Well I think my 3bar is working correctly in evoscan

Just need to find a connector to fit the MAT sensor now

 

wreckleford, do you have MAP in your logs already reading the correct values?

If so, you just need to apply the correct scaling to the top row of the MAP VE table.

If not, you need to follow the guides in other threads about how to add a MAP sensor to your ECU and get that working first with no CEL and correct logging values on the standard MAP logging request ID of 38.

 

ok, I've got jcsbanks' speed density v3 patch fully ported over to the 88590015 ROM. A patched ROM and the xml for the patch are in the linked zip file. As I mentioned before, you'll need to have a 3+ bar MAP sensor and a MAT sensor installed along with my patches for those two pieces of hardware. The tuning tables and MAT scaling table are setup for a JDM MAP sensor and a GM IAT sensor. With this patch, the MAF can be completely removed with no CELs.

Temporary DISCLAIMER: I have not yet tested the patch on my car (but have tested it in a simulator). Try it first at low load conditions, and log data during your testing.

http://se30.dyndns.org:8080/EVO/8859..._JCSBSD3M2.zip

I have a comment for the SD theory folks. In the process of setting up the IAT -> MAT conversion, I got to thinking about whether it makes sense to substitute MAT in place of IAT for this patch. If MAP can truly be converted to load with the MAP VE and RPM VE scaling tables alone, then it makes no sense to replace IAT with MAT for the following reason: The temperature oompensated load and other things that are scaled with air temp (esp. IPW) have nothing to do with the baseline load calculation and were designed to provide the correct air temperature compensations based on IAT. Substituting MAT in place of IAT will just screw things up because of the higher air temps in the IM. My thought is that for the sake of ditching the MAF, another IAT sensor needs to be installed, but rather than installing it in the IM, it makes more sense to put it in the UICP. Comments?

EDIT: Instructions for applying the SD patch are as follows:

1) Copy the new xml definitions from the zip file to your ECUFlash ROM xml file.

2) Launch ECUFlash and open your current ROM and the stock patched ROM in the zip file.

3) In ECUFlash, copy the values in the new tables from the stock patched ROM to your ROM.

4) Save your current ROM and upload it to your ECU.

Assuming you've already done the JDM MAP and the GM IAT patches, then that's it. It works for a tephra ROM or a non-tephra ROM.

 

But the IM is still more accurate in terms of air temp going into the cylinder right? Or are you saying it doesn't matter?

Thanks for the great work. Three day weekend coming up i should have some testing done.

 

The temperature (& baro) compensations are to provide air volume to air mass conversion using our volume airflow meter.

For speed density, to get an accurate air mass measurement you want the air temperature and pressure measured in the same location. Since the pressure is measured after the throttle, in the inlet manifold, it is better to measure the temperature there too. UICP or IAT are 2nd and 3rd best options, they will still work. Heatsoak of sensors is another issue, but that also affects IAT.

The reason I say that IAT is the 3rd best option for air mass compensation is that you will get an error in air mass estimation as the delta between MAT and IAT increases as you heatsoak the intercooler and the inlet manifold temperature heats up from transfer from the cylinder head, TB water circuit if still present, and from the intake air.

UICP is inferior because of the pressure and temperature gradients across the throttle body.

Speed density ECU manufacturers always recommend a post intercooler location at least.

In my testing so far, I found that the wideband AFR has been consistent, it would have been 5% richer if using IAT in some conditions, but not in others. With good airflow (high speed) on a cold day, using no boost, the IAT and MAT can have a small delta, with low speed and/or boost use there is a delta, and this would give you an error if using IAT as your compensation. The error would be in the generally safe direction of retard and enrichment, but so does not running temperature compensation at all.

Not having baro I think is a very small concern. Air mass will tend to be overestimated slightly at elevation or on days with lower baro, but the error is only due to the turbine and compressor efficiency being lower due to the higher rotational speed required to provide the same MAP. It is nothing like the sort of error that you would get with a volume airflow sensor without baro correction (error would be 8% on the stock airbox at high revs).

 

^ now after typing a useful post I'm going to admit I had wondered about going VTA on my stock IX metal bypass valve With Subaru recirculating valves, they would leak at idle, anyone know if the Evo metal one does?

Generally I hate the sneezing noise, but I wondered about trying it for giggles before becoming a sensible doctor again

 

Truthfully, I don't think it matters, since it can be tuned either way (IM or UICP) with the tables that we have from the patch.

In my opinion though, I think the temps from the UICP would be less likely to heat soak and cause drifts in the tune during low airflow/idle situations, etc. But, since we still have closed loop operation and everything else working as normal, I don't think it should be an issue.

 

I believe that the stock DV is designed to be slightly open at idle vacuum levels. I can't confirm this though, as I have never checked.

Anyway, running VTA in my opinion is just a waste...all you get is decreased turbo response between shifts and some more noise. : )

 

I can't wait to run VTA, my inner ricer wants to buy the loudest BOV possible <- not joking.

 

Agree, just fighting with my inner ricer. You've beaten him to a pulp and my usual refined behaviour has resumed

 

I figured this might be the best place to ask...does anyone have any extra EVO VIII ECU pins laying around?

If/when this SD patch ever gets converted to 96530006, I will need two pins to setup mrfred's fuel temperature/mat patch and to wire my GM 3.3bar into the mdp pin on the ECU. I can do both without the pins I guess, but it would be a lot clean with pins and not cutting wires/etc. I looked in the links from jack_of_trades thread and it says it's like a 54 day wait for them.

For the map sensor I was even thinking about buying another that will replace the stock 1-bar (and run with two map sensors ), but since I already have the GM 3.3bar and it's working fine, I figured I would just use that.

 

You make me laugh, John. Of course, my comment wasn't meant to condescend at all...simply info from my experiences in the past.

I remember when I went blow-through on my Eclipse years ago, I used a Tial BOV and changed it out within about a week to a recirc DV again. The loss of spool and power between shifts just made the car feel like crap.

 

I have used MOLEX 90119-2110 for ECU bench, they fit on ECU pin, but I do not know if they click in the connector thou, you may find them on farnell.com

 

Why would it cause decreased response from a technical standpoint?

 

you can avoid the stock IX bov from leaking at idle. You have to remove the bottom 3 screws from it and take it apart. There is a vent hole that allows air into the bottom. This vent cases the valve to slightly open at idle. This is due to vacuum on the top of the diaphram and pressure in the intercooler pipe. If you plug the hole with JB weld, it will act like a true BOV and stay shut at idle.