Thanks to mrfred, I got my jdm map sensor and went to town. I set everything up according to his how-to. I did some logging and noticed a couple of things. The sensor max's out at 33psi and it doesn't track exactly with the zt-2 map sensor. Here's a screenshot of a 1-4th gear pull from a stop (checking 0-60 time) showing both MAP sensors.



The zt-2 sensor is connected to the vacuum line that runs to the stock boost gauge (tee'd in behind the glovebox). The jdm map shows a 31psi spike in first @4400rpm while the zt-s shows 24.4 @6200. Goes to show that sensor response is greatly affected by location.

 

Did you set up the sensor correctly with your altitude and change it tothe right atmospheric pressure?

 

Yep, I'm on the CT coast, so I am only a couple of hundred ft above sea level at most. I used 14.6 for the correction, which shows as 0.0004psi with the key on, engine off. I'm attributing the difference to sensor placement. The JDM map will be more accurate since it is directly on the manifold, whereas the zt-2 map has a delay due to location.

 

I'm not so sure about this. The MAP sensor I use on the dyno is about 12-15ft away from where I T the line on the intake manifold. On my personal car with the zeitronix Map sensor I see approx .5-1 psi difference between it and the dyno map sensor.

7 psi difference makes me something else is at play here.

 

Not that I dispute your results, but like Bryan something else is going on here. My JDM is dead on with my Autometer that has also been crosschecked by an Isspro and Defi boost gauge.

 

Finally!! Someone else noticing the same thing!

My sensor tells me that I'm running 29psi on Cali 91 octane pump fuel.

I know it's not true, but couldn't really prove it without another MAP hooked up (and logged) at the same time. I posted about this a month ago, but nobody chimed in with supporting evidence (or comments).

Thank you so much for posting this. Now let's figure out the algorithm to make it log accurately.

 

Looking at mixmastermatt's boost log, I can see a pretty big lag in the response of the Zeitronix in the lower gears. In the higher gears, the lag is not so evident, and near the top of the last gear where the boost is not changing as rapidly with time, the Zeitronix and the JDM sensor are nearly the same.

It feels to me like a really really long boost hose or a leak somewhere along the boost hose.

What would be cool is if you could get another log where you roll up the boost through the high boost range very gradually so that there is sufficient time to allow the Zeitronix to track with the JDM MAP sensor.

 

I will be installing my JDM sensor this week.

Next time I get my personal car on the dyno I can cross reference the JDM MAP, the dyno MAP, the zietronix MAP and my greddy boost gauge.

Maybe if I average out all 4 I can get accurate boost.

 

well, a good test would be is to disconnect the boost controller, and run a hose from the turbo IC pipe straight into the WGA. you should see 11psi.

 

Guys, please take into account what his boost curve would *reasonably* look like. To me, it looks like the non-JDM MAP sensor is tracing a logical curve for his mods on the stock turbo. His JDM curves look exactly like mine on 91 octane and my GM 3Bar spit out curves much like his non-JDM sensor. It's just not reasonable to assume that he's hitting 32psi on the stock turbo using pump fuel, nor is it reasonable to assume that I'm getting away with 29psi on straight Cali 91 with no knock.

It seems to me that there's a calculation routine error here which is producing the high boost misreadings shown in this post and mine. Either that, or the sensor itself isn't capable of logging the true boost curve (I really hope this isn't the case).

 

This was a 4th gear pull from 2200 to 7600 rpm:



I can't log over 25psi on the zt-2 map sensor in mitsulogger until MJ releases the next version. Peak boost was 31.7 according to the peak hold function of the zt.2. I max'd the jdm sensor out at 33psi. The curves are a little closer and I personally think the discrepancy is due to sensor placement.

I don't know what I was doing, but I noticed this in my logs:



At lower boost (under 10psi) they were right with each other. This was at less than 50% throttle.

 

One of the first things that I had Pd1 try was to log boost with the ign on and engine off. It read 0 psi as it should (just like mixmastermatt's). This is a good test because these sensors are MAP sensors (vs MDP sensors), and if the scaling were off, it would be quite evident at atmospheric pressure. I also used Pd1's sensor for several days in my own Evo 9, and again, it gave the exact same boost numbers as my own JDM MAP sensor (23-24 psi at peak boost) which also matched my MR boost gauge.

To be straightforward, its not the JDM sensor, and I don't think its the ROM or ECU. I say its not the ROM/ECU because all the logging programs read the JDM sensor voltage directly in "ECU units" of 0-255, and none of the changes I have people make to the ROM change the translation from 0-5V to 0-255. That translation happens somewhere else in the ROM. And if for some reason that translation were different for their ROMs, then the boost measurement with the ign on and engine off would read incorrectly too.

Before I had seen mixmastermatt's boost logs in this thread, I was pretty confounded by Pd1's issue, but after seeing these boost logs, I feel that its something with the ZT-2 boost meas setup. The smoking gun for me is the delayed response of the ZT-2 both in boost rise during spool up and in boost fall between shifts. There are substantial response delays of the ZT-2 in both magnitude of the reading and the speed at which the reading changes. Now that I think more about it, the delay in response between shifts suggests to me that its not a leak in the boost line going to the ZT-2 sensor. If there were a leak, the ZT-2 boost response would be delayed on spool up but not between shifts. Its as if there is a really long boost hose or the ZT-2 sensor has a slower response rate than the JDM MAP sensor.

 

Damn I forgot about that bug in the Zeitronix logger I wrote.

 

That's why I got the JDM sensor. I didn't want to wait on you anymore.

 

FWIW, my JDM MAP sensor read a peak of 19.7 psi on my bone stock EVO 9. We know that a stock 9 has a 20.3 psi peak boost. The 60% humidity lowered my number a bit, but I know that the JDM MAP is accurate from the 19.7 result that I got. I set my sensor to 14.7 psi with the ignition on and engine off.

Yesterday, I installed my innovate logging equipment in the car. I finally had the time to install my HFC with two bungs, one on the mouth of the CAT and one on the tail of the CAT. The tail bung is for the stock sensor and the mouth bung is for the wideband sensor.

I also hooked up my LMA-2 to log boost from the JDM MAP via the LM-1. I alos used the Tactrix cable to log boost via the ECU. I looked at both gauges, the one from the LMA-2 and the one from the ECU. As I expected, the gauge readings from the ECU is LAGGY and SLOW compared to the gauge readings from the LMA-2/LM-1. I had the same results logging RPM from the ECU and from the LMA-2/LM-1. The ECU data is slow and the more parameters you log the slower it becomes.