power loss of K8 ECU & KL-DE?

[masonholmes]

Anybody got ballpark figures on power loss from utilizing the K8 ECU on a KL-DE? (KL02 VAF)
Thanks,

[lakersfan1]

I wonder this also if anyone's dynoed thier car with the K8ECU/KL02 setup, then tried the KLZE ECU/JE50 setup on the dyno with no other changes?

[veltpak6]

come on! someone holler

[hgallegos915]

wouldnt u run lean with the k8 ecu at high revs this destroying the engine quietly and internally? just a rumor i heard.

[veltpak6]

wouldnt u run lean with the k8 ecu at high revs this destroying the engine quietly and internally? just a rumor i heard.

someone else said it was too rich

[hgallegos915]

hmm.what to believe.. I tought the k8 injectors where rated less than those of the millenia.. can someone please carify..also can someone clarify that engines crap out when running lean please?

[Tunes67]

You can damage your engine from running too lean. When you run to lean of a fuel/air mixture it causes a much hotter explosion in the combustion chamber. This can lead to detonation (knocking) and burned valves if allowed to happen for a long enough period of time. So yes.. running to lean can eventually kill your engine.. but it wont happen over night either. Conversely.. running to rich of a mixture isnt good either.. though arguably it is a bit easier on your pistons and valves. But with too rich of a mixture.. you can get excessive carbon build up on the piston and valves.. this leads to hot spots in the combustion chamber and can cause pre-ignition (Pinging). Pre-ignition is hard on your valves and pistons and can wear on them more than under normal conditions. Extreme pre-ignition can even bend valves.

Tunes67

[tatsu]

I doubt anyone has actually dyno'd the two ECUs back to back on the same motor. When I do my swap, I'll try to get some numbers using a G-Tech because dyno time is too expensive!!! Unfortunately, that probably won't be until about a year from now - until then, here is another long-a-- brain-dump:

Basically, my understanding is that the ECU has fuel and timing maps that tell it for a given rpm value and throttle position the air/fuel ratio should be X, and the ignition timing should be Y.

With regards to ignition timing, this can be dynamically retarded by feeback from the knock sensor, but it won't advance beyond the base setting. In other words, if it's pinging, the ECU pulls back timing - if it's not, it goes back to the base timing map.

With regards to fuel mapping, the ECU itself does not directly measure how much air and how much fuel is being delivered. In reality, for each point on the fuel map the ECU simply instructs the injectors to open for a certain amount of time based on the input voltage from the VAF. This is adjusted on an ongoing basis by feedback from the O2 sensor(s).

My theory is that all of the Otto-cycle K-series motors' ECUs are generally close in their fuel mapping with slight differences for fine tuning and recommended fuel grades, and that the major difference is found in the calibration of each model's VAF to work with the correct sized injectors for that model. I believe that this is done by sending a correspondingly higher voltage (I.E. a "lower" reading, because the voltage from the VAF drops off as airflow increases) to the ECU for a given amount of airflow to compensate for the larger injector size.

In other words, a KL VAF is sending a "lower" reading to the ECU for a given amount of airflow so that the ECU opens the larger KL injectors for a shorter time so that they deliver the appropriate amount of fuel. Conversely, the K8 VAF delivers a "higher" reading to the ECU for a given amount of airflow because the K8 injectors need to be open longer to deliver the same amount of fuel.

If I am correct in my assumptions (I have to assume, because it seems nobody has done any instrumented testing on the different combos of VAF/ECU/injectors yet), as long as you are running the injectors to match the VAF you have, your mixture should be generally OK, although the tuning would not be optimal if you don't have the correct ECU.

This makes sense when you take note of the fact that when a KL VAF is installed on a K8 motor with K8 ECU and injectors (like what Corksport tried) you lose power all around, but a KL (even a ZE) with KL VAF and injectors seems to run OK (but not optimal) when a K8 ECU is used.

Getting back off of the VAF/injector tangent, aside from the differences in the fuel and ignition timing maps between the ECUs, which I believe to be minor enough not to kill your motor but may affect power, there are differences in VRIS activation points that definitely WILL affect power. This is true between the 1.8L, 2.0L and 2.5L motors, and also between the DE and ZE motors - the ZE's run somewhat leaner fuel maps and more aggressive ignition timing because of their "high-performance" orientation (and Japanese premium gas) versus the everyday durability orientation of the DE, which was designed to run on regular unleaded. All five types (K8, KF, KF-ZE, KL, KL-ZE) of motor will use different VRIS activation points to optimize their performance based on displacement and cams.

In summary, (again) IF my assumptions are correct, there are three key points with regards to VAF/injector/ECU combos:

First, DO NOT mix and match VAF and injectors from the different-sized motors. With a K8 VAF and KL injectors, you will run very rich because the bigger injectors are being told to stay open too long. With a KL VAF and K8 injectors, you will run very lean because the smaller injectors are not being told to stay open long enough. Treat the VAF and injectors as a combo, and always stay with ones from the same size motor, if not from the same motor.

Second, and this should be obvious, DO NOT run the K8 VAF/injector combo on a 2.5L. The injectors are not sized properly to deliver enough fuel for the larger motor, particularly if you are running a ZE or have a tuned DE. You should be able to run the KL VAF/injector combo on a 1.8L with no problem.

