- skeurton’s Garage
Aug 26, 2013 | Views: 3,332
Filed under: /blogs/browse/t/vehicle/v//p
I use Cobb's AccessTuner and AccessPort to make all the necessary changes to the ECU mapping, as well as to record realtime data.
I also use Virtual Dyno (VD) as a tuning aid in order to quantify map changes and to see their effect on overall power output. VD uses the datalogs recorded with the AccessPort to generate a projected dyno plot. It's easy to be skeptical about the results that VD produces, but I've compared them to my real world results on a Dynojet, and VD (with its correction factor set to Dynojet) was within a 5% margin of error for horsepower and within 6% for torque. That's not bad at all. It's still important to best minimize the number of variables when recording datalogs for use with VD, so I stick to the same stretch of road for all my pulls.
So here we go....
Unfortunately, I did not think to get a datalog or put my car up on rollers while it was still stock. However, I've seen several dyno sheets for a stock 2011 Mazdaspeed 3, and I would place it at around 240HP / 259TQ.
I had quickly modified my way through all the stage 2 parts and then some (Ref: Mazdaspeed 3/6 Modification Guide 2006 to Current). At this point, I knew I was going to be upgrading to a bigger turbo eventually, so I figured it would be a good time to get the car up on the dyno. This particular map was targeting 19psi and 11.8AFR. Here are the results:
(Click on images for larger version)
For comparison and VD accuracy sake, I recorded datalogs while on the dyno of runs #2 through #5 above. To reitierate, VD was within a 5% margin of error for horsepower and within 6% for torque.
300HP was my goal for the stock turbo, and I just about got there. Now I was on a mission for 350HP.
In September 2012, I installed a BNR Stage 3 Turbo (click link for photo album). It is basically a gt2871r compressor with a custom housing and stock Mazda hotside. It's not exactly a BIG turbo, but it's good enough to get you to about 400HP.
I wanted to see what it could produce using only 93 octane fuel. At 22psi and 11.8AFR, the motor was severely knock limited in terms of timing. I was only able to advance 13° at redline.
The following VD plot is a comparison between two datalogs recorded on the same map --- one log is from right after I finalized the tune in September 2012, and the other is from July 2013 (10 months later). The tune held up well and I was pleased with the consistency.
The final step in my quest for power was to install an alcohol injection system. After doing much research, I settled on the Aquamist HFS-4 (click link for photo album).
I started out injecting a 50/50 mix of water/methanol. Charge temps lowered significantly, but I was not able to increase boost and timing the way I wanted to. So I stepped up to 100% methanol. This motor seems to like the octane boost it gets from straight methanol, while still maintaining really cool charge temps. Again, my end goal was to achieve 350HP. With max boost set to 26psi and timing advanced up to 19.5° at redline, I got a little more than I bargained for:
That's way more than I was shooting for, but I'll take it!
As far as power-adding modifications, I think I'm pretty much done. The next step will likely be to dial back the tune a bit until I can build the motor and replace the clutch.
Permanent Link to this Blog Post: