Disassembly: Manifold Pressure Differential Sensor

[Greg E]

It's late and I'm very tired so I'm not going to get into too much detail about this.

The stock MDP sensor is a simple 1bar sensor on top of the plenum in OBD2 cars to monitor if there's positive pressure in the plenum so the ECU knows when to shut the purge and EGR valves.

For a while now, I've been testing the EVO X map sensor in my car. The idea is to datalog boost pressure using EVOScan and the flashable ECU. The EVO 10 uses a 3.25 bar MAP sensor that can measure up to 32 psi of boost. This sensor is nearly identical in appearance to the stock sensor, so it can be easily swapped in for some plug and play fun.

If you pop off the stock sensor, you'll see a space with a small hole. I believe this cavern is just an area to keep air turbulence away from the sensor so the readings are more accurate.

http://i.imgur.com/RvX8N.jpg

Here are the 2 sensors side by side. Note: the EVO 10 sensor is the one with the yellow cap on it.

http://i.imgur.com/Jwdpv.jpg
http://i.imgur.com/ZcMiZ.jpg
http://i.imgur.com/aaFEu.jpg

No one would have a clue...
http://i.imgur.com/UE3Ad.jpg

Part number for this sensor is 1865A052 and it's available for $145.16 plus shipping thru MitsubishiParts.net - Your #1 Source for OEM Parts and Accessories (http://www.mitsubishiparts.net/cart/?pn=1865A052)

Anyone who is running or planning to run the EVOX MAP sensor, you will need this information in order to keep the periphery bit enabled and pass emissions. Go ahead and add this to your V1 Chromed ROM.

V2 update! Since I've figured out how the emissions tests work, most of these values do not need to be updated. Only the EGR Offset table. I've also eliminated the second EGR table (both tables were the same values anyway).

http://i.imgur.com/Mje9i.jpg

Change the EGR MAP Offset table to the following values:

750 84
1000 84
1250 73
1500 65
1750 60
2000 55
2500 48
3000 43
3500 23

http://i.imgur.com/qUE36.jpg

In V2 you need to set the sensor scaling in order for EVOScan to show accurate numbers. Here are the settings!

http://i.imgur.com/tlWPR.jpg

These values are ONLY for the EVO 10 MAP sensor! I have no clue how to make any of the other sensors out there work with the emissions.

[TurboSinceBirth]

I was actually just looking at the pinout for the manifold pressure differential sensor wondering what in the world I would be using it for but now I know. :) I might keep this in mind although I think I was going to have EVOScan get boost from my BlackStealth LCD boost controller.

[Greg E]

I was actually just looking at the pinout for the manifold pressure differential sensor wondering what in the world I would be using it for but now I know. :) I might keep this in mind although I think I was going to have EVOScan get boost from my BlackStealth LCD boost controller.

I'm trying to find the MAP sensor voltage scaling table. Although this patch fixed the OBD2 check, idle is still lopey.... Just so you're aware!

If you're an OBD1 car, disabling the peripheries for the MAP and EGR work just fine. I did this dis-assembly so OBD2 guys can still use this sensor an pass inspection.

[Greg E]

Just confirmed that my car passed OBD2 inspections with these settings for EVOX map sensor!

Also figured out why the idle is lopey. The periphery0 EGR bit is backwards! Leaving it switched to 0 means it's enabled. Switching it to 0 screws with the write to R0 for the next sub. I will patch this in V2 so it works like we think it should.

[awdonly]

I'm trying to find the MAP sensor voltage scaling table. Although this patch fixed the OBD2 check, idle is still lopey.... Just so you're aware!

If you're an OBD1 car, disabling the peripheries for the MAP and EGR work just fine. I did this dis-assembly so OBD2 guys can still use this sensor an pass inspection.

Thanks Greg - I'm planning on installing an EVO X MAP in my '95 (got an adapter and pigtail) and will be hooking it up to pin 92 of the ECU (its also feeding my LCDBoostController). Will this setup allow me to enable peripheries for the MAP and EGR to get EGR to operate normally? Will it add any other value/functionality asside from ECU boost control?

[Greg E]

Yes it will.

Not right now.

[awdonly]

Thanks Greg,

Will these changes solve the evap purge solenoid clicking problem as well? If not, how do I disable?

[Greg E]

You're going to need the patch to bypass the fuel tank pressure sensor. This will come in V2. If you're interested, I need a beta tester. :)

For now, just unplug the solenoid. It shouldn't be clicking if you have bit 15 of periphery 2 disabled though.

[awdonly]

I can help beta test, but I don't have a stock set-up for a '95 (Cali) since I added the evo x map sensor.

Let me check that bit again when I get home.

[Greg E]

Doesn't matter if you're setup for Cali emissions or not. When OBD2 became standard every car had to pass the same laws statewide. All years/models had to have a purge system. The 98/99 ECU handled that system a bit differently. It my hopes this patch will still keep that function without the added sensor from the 98/99 car.

All I'd like you to test is if the solenoid works with my ROM or not. Seems to work on my car but it would be wise to confirm its functionality on a few vehicles before going public.

[awdonly]

You're going to need the patch to bypass the fuel tank pressure sensor. This will come in V2. If you're interested, I need a beta tester. :)

For now, just unplug the solenoid. It shouldn't be clicking if you have bit 15 of periphery 2 disabled though.

Currently on Chome 1.0 with the MAP mods for EVO X MAP sensor entered. Still getting constant purge clicking - Periphery 2 bit 15 (P0443, P0446, P0403, P0243, P0090, P2263) =0

[awdonly]

Doesn't matter if you're setup for Cali emissions or not. When OBD2 became standard every car had to pass the same laws statewide. All years/models had to have a purge system. The 98/99 ECU handled that system a bit differently. It my hopes this patch will still keep that function without the added sensor from the 98/99 car.

All I'd like you to test is if the solenoid works with my ROM or not. Seems to work on my car but it would be wise to confirm its functionality on a few vehicles before going public.

Happy to help.

[Greg E]

Happy to help.

Can you do a quick test for me?

Open up your ROM and set Periphery2 Bit 14 to 0.

CHANGE NOTHING ELSE and flash this to your car. Tell me if the solenoid still clicks.

If that doesn't work, leave bit 14 at 0 and switch bit 13 to 0 as well. Again, change nothing else!

[awdonly]

I'm afraid I've been playing quite a bit with the code lately - now at 1.1. Also I got a CEL P0400 (EGR) and figured out from your notes that periphery 0 bit 11 was reversed so I cahnged that back to 0 to enable EGR. While I can't swear to it - I think that was the change that stopped the clicking. I don't think Periphery2 bit 14 is the culprit as it has been at 0 for awhile since that is part of the aaronatstate setting I used.

[Greg E]

I'm afraid I've been playing quite a bit with the code lately - now at 1.1. Also I got a CEL P0400 (EGR) and figured out from your notes that periphery 0 bit 11 was reversed so I cahnged that back to 0 to enable EGR. While I can't swear to it - I think that was the change that stopped the clicking. I don't think Periphery2 bit 14 is the culprit as it has been at 0 for awhile since that is part of the aaronatstate setting I used.

Thank you!

Could you quick re-enable those periphery 2 bits (13 and 14), flash to your car and tell me what happens?

[awdonly]

OK - so now I'm confused. I enabled 2 bits (13 & 14) and clicking began within a few secnds of startup. I wasn't getting any clicking after a good 2 min of warmup (post 14) before enabling. I disabled again, reflashed, and .... clicking remains....? Is there a temperature threshold on the system? Its possible it wasn't clicking during my first test because car was cold and once it warmed up it started clicking (timing could have coincided with rom change). Now I'm really confused.

[Greg E]

OK - so now I'm confused. I enabled 2 bits (13 & 14) and clicking began within a few secnds of startup. I wasn't getting any clicking after a good 2 min of warmup (post 14) before enabling. I disabled again, reflashed, and .... clicking remains....? Is there a temperature threshold on the system? Its possible it wasn't clicking during my first test because car was cold and once it warmed up it started clicking (timing could have coincided with rom change). Now I'm really confused.

There's an intake air temperature threshold where the system is disabled below 14 degF and engine coolant temp is below 113 degF.

What's going on is since you don't have the fuel tank sensor, the ECU is seeing a constant reading of 0. This is telling the code that there's excessive vacuum in the fuel tank and it needs to open the purge to equalize the pressure.

Sit tight for a bit. I'm going to write you a patch to test.

[awdonly]

OK. Thanks. I have another EvoX MAP sensor question. In Chrome 1.1. am I correct in saying the boost guage is reporting directly on the MAP input? