and Washington. E85 is only sold at one local station, but I work at a shop with a number of race fuel only vehicles, so I may just do that. Don't have any ethanol concerns, and once I tune it, I can turn it back down to pump gas limits, and just top off as it runs out, rather than re-tune.

Excellent, it will be interesting to see how close I come to that 330 number on 93 on the east coast here! I would be really happy with that on pump... e85 is supposedly in our area, so I set up my flex sensor, but not sure how available it really is. Good numbers!

Solid power on pump gas, great job man!
I agree about small/compact fpr's (same with tomei): if you're upgrading, go for the big boy each time, the compact ones just don't work as well.

are you at elevation? I'd expect the efr to not have that dip between 3-4k. (totally just curious)

Only 500 feet. That may be a VVT tuning issue? I never messed with it. This EFR has a very similar spool curve as the evox/msm turbo I was running previously, graph below (red is efr, orange is msm). Boost is at 100% dc until 3250, it may be that as well. Could have kept pushing it, but I didn't have a significant peak before MAP leveled off, so I left it. The red line that does have a peak is actually closed loop boost control. I enabled it on the last run to make sure it was worked reasonable well, in the hopes that on a cooler day it would pull some DC before shooting straight to my 250kpa overboost cut.


https://cimg3.ibsrv.net/gimg/www.mia...b8af05bf90.png

In that case yeah a few simple tweaks and you'll pick that back up, and it's really minor anyway (but who would say no to more torque under the curve amirite? :likecat: )

Here's Ed's "10ae" vs. my Kraken setup, his run is in blue. I forget what turbo he had, but it's an absurdflow setup. I labeled his run "hail mary", I believe we threw everything at it to see what'd it make before backing it down to 240 low boost, 320 high boost.
https://cimg8.ibsrv.net/gimg/www.mia...e23e8730d2.png

Very interesting, thanks for posting this. I have a FPR that is mounted on the NA8 fuel rail and looks like the same diaphragm configuration as the Compact EFI. I have a FPR gauge, but it isn't wired into the MS to log currently. Looks like I'll need get my TinyIOx working sooner than later to log this.

I honestly think fuel pressure is just as important as anything oil related. You wouldn't put a shitty oil pump in the car, why put a shitty regulator. That goes for monitoring it. I always try to monitor both with the ECU. I tried to get the fuel pressure feature working, so it uses it in the fuel calculation, but couldn't figure it out. natewin PM'd me about some timing tables, but since you can't attach anything in a PM, I'll post them here. I use cold advance to pull timing when over heating. Between MAT retard, cold advance, and knock, if the three of them happened at the same time, I'm pretty sure the car wouldn't run...



https://cimg1.ibsrv.net/gimg/www.mia...ed8173b4b3.png
Nate

Curious about the 2300 rpm timing bump theory?

The maps far from perfect, if I had to guess, I adjusted some idle stuff, some in boost stuff, and the 1800’column sort of sits between the two and I never interpolated between the two to bring it up a bit.

Another question, I maybe just missed it:
What fuel is that timing map based on? Single digit timing in the mid range at 50kpa boost is the lowest I have seen in a long time. I'm sure as experienced as you are, you have your reasons. Is it really that knock limited?

This was on a tank of 92. I haven't touched the map after the dyno, so no smoothing has been done, teaches me to post something before looking at it. That day I only did 2 runs under 200kpa. I monitored knock, but didn't start dialing in the ignition table until 6 or 8 runs in while I worked on boost control.
I can disable EBC on the fly, and those runs were to make sure fueling looked good. It's meant to be a low power map, although those ignition figures are probably safer than they need to be.
Also, I can't reach 155kpa til 3400rpm, out of that single digit area.

Here's a smoothed map that might make a little more sense, I'll have to do some logs to make sure the low boost advance I added didn't fuck anything. Thanks a lot, not!


https://cimg1.ibsrv.net/gimg/www.mia...d424753deb.png

And looking at my logs, here's a map shows you what actually matters, I'll hit none of the zeroes. These are WOT pulls from 1500rpm, so if I started at a later RPM, spool would be even lower.


https://cimg8.ibsrv.net/gimg/www.mia...0fb5f2d583.png

I didn't want to offend you in any way :D

The smoothed out map looks much more reasonable, I just wanted to know where this came from. You know, to learn :)

Sorry to drift from informative fuel pressure thread.
Got me doing way too much reading, Didn't know you can adjust fuel and spark per cylinder with any Megasquirt?
Really amazed by your knock graphing. Every log I have ever done / seen has much more noise on #3 and less on others as they get further from knock sensor.
Sure seems hard to separate / filter knock from noise.

Yup, under ignition options enable spark trim, and under engine and sequential settings, enable injector trims.

My biggest issue on the dyno was interpreting the data. Here's what it looked like at first, I couldn't quite tell what cylinder was worse visually, unless I looked at the MAX value, and it all looked like my engine was about to explode.

https://cimg5.ibsrv.net/gimg/www.mia...aeee301121.png

So I went to "calculated fields" and "Field Min and Max editor", and changed knock 1,2,3,4 to not be auto, with a min of 0 and max of 50. Here's what that produced. This made more more sense, and I could tell I wasn't seeing a lot of knock. You also need to do a log revving the engine semi slowly in neutral, and use that as your threshold, as noise will inherently increase as RPMs increase. You're not looking for a flat line. You also want to start by adding a bit of fuel to cylinders that look like they're noisier than the rest, I saw this in 3 and 4, which in a longitudinal engine (or intake manifold at least), gets a bit more boost than 1 and 2. The way it was described to me, if your MAP reading 200, your cylinders might be seeing closer to 198, 199, 201, and 202 for cylinders 1,2,3 and 4 respectively. So I trimmed them with a percent or two of fuel before trying any ignition retard. These trims should be in the 1-3% range IMO, anything more and you really should be making global changes.


https://cimg1.ibsrv.net/gimg/www.mia...8108a47116.png

Just remember this is my first foray into knock tuning, with an un-tested BMW sensor and default MS3 basic settings, but I was getting reasonable data so I went with it. If my knock settings are too aggressive, I could be leaving a lot of power on the table. I had a TS gauge showing knock retard, and was aborting any run and making changes if it ever left "0".

Wow, I am running A LOT more timing on my car with 93 on my setup. Interesting