BRZ Dyno Testing from PERRIN

[PERRIN_Jeff]

For all you FR-S guys, replace the letters B, R, and Z with F, R, and S and then this will all make sense.

There were exactly 69 miles on the ODO of our BRZ, and it was only two days old, then it was bolted to the dyno. This may be extreme to some but this is what we always do. The last time a new Subaru came out, we had it on the dyno with-in hours of owning it. This time was different as the car came in over Mothers day weekend.

The dyno we use for all our testing is our Dynapack AWD dyno. This is a type that you remove the wheels from the car and bolt on adapters which then fit into pods which do all the magic. Some people say they read high, some people say they estimate flywheel HP, but in the end, it doesn't matter. What matters is, what does it read stock, and what does it read after the modifications.

You can see from the below graph how consistent the runs were. Only one of them was different than the rest and even then it was only a few HP off. Looking at these numbers, 168 Whele HP isn't too far off from what i expected to see. If Subaru rates the engine at 200 HP this makes for a drivetrain loss of about 15%. Not terrible at all. What is interesting is that Subaru says 150ft-lbs of torque and normally we see wheel torque numbers very close to this, and in this case we did as well. We also are not running 93+ octane, like a lot of US, but rather 92 octane, which means there could be even more TQ on cars running 9oct.



So how does this compare to another dyno? Locally we have Cobb Tuning(Surgline) and they had a chance to dyno a BRZ recently as well. Their dyno is a Mustang Dyno, which typically read low compared to some. We have found our dyno's both read very similar. You can see from the two graphs, that both cars made very similar HP. (168 vs. 164)



From both our graphs you can see the large dip in torque. This has been up for some debate for a while. Its a 25ft-lbs loss right in the middle of things. There has been speculation that its from cam timing, or fueling or something else. There is still some mystery to this, but I think its done on purpose. Since most people drive in the 1000-4000 range during normal driving this creates a powerband that drives people to shift early and save fuel. Then you have those times where you are shifting at redline, and in these situations you are always about 5000RPM or higher. This falls right into that second hump in TQ and power. Can we tune this out? I say yes, but only time will tell.



So how does the BRZ run? The data logging parameters are very limited and at best not very accurate with our Innovate Datalogger. What we can see is Air fuel ratio from the dyno. The leaner air fuel ratio was expected, but not the super rich condition from 6500 on up. The 02 sensor is behind two cats and from what we have leaned over the years is that this causes the AFR (air fuel ratio) to leaner than it really is by 1 full point. Meaning if you see 12.0 AFR, its actaully 11.0.

The above graphs are in Lambda which is 14.7 multiplied by the Lambda number. So (.85) Lambda is about 12.5 AFR, but because its behind the cats, its actually 11.5. Which is a very rich AFR for a non turbo car. Especially an efficient direct injection engine like this. What is a safe AFR? There are some basic rules to follow. Under full throttle an engine can run very lean but not for very long as the heat can melt things. "Safe" could be a wide range of AFR. Safe for this engine could be that 10.0-11.5 range. We won't really know until we can start tuning the BRZ. Then we will find out how lean we can run before knock occurs.

What AFR makes the most power? 12.5 AFR is a good range to shoot for. 13.2AFR is a known best TQ AFR, but this is where things could run closer to the edge. Again, we really won't know until we can tune these.

Here is a screen shot of a Logworks run. I logged tons of things and there were a few that makes sense, but a bunch that don't. STFT is always changing under full throttle, but we found evidence that its not active as different MAF housing throw off the AFR. If STFT was actually working, it would compensate.



Here are some other comparisons of some other cars. First one is a 2.5RS. When you look at these two cars side by side, you will see that the BRZ will blow the doors off a 2.5RS. You then add the weight factor to it and you really have a much faster car with the BRZ.



Here is another 2.5RS, but this time its modified with cams, pistons, and tuning. Even then its still no match.



Lastly is a comparison of a MINI Cooper S and BRZ. The MINI's dyno very high for some reason. But you can see a car that has similar overall Wheel HP to the BRZ, but the MINI has way more torque. Boost is the answer here.



This is just a snippet of what is to come with this car. Be on the look out for more dyno results this week!

[slow_frs]

judging by the powerband vs afr i think its safe to assume that peak tq really wont be raised with just a tune as the afr is about right where it needs to be for peak tq unless timing can be added, also maybe the rich burn at high rpms is attributed to toyota DETUNNING the engine form 230 to 200 for emissions

[slow_frs]

either way looks like i need to hurry up n slap a NOS sniper kit on one

[Tradewind]

Good work, interested to see tuning methods/software and more

[Matador]

Re: Torque dip.

Some have speculated fueling. Could it be possible that this has something to do with the port/DI mix? I've noticed that dip in the dyno plots of the D4-S V6s and V8s as well. One probably wouldn't notice as these are much bigger displacement and more powerful engines.

2GR-FSE

[serialk11r]

Originally Posted by slow_frs

judging by the powerband vs afr i think its safe to assume that peak tq really wont be raised with just a tune as the afr is about right where it needs to be for peak tq unless timing can be added, also maybe the rich burn at high rpms is attributed to toyota DETUNNING the engine form 230 to 200 for emissions

Er, engines are not tested at high rpm for emissions I think. Running more rich is going to emit many many times more hydrocarbons anyways. Tuning isn't just about fuel anyways, there's also spark timing, and possibly fiddling with cam timing.

Originally Posted by Matador

Re: Torque dip.

Some have speculated fueling. Could it be possible that this has something to do with the port/DI mix? I've noticed that dip in the dyno plots of the D4-S V6s and V8s as well. One probably wouldn't notice as these are much bigger displacement and more powerful engines.

2GR-FSE

lol Matador I thought you'd be more on top of your s*** :P

Port injection is gone by 3000rpm (check out the pretty articles and diagrams arghx7 posted). Tons of motors have that torque dip in the middle, so I'd guess it has to do with the way manufacturers do their intake tuning. (my hypothesis is sorta like what Perrin said, a "dual powerband" with low rpm torque for better driveability)

[PERRIN_Jeff]

Originally Posted by slow_frs

judging by the powerband vs afr i think its safe to assume that peak tq really wont be raised with just a tune as the afr is about right where it needs to be for peak tq unless timing can be added, also maybe the rich burn at high rpms is attributed to toyota DETUNNING the engine form 230 to 200 for emissions

Peak TQ should be able to be raised up some. I highly doubt that Subaru/Toyota maximized timing, fueling and cam timing. There has to be some room on the table. If they didn't, there is still more redline to add!

[ahausheer]

[Tons of motors have that torque dip in the middle, so I'd guess it has to do with the way manufacturers do their intake tuning. (my hypothesis is sorta like what Perrin said, a "dual powerband" with low rpm torque for better driveability)[/QUOTE]

Yeah most toyotas and Lexus display similar dips due to intake tuning/variable intake switchover points. I dont know if the D4-S has a variable intake but if it does the dip cannot be tuned out with the stock intake. Google ''T-VIS''. Lexus uses a three stage system and some Toyota's use a two stage system. Note that T-VIS may not be the current name of the technology but variable intake lengths/volumes are still used. Without it you get one hump or the other but not both. I hope the dip is related to something else so that is can be tuned out but I doubt it.

[ChristosG]

Originally Posted by PERRIN_Jeff

What is interesting is that Subaru says 170ft-lbs of torque and normally we see wheel torque numbers very close to this, but in this case, we do not. We also are not running 93+ octane, like a lot of US, but rather 92 octane. That could have something to do with it.

Actually it's rated for 150 lb-ft so it doesn't have much of a loss at all.

[PERRIN_Jeff]

Originally Posted by ahausheer

[Tons of motors have that torque dip in the middle, so I'd guess it has to do with the way manufacturers do their intake tuning. (my hypothesis is sorta like what Perrin said, a "dual powerband" with low rpm torque for better driveability)

Yeah most toyotas and Lexus display similar dips due to intake tuning/variable intake switchover points. I dont know if the D4-S has a variable intake but if it does the dip cannot be tuned out with the stock intake. Google ''T-VIS''. Lexus uses a three stage system and some Toyota's use a two stage system. Note that T-VIS may not be the current name of the technology but variable intake lengths/volumes are still used. Without it you get one hump or the other but not both. I hope the dip is related to something else so that is can be tuned out but I doubt it.[/QUOTE]
There is no variable intake track on this engine. There isn't even TGV housings like on the 2012 Impreza or all other turbo Subarus. Its too bad i can't log more data, or we could see what is causing this.

Originally Posted by ChristosG

Actually it's rated for 150 lb-ft so it doesn't have much of a loss at all.

Ooops! Not sure where I got that from last night. So then that makes a lot more sense.

[Matador]

Originally Posted by serialk11r

lol Matador I thought you'd be more on top of your s*** :P

Port injection is gone by 3000rpm (check out the pretty articles and diagrams arghx7 posted). Tons of motors have that torque dip in the middle, so I'd guess it has to do with the way manufacturers do their intake tuning. (my hypothesis is sorta like what Perrin said, a "dual powerband" with low rpm torque for better driveability)

**bends head in shame** the day job really is getting the better of me these days.

[serialk11r]

Originally Posted by ahausheer

Yeah most toyotas and Lexus display similar dips due to intake tuning/variable intake switchover points. I dont know if the D4-S has a variable intake but if it does the dip cannot be tuned out with the stock intake.

No variable intake, I think this is how Subaru motors usually go. I believe the primary reason is that for the engine to make any kind of power high in the rpm range, they need very long duration cams, which is super bad for low rpm torque. So more specifically my hypothesis is that they tuned the intake for low rpm torque so people wouldn't complain excessively about "lack of torque", but this in turn reduces high rpm torque, which is why we end up with a relatively modest amount of power.

As Dimman likes to point out, the 3S-GE can make 245Nm (122.5Nm/L, more than a Ferrari 458!) with cams, ITBs (with correctly tuned runner lengths), and 13:1 compression pistons. That TRD motor also has completely laughable low end torque. The FA20 has the advantage of direct injection, and a better combustion chamber design, so it's possible that putting on an intake that moves the peak volumetric efficiency up to 6000+rpm can give a decent increase in peak power at the expense of low end torque. Of course the people who know what's going on with the cams and stuff haven't said anything yet, and it could turn out that the cams have some compromises of their own, but I think this is as good of an explanation as we can give without knowing more details.

[arghx7]

Somebody hook up a universal OBD scanner and look at the estimated catalyst temperature and commanded AFR. I bet it's keeping catalyst temp near or below 850C.

[Dimman]

Originally Posted by serialk11r

Er, engines are not tested at high rpm for emissions I think. Running more rich is going to emit many many times more hydrocarbons anyways. Tuning isn't just about fuel anyways, there's also spark timing, and possibly fiddling with cam timing.



lol Matador I thought you'd be more on top of your s*** :P

Port injection is gone by 3000rpm (check out the pretty articles and diagrams arghx7 posted). Tons of motors have that torque dip in the middle, so I'd guess it has to do with the way manufacturers do their intake tuning. (my hypothesis is sorta like what Perrin said, a "dual powerband" with low rpm torque for better driveability)

I said that months ago...

[Harlequin]

Could it be noise emission relevant?

In EU, u did need to pass a noise emission test an FULL throttle from 50km/h (forgot which gear)

The Cayman S has the same "problem" when accelerating from 50km/h.

[PERRIN_Jeff]

Interesting thought for sure. The dip could help control noise/drone. But its funny because the car is so quiet to start with, it needs some sound!

[ajc209]

Its worth mentionig that the torque numbers for the mini are for a 1:1 gear ratio that will hit 90 something MPH, whereas with the BRZ it will hit 130MPH in 5th gear (1:1) because it will rev to 7450RPM.

My point is that the mini whould have to shift into an overdrive gear to be able to get to 130MPH and the wheel torque will go down a bit.

If we where to look at the wheel torque of both cars at say 125MPH in the highest gear possible it should be quite similar. Since torque and RPM of the wheel is what makes up WHP.

[PERRIN_Jeff]

Not sure what you are trying to say. The TQ readings on they dyno are comparable regardless of the gear ratio. For instance the wheel TQ (at the hubs) is very high and the gear ratio that we input on the dyno takes this wheel TQ and devides it by the gear ratio. This comes up with Engine TQ.

The only thing that the dyno graph shows is how much more TQ the MINI makes during normal day to day driving. For redline type shifting the MINI and BRZ are very close in their TQ and HP being delivered to the ground. Which is why their 0-60 and 1/4mile times are close(MINI 6.9/15.0, BRZ 7.3/15.3). STill the MINI is faster because its got more HP and TQ to the wheels.

Trust me, I wish the BRZ had the MINI Kind of TQ, but it just doesn't....yet

[trackmagic]

What are you programming with? Did you already crack the ECU or are you using something else?

[PERRIN_Jeff]

No, we have not cracked the ECU (we will leave that to Cobb Tuning). So there is no way to tune these at this point. I should say there is no way to safely tune these

[slow_frs]

Originally Posted by PERRIN_Jeff

No, we have not cracked the ECU (we will leave that to Cobb Tuning). So there is no way to tune these at this point. I should say there is no way to safely tune these

Pssh you want leaner burn and more timing, no problem my friend enlarge maf housing , forget the CEL for maf out of range, poor mans tune

[Scooby South]

Originally Posted by PERRIN_Jeff

Peak TQ should be able to be raised up some. I highly doubt that Subaru/Toyota maximized timing, fueling and cam timing. There has to be some room on the table. If they didn't, there is still more redline to add!

Thanks for the work Jeff..
Can you post up an 08 STi with Dual AVCS... The same dip was apparent on the STi about the same rpm level due to the AVCS Crossover. Which for the most part can be tuned out...Once there is some Software that will allow tuning..I am confident there will be some major Gains to be made with some timing, some fuel pressure adjustments, and some RPM extension.. I have invited some other people to look at this and hopefully, collectively as a group, we can tackle this.. along with COBB, Crawford and Topspeed Perfomance..IMO this is going to be the Achilles heel of this thing if someone doesn't find a fix... A flatter Curve would be wonderful...

Agian thanks for the work your putting in on this thing.. For this thing to be successful..its going to take the whole community ...not just one or two..

Bill