Third, you can most likely run any of the ECU's safely with any of the motors as long as you are using the correct VAF and injectors, but each motor will perform best with its matching ECU, or even better a dyno-tuned, custom-programmed ECU.

[veltpak6]

... not to kill your motor but may affect power, there are differences in VRIS activation points that definitely WILL affect power

This is exactly what i'm experimenting with. I have a KLDE with ZE intake mani and (half the) ZE cams... actually KL01 cams on the rear and KL31 cams on the front :p and the KL01 computer seems to run it just fine, EXCEPT for the VRIS points... I researched the K8 Vris points and for the V1 & V2 opening, theyre in more ZEish spots... the top end closing point is a little too early (6300) where it should be 6900-6900 ideally but i can live with that... oh also the K8 ecu opens the V1 & V2 between 1900 and 2800 (how weird is that! the K8 probably benefits from that i guess....) but on the KL's its not a great idea... but thats just b/w 1900-2800... i dont really care too much down there...

[tatsu]

It appears that there is a significant exception to my VAF/injector combo theory - the program in the KL-ZE ECU apparently was designed to use the JE-50 VAF (same as the 1.8L MX-3) with the larger KL injectors (which I believe to be the same between the KL-DE and KL-ZE).

Regarding the VRIS activation points, the following pictures and info will more clearly illustrate the difference. This is from the factory service manual for the K8:

• 

It's kind of hard to differentiate between the lines, but if you basically follow the highest line on the graph at all times, you'll see the torque curve for the K8 using the following VRIS activation points:

• 0-1,900 rpm (VRIS 1 and 2 closed)
  
  1,900-2,600 rpm (VRIS 1 and 2 open)
  
  2,600-3,900 rpm (VRIS 1 and 2 closed)
  
  3,900-4,700 rpm (VRIS 1 open, VRIS 2 closed)
  
  4,700-6,300 rpm (VRIS 1 and 2 open)
  
  6,300+ rpm (VRIS 1 and 2 closed)

From the Probe/MX-6 FAQ web page, here are the VRIS activation points for the KL-DE:

• 0-3,250 rpm (VRIS 1 and 2 closed)
  
  3,250-4,250 rpm (VRIS 1 open, VRIS 2 closed)
  
  4,250-6,250 rpm (VRIS 1 and 2 open)
  
  6,250+ rpm (VRIS 1 and 2 closed)

And the matching graph of the KL torque curve from an SAE document on the K-series motors:

• 

You can deduce from the two graphs that using a K8 ECU on a KL would result in a loss of torque (and therefore horsepower) from 1,900-3,900 rpm, and again from 4,250-4,700 rpm, leaving aside any differences in fuel/timing maps.

You can likewise deduce from the two graphs that using a KL ECU on a K8 would result in a slight loss of torque/power from 1,900-2,600 rpm, a BIG loss from 3,250-3,900 rpm, and another loss from 4,250-4,700 rpm. Losing all that torque right in the mid-range would be bad news for the 1.8L!

I don't have data or graphs for the KL-ZE, but I imagine that non-optimal VRIS points would have a similarly dramatic effect on the powerband. One would also suspect that changing intake manifolds and/or cams would also require a change in VRIS points to optimize performance.

If you have modified your motor and want to optimize your VRIS activation points, you would ideally do three dyno runs - one with both open, one with both closed, and one with only VRIS 1 open (VRIS 2 by itself does nothing beneficial). You would then overlay the resulting torque curves and program the switchovers to always keep the torque curve as high as possible, just like Mazda did right from the factory!

Bearing all that in mind, here are the possible combos that should be generally OK to run for fuel delivery, and would have the correct fuel/timing maps and VRIS activation points by virtue of using the correct ECU (duh!):

• K8 motor + K8 ECU + K8 or ZE (JE-50) VAF + K8 injectors
  
  K8 motor + K8 ECU + KL VAF + KL injectors
  
  KL-DE motor + KL ECU + KL VAF + KL injectors (not recommended for KL-ZE due to different VRIS activation points and probably different fuel/timing maps in the higher rpm ranges, although many people have apparently done it without any reliability issues)
  
  KL-ZE motor + KL-ZE ECU + K8 or KL-ZE (JE-50) VAF + KL injectors (optimal combination for KL-ZE)

Incidentally, this would also suggest that the stock VAF is not a significant bottleneck for the K8, since the same one is used on a motor with 50% more power and 39% more displacement.

Hope all of this helps!

[veltpak6]

You can deduce from the two graphs that using a K8 ECU on a KL would result in a loss of torque (and therefore horsepower) from 1,900-3,900 rpm, and again from 4,250-4,700 rpm, leaving aside any differences in fuel/timing maps.

again, good detailed info! The KL-ZE has different VRIS points than the KL-DE because of its different intake manifold. I have a KLZE intake mani on my KLDE and the KL01 computer runs the engine fine, it just switches the VRIS too early, so I'm alread losing torque... KLZE VRIS points are basically perfectly matched to the K8 VRIS points with the exception of the 1900-2800 spot, and the 6300rpm closing , although its better than the KLDE's 6250 closing. ZI IM requires at least 68001&2 closing to make the most top end torque... but aside from those two the K8 has perfect VRIS points for the ZE....if the fueling on a KL is good with kl02 VAF, KLinjectors and K8 ecu combo, as some people have reported then its something to try out