K04 AMB Head vs K0R AEB Head Dyno Graph

[tjb616]

"AEB head won't help on a small turbo."
"AEB head just moves power right, no point with a K04"
"There is no difference in actual power between a K04 and a K0R"

All quotes I read many, many, times, before ignoring them all.

My Ultrasport, 2 dyno graphs overlaid.

Same dyno (Mustang, has "The Heartbreaker" in vinyl on the side, reads 5% lower than neighboring Mustangs)
Same tuner (Roshan @ Powerhaus)
Actually 1 calendar day apart in 2014 and 2016, so no astrological variables ;)
Run 1: 50 degree ambient temp on K04/AMB/TT225 map.
Run 2: 55 degree ambient temp on K0R/AEB/440cc map.

All other mods remained the same:

Autospeed FMIC
Turbo Inlet
FSI Coil Packs (with the popular NGK plugs, forget the part number)
Test Pipe to APR Stealth Turbo-back
Stock Intake with K&N Drop-In
TT225 MAF
I did switch from a Kentic manifold on the K04 to the supplied manifold with the K0R.
19.5PSI tapering to 16-17PSI on both tunes.
93 Sunoco (same gas station, actually) no Meth.

Notes:

I a crack formed in my AMB head. Found an ultrasonically cleaned and freshly rebuild AEB head and jumped on it.
Kept VVT.
Had already deleted SAI on the first K04 tune.
The dip in power around 5.2k-5.6k is likely (according to the tuner) a result of poor intake flow due to a VERY poor gasket match by my engine builder. Too wide, not tall enough, and only goes 1/4" into the runner. It will be replaced with a properly ported AMB mainfold (ISO, if anyone is selling.)
Fresh piston rings and bearings, both OEM, were installed when the head was replaced.
The motor had strong compression and burned no oil (nobody will believe this... but that's okay with me) but it the bearings were found to be barely hanging on when inspected.
Stock rods (The idea was that I have a GT28 B7 and a stage 3 B5 S4, this is my beater, it's never getting a bigger turbo without a motor swap, but as you'll see below, I'm starting to regret not buying rods...)
Tune was after a 1k mile break-in on the new rings.

Draw your own conclusions, but I was hoping to pick up 10-15whp. The results:

Before: 164whp / 185wtwq
After: 209whp / 214wtq

Exact gains: 43.6whp / 28.5wtq

What really impressed me, is that I didn't lose a single tick of power down low. There isn't a mm on the graph where car pick didn't pick up power under the curve.

Massive thank you to Roshan @ Powerhuas for the tune, Dave @ Brother to Brother for building a strong motor, and RAI for a great product (no affiliation, although I do hope he sees this thread.) This car (my work beater, takes 500 miles a week) is now actually really, really, fun to drive!

[7speed]

That's pretty impressive.

[texasboy21]

What is a K0R turbo?

[Matt Devo]

That's pretty impressive.

yes, but it's also clear that the before engine with the AMB head was massively under-performing. 164awhp is low even for a 1+ tune, let alone a K04. I put down ~200whp with a crappy Neuspeed tune and 318cc injectors a decade ago, on a similar Mustang dyno. And while I know you can't directly compare, there's just no way that before graph is representative of a healthy engine with a properly tuned K04-015.

[7speed]

This is also true

[tjb616]

What is a K0R turbo?

RAI's hybrid K04.

yes, but it's also clear that the before engine with the AMB head was massively under-performing. 164awhp is low even for a 1+ tune, let alone a K04. I put down ~200whp with a crappy Neuspeed tune and 318cc injectors a decade ago, on a similar Mustang dyno. And while I know you can't directly compare, there's just no way that before graph is representative of a healthy engine with a properly tuned K04-015.

You don't have to believe it is, but it is indeed a fact. A 165whp car on this dyno puts down 205 on a Dynamics a couple towns over, and I would say that's right in line with K04 AMB car running 19 peak psi.

To give you an idea, a 991 911 Turbo S with bolt-ons and a tune put down 480whp on this dyno in low February temps. Thats a LOW 10 car all day long.

But like you said, you can't directly compare.

[CaptainJohn]

The HP increase is a little deceptive. The first run didn't go to the same RPM, which is where the last ~10hp were made.

Thank you for sharing. I'm not sure exactly what this means (differences in turbo, engine condition, etc), but, I'm excited having recently installed an AEB head on my A4.

[Doug@FrankenTurbo]

In our testing I came to the same conclusion. The "torque-loss" of a wide-port cylinder head is a red herring. On a small turbo that knows how to spool, that AEB head is a nice add.

[diagnosticator]

The HP and torque numbers stated vs the graphed results are not consistent, there is a significant difference between the stated HP and torque numbers and the graphed peak values achieved

For example, the "After" stated maximum HP is 209 HP, but the graphed HP curve maximum value is ~ 190 HP.

Both of the "After" HP and torque values graphed above 5270> RPM are lower than the "Before" values, therefore, the statement "There isn't a mm on the graph where car pick didn't pick up power under the curve", is not entirely accurate. This is evident since each partial region of the total area under the curve is defined by the lower and upper boundaries by vertical lines from the curve to the x axis. For the area region from 5270> RPM , Before is > than After.

[Kevin C]

The HP and torque numbers stated vs the graphed results are not consistent, there is a significant difference between the stated HP and torque numbers and the graphed peak values achieved

For example, the "After" stated maximum HP is 209 HP, but the graphed HP curve maximum value is ~ 190 HP.

Both of the "After" HP and torque values graphed above 5270> RPM are lower than the "Before" values, therefore, the statement "There isn't a mm on the graph where car pick didn't pick up power under the curve", is not entirely accurate. This is evident since each partial region of the total area under the curve is defined by the lower and upper boundaries by vertical lines from the curve to the x axis. For the area region from 5270> RPM , Before is > than After.

My quick take is that the colors are reversed on the graph for the "after" results ( good catch!). On the after, blue is the HP and red is torque. Each data set intersects at 5250 (where mathematically HP and torque are the same). Below 5250, mathematically, torque must be higher.

[diagnosticator]

My quick take is that the colors are reversed on the graph for the "after" results ( good catch!). On the after, blue is the HP and red is torque. Each data set intersects at 5250 (where mathematically HP and torque are the same). Below 5250, mathematically, torque must be higher.

I considered that possibility, but the values stated still are not consistent with the graph.

[Kevin C]

I considered that possibility, but the values stated still are not consistent with the graph.

This is how I read it:

35 hp per division... On the after test, the graph line is about 1/3 of a block above the 195 hp line, 35/3 = 12.

195+12 = 207 HP, the tests states its 209. Seems right to me.

Same with the before reading, that's about 1/8 of a block above the last line. 35/8 = ~ 4 the last line is 160. That works out to 164.

That said, 35 HP per division is kind of an odd increment ( no pun intended).

[diagnosticator]

This is how I read it:

35 hp per division... On the after test, the graph line is about 1/3 of a block above the 195 hp line, 35/3 = 12.

195+12 = 207 HP, the tests states its 209. Seems right to me.

Same with the before reading, that's about 1/8 of a block above the last line. 35/8 = ~ 4 the last line is 160. That works out to 164.

That said, 35 HP per division is kind of an odd increment ( no pun intended).

It is a difficult graph to interpret. Graphed data should be easier to read. I meant to critique the graph, not the numbers.

[tjb616]

It is a difficult graph to interpret. Graphed data should be easier to read. I meant to critique the graph, not the numbers.

Honestly, it took me about 15 minutes too long to make the OP because I kept confusing the values. I agree, it's hard to follow.