Super-Cheater K04

[ConeKiller]

Ported K04-882 (Mazdaspeed) Turbine Housing
-Heavily ported turbine entry housing & wastegate. No porting past the first bend of the housing.
-Bored and radius'ed for larger turbine wheel
-Wastegaste bored 2mm over.

Ported K04-582 (CX7) Compressor Housing
-Light porting of whole A/R.
-Bored and radius'ed by Turbolab for larger billet compressor wheel

Turbine Wheel
-Mamba "Extreme" K04 Turbine
-Clipped 10 degrees and balanced with comp wheel by Bullseye Power
K04 Extreme Turbo Turbine Wheel Borg Warner K03 K04 AUDI VW

Compressor Wheel
-Turbolab billet wheel

Bottom line:

Why are you doing this, when you could buy an atp or S3?
I wanted to have fun with non-stock turbos, but be obviously breaking stock/mod class rules of AX/rallycross based on visual inspection.

Will it work?
I don't know, let's find out.

Well, if it works, how much power will it make?
I guess- IF it works, it might make somewhere between an S2 and a S3, just based on the size of the turbine wheel and compressor wheel.

Why did you use a CX7 Compressor housing- isn't a Mazdaspeed housing better?
I bought a used turbolab modified turbo to start this project that came with that housing, but as far as I can tell, the only major difference between the CX7/MS compressor housing is the inlet diameter (which was bored out for the larger compressor wheel anyways).


Reference:

 

[SharksInSpace]

Will it work?
I don't know, let's find out.

I like where this is going

 

[Redline]

Eager to see what power levels you can push.

I've actually considered buying a new (v2) S3 turbine housing from BNR, sending off the BNR to get reworked, maybe a new turbine wheel/porting, definitely porting WG, and getting a billet wheel in there. I'm sure I could get the 42-43 lbs/min closer (or maybe even above) 50 lbs/min. That'd make it a 450whp turbo easy, if my ricer math is correct.

 

[NavysailorMS3]

I dont think i have ever heard of someone doing this to a K04. Would be interesting to she the limits that can be pushed with it.

 

[VashEXE]

 

[VTMongoose]

Very cool stuff Robert, I hope this turns out well. I have thought many times about how to get a more ideal turbo on my car than the BNR S3 or the K04 (that isn't as expensive as an EFR). Since my S3 was stolen and I'm back on the K04 again, the weaknessess of both are really apparent. I really wish there was an option that slotted between these two. I asked Bryan about the S2 over the phone, but he said it's down about 60 horsepower from the S3. This puts it barely above the K04 in terms of power output - probably not worth it. I look forward to seeing your results.

 

[Easter Bunny]

I feel like someone on the other forum did something similar with a super ko4 and determined it wasnt cost effective. but he wasnt trying to cheat in which case a slight improvement over stock would be perfect.

 

[ConeKiller]

I feel like someone on the other forum did something similar with a super ko4 and determined it wasnt cost effective. but he wasnt trying to cheat in which case a slight improvement over stock would be perfect.

I found a few threads like this, and I agree with the idea. With the cost of used S3's these days, the time, effort and money invested in this can't beat a used S3- even if this modified turbo out preforms my expectations.

This is mostly an experiment with modification bits that I found on the cheap, that don't community doesn't have a lot of good information about.

Oh yeah. Turbo makes boost. Stock tune is stock tune though, and BCS systems are doing their job. ~14psi on 30-50% relative throttle and 30% wgdc. Plenty of room to grow.

 

[VTMongoose]

What kind of BCS setup are you using? Are you using the stock restrictor pill with bleed mode?

I should have ported the wastegate on my current K04 before installing it. I am getting 16 psi on spring pressure and the usual problems with boost spikes, which I almost forgot about while I had the BNR.

 

[ConeKiller]

What kind of BCS setup are you using? Are you using the stock restrictor pill with bleed mode?

I should have ported the wastegate on my current K04 before installing it. I am getting 16 psi on spring pressure and the usual problems with boost spikes, which I almost forgot about while I had the BNR.

My engine/tip/exhaust setup is stock, including wga, restriction pill, etc.
Actually, I'm not noticing boost spikes like I used to on stock k04, but no noticeable lag. Benefits of a ported WG housing?

 

[CorkSport - Derrick]

We used to do something similar with the VJ-14 turbos back in the day for the 323 GTX/GT. Clean up the castings in the exhaust housing, port the wastegate, install a little larger wheel from the Mx6 turbo VJ-20 and be good to 200 whp without boost creep. Going a little further port matching the manifold to the turbo was good for a small decrease in spool time as well.

-Derrick

 

[BAT-man]

What boost spikes do people get on the K04? I've been tuning one for 75k miles and never had a spike. The only time I had overboosting is when I hooked up the 3port EBCS wrong.
My hybrid turbo is similar to OPs, except that the exhaust side is a clipped stock K04, and the intake side is not ported on the outlet. Bosts fine, hitting g/s of 300 right now at 19psi. Was close to hitting 300 g/s with stock K04 and intake/dp though, so need to run more boost to see the true benefits.

 

[Redline]

What boost spikes do people get on the K04? I've been tuning one for 75k miles and never had a spike. The only time I had overboosting is when I hooked up the 3port EBCS wrong.
My hybrid turbo is similar to OPs, except that the exhaust side is a clipped stock K04, and the intake side is not ported on the outlet. Bosts fine, hitting g/s of 300 right now at 19psi. Was close to hitting 300 g/s with stock K04 and intake/dp though, so need to run more boost to see the true benefits.

None here, except when initial (spool-up) WGDC values and WG Dynamics values were too high. Once the tune was adjusted, the numerous K04s I tuned didn't have any spiking at all.

 

[Mauro_Penguin]

None here, except when initial (spool-up) WGDC values and WG Dynamics values were too high. Once the tune was adjusted, the numerous K04s I tuned didn't have any spiking at all.

But you are freek tuned?

 

[Redline]

But you are freek tuned?

Now I am. I wasn't for years.

 

[VTMongoose]

What boost spikes do people get on the K04? I've been tuning one for 75k miles and never had a spike. The only time I had overboosting is when I hooked up the 3port EBCS wrong.
My hybrid turbo is similar to OPs, except that the exhaust side is a clipped stock K04, and the intake side is not ported on the outlet. Bosts fine, hitting g/s of 300 right now at 19psi. Was close to hitting 300 g/s with stock K04 and intake/dp though, so need to run more boost to see the true benefits.

My original K04 would spike ~0.35-0.40 above my target calculated load on the reg, and that's with me lowering WGDC to the point where it could barely reach my targets in the 3000-3500 vicinity. My new one is way better, I tune it like I would any other turbo and it only misses by about 0.15. Same exact setup (I use the stock bleed mode EBCS config), only difference is I have a 3" CBE now compared to before. There really isn't any reason why the two turbos should be different. One thing I have thought about doing is enlarging the ID of the factory installed restrictor pill in the K04 compressor housing (in fact, my original restrictor pill being out of tolerance or even partially clogged with something may have been the source of my problems). However I am far too lazy to do this and I do not plan on keeping the K04 long term.

 

[ConeKiller]

periodic update:

Tuning adventure with the super-cheater k04 hit a temporary wall when I maxed out the stock airbox maf voltage (>4.9v) at ~2.3 load.

So I relocated the battery to the trunk and installed a 4" HTP intake.

In the process of which, I found this:

Caused by this:
(A lock washer that somehow made it into the airbox)

In the spirit of using a broken part justifying the expense of a new, more ridiculous car part, I bought this to replace the broken turbo labs wheel:

With these specs:

Inducer Dia.: 51.03 mm (vs 50mm Turbo Labs)
Extend Tapered Tip Exducer Dia : 66.13 mm vs (65mm Turbo Labs)
Exducer Dia. : 61.98 mm vs (60mm Turbo Labs)
Wheel: 7+7 (vs 6+6 Turbo Labs)

I bored out and polished the housing for the new wheel:

The turbine and compressor wheel are currently being balanced at R&S Distributing in minooka, il: hopefully it will be ready for pick up and install tomorrow.

In the meantime, I've played around with squeezing what I can from the stock mazdaspeed k04 as a baseline. Note that the AFR is actually Lb/min numbers. Only mods are: Autotech internals running E25-E30, and 4" intake (both OEM cats still there). 2.2 max load, tapering to 1.5 in the dyno below:

 

[Trithemius]

periodic update:

In the process of which, I found this:

Caused by this:
(A lock washer that somehow made it into the airbox)

That sucks. I did the exact same thing to my stock K04 but with a nut from my new shift cable setup. Amazingly the thing still makes stock boost pressure, it just whistles while doing it. Gotta get that fixed this spring.

 

[ConeKiller]

That sucks. I did the exact same thing to my stock K04 but with a nut from my new shift cable setup. Amazingly the thing still makes stock boost pressure, it just whistles while doing it. Gotta get that fixed this spring.

If you wait on replacement, plan on replacing the whole turbo, or at least fully rebuilding with new turbine and compressor wheel. When the compressor wheel gets screwed up, it's very likely to be out of balance as well- which does bad things to the turbine wheel shaft and journal bearing.

 

[ConeKiller]

Update:

Super-cheater K04 works great, and flows ~46lb/min or 4 volts on a 4" htp intake.

I've run into a tuning problem that I can't make heads or tails of.

In high 30lb/min & >3.5 maf volts the car has trouble keeping up and is doing some weird fuel trimming which results higher requested lambda/AFR's(see attached log).

I must be missing something, but I not sure what. I recreated my tune, based off a high load OTS map in versatuner, but still have the issue.