EFR7163: Midnight in Moscow - is sunshine in LA

[Jackis]

Hello!

I would like to describe our experience with bolting and tuning EFR7163 ms6 "bolt-on" kit.
The car was not mine - one of my fellows in Moscow. I have even never seen it in real life, only on photos.

Car is true-red Mazda 6 MPS (aka MazdaSpeed6 for oztralia and europe).
Owner name is George, and he owns this car for quite a long time.
Mazda 6 MPS ?? EFR 7163 "????"

Funny began this winter with rod coming out of the block.





I had Carillos Pro-SA rods laying here with me, but with 22mm wrist pin. I bored and honed them for stock pistons, George got fresh short-block with Gen2 pistons, put carillos on and assembled.


George assembled motor on fresh short-block, carillo rods and stock pistons, put new chinese CHRA in the stock turbo, cpe injector seals, EGR was shut:





I was tuning him at summer on stock-turbo, but chinese CHRA after 3k miles started to blow oil as usual, and George decided to fix it completely... Digged here and there and ordered Full-Race EFR7163 kit. Wise guy!






Well concerning bolt-on... Instructions came were for MS3 Gen2, which is completely different. George was using @Alpha's build for reference purposes. Information about canister and which is pre-installed was just absent. Crank-case ventilation was not intended, and George welded it into inlet for proper operation.
Everybody has its own opinion, but me presonally, I prefer having vacuum in inlet under boost to suck vapors and free up excessive crank pressure, this is what is intended to and this is how I want it to work.
Turbo-inlet was absent at all. Guys fabricated it theirselves. They also atatched air-return from EBCS to have all air metered.
Intake pipe with MAF had no fastener and no PCV recirc (also to have all air metered and to relief PCV pressure under boost).
FMIC pipes had no fasteners at all, and George had to wled them on, to bolt them to engine where appropriate. Not having them for such a price is ridiculous from our point of view.

All in all, it was installed and now in the process of tuning. I am using VersaTuner and tuning George remotely.
Several tunes iterations are listed below
[img][/img]

So far we are having slow spool (I guess because of not very proper wg preload) and strange knocks at 5000-6500 even with kiddy ignition, which could be VVT or inproper timing issue.

[GartzSpeed]

If you're gonna blow, blow large.

Well done sir, good look on the tune. Will this be the 2nd Speed 6 with the EFR 7163?

[predapio]

In Russia Turrrbo bigger than car.

[Jackis]

Tune has only begun. I will share it now from home PC.
We are tackling slow boost rise issue (22psi boost on 3600 only, flooring her at 2600-2700rpm) and strange annoying knocks, and still trying to make flat AFR (I am targeting 11.47 there, where it waves like Michael Phelps)

[Jackis]

Here is current tune with very very low timing and (I hope) final MAF curve.
I had to add a .rar extension to make versa file load.

[mituc]

How does a log look like with this tune on that car? Also, isn't he using a 3BAR MAP sensor?

Later edit: The tune is very strange... a lot of the soft limits are equal to the hard limits (e.g. throttle close versus fuel cut, high bat versus low bat, and so on), the ignition retard tables seem to have been scaled linearly, the load targets will result into boost way above what the factory MAP sensor can read (at least the calibrations are for the factory MAP), the load correction tables seem to be stock while I'm pretty sure an aftermarket EBCS is used (not that it matters since the load targets are so high and the compensations applied will always be from the positive side of those vectors) and so on.

[Jackis]

Now load limits is quite raised up and if you can see - I have forced WGDC increase percentage to zero. So it aint trying to raise boost if target is not hit, but lowers if it it has happened.
Ignition tables were interpolated in low-mid load range just to have smooth angles (I like when all is smooth). Upper load region is hardly set to manual values to have defined ignition there.
Soft limits are equal to hard limits because MAP sensor is stock so far and I just moved all the cuts away for now (out of 255 kpa).
High bat and low bat tables are the same because... well because fuck the high bat! One should keep its bat as low as possible lol. There is again bat/ect retard table which is applied always, so me personally I do not think we should decrease load output, when timing retard is applied already.
Seriously, that is so as for now, while we are searching for load targets we could achieve vis boost-tune. Then I move them to Desired Loads and each HighBat table will be decreased a bit. But while we are searching for boost and angles - I see no reason in polishing those tables.

If I have not answered some questions - feel free to ask more.

UPD: attached log

[Enki]

6 counts of knock with 5 degrees timing at 6500 rpm? Something is definitely not right here.
Can the owner hear it ping or hear some other rattling? Does this car still have balance shafts?

[Jackis]

Yes shafts are presented. No no unusual sounds. But he has cpe exhaust, and on 1.55 boost it is like a jet, so I bet he would not hear even nuclear boom...
Thanks for your help Enki!

UPD: he has just checked the ignition and claims that after locking crank into TDC, they see the shafts are parallel, and the crank sensor is against 20th tooth.

[Enki]

I wonder if he has hotspots on the piston or something then, or the wrong kind of spark plugs. When was the last time he checked the plugs or scoped the cylinders to check for true detonation? Can he get a gallon or two of alcohol in the tank?

[Jackis]

Last time he put some ethanol in the tank - he broke his rod LOL
But we definitely will be looking inside the cylinders and I will tel him to put some colder plugs. He is using stock NGK now.

[mituc]

Colder plugs may help indeed, at 360+ g/s of air the stock heat range plugs need some kind of alcohol (ethanol or methanol, whatever).

So now after you explained I understand what you're doing. Boost tuning with Versatuner should work with your approach (0-ed compensation tables, load set high enough not to become an issue, and focus on the WG tables). This is how Steve/Ugnius explained at some point that boost tuning would work with the public version of Versatuner, and I've also had a tune with a similar approach at some point.

[Jackis]

Yes, the plugs are on their way now. Should try ITV22 regapped to 0.7mm (0.027 inches). I hope they had addressed theire insulator issue.

Yes, @mituc, that is what I am doing eactly with any car (dont matter cobb or versa I use) I tune, just to dial loads based on flat or raising boost curve. I hope it was called hybrid tuning, but final tune is totally load-based. The car is going directly by WGDC, so I directly get the loads it achieves with full throttle for any RPM. When we are fine with boost curve - I put it into Requested load tables, and put some flat limiter lines for 1st and 2nd gear for Speed3 and put no any to Speed6, because race cars LOL

[mituc]

You should gap them to 0.65mm (0.026").

With so much boost and a built motor maybe you should get the NGKs simply because of the excess of ceramic material which should help transferring more heat.

The ITV22s had that insulator issue for only one batch or a limited amount of batches AFAIK. Also, the insulator cracking issue was occurring in the days when the Mazdaspeed community was looking at the HPFP upgrade only as a nice to have, not a necessity.

[Jackis]

Yes I tried to make him get those NGK, but they are hard to get here in Russia, and Densos are in-stock everywhere. So we tried ITV22 and gapped them to 0.7mm as for the beginning. I decided to start with 0.7 because it is much easier to do 0.65 afterwards, than go back from smaller to bigger gap.

I have made timing even more kiddy in all the range (got 3 degrees from the table in higher ranges). 10 up-top, thats just disaster...
I did separate-segment maf calibration for it.
I left stock VVT values from 1.25 load up to 5500, and then zero.

And we have got much cleaner log, without huge knocks. Atrtaching the log and the graph also (comparing to previous knocky one).

[Enki]

There's still knock present, but it is significantly less. Only the spark plugs were changed for this result?

[Jackis]

Not only.
Ignition was moved 2-3 degrees even less, spark plugs were changed, VVT was moved to stock. As I had 3 things which alters combustion, I am not sure which one was the main one. Oh, and I did very tricky maf_cal (splitting of several pieces of exponential trend-lines, damned excel, usually 1 power function covers the whole maf voltage range, here I had to use 3 different piece-wise functions)
But at least now we can start adding timing and see how it goes.

[Enki]

Well the car is hitting higher timing than it was now, so that's good.

[Jackis]

Absolutely, sir

[Vadim aka ScooteR]

will he beat me next season? )

[mituc]

Originally Posted by Vadim aka ScooteR will he beat me next season? )

Technically that's possible but it's proven by now that your car is a bit on the extreme side of the ridiculous

[Vadim aka ScooteR]

Originally Posted by mituc Technically that's possible but it's proven by now that your car is a bit on the extreme side of the ridiculous

There are several MS that technically can be faster then i)

And that real whp of this red MS with my last run
they took wrong size of tires(((

[mazdakaps]

Nice build. Did you log your injector duty cycles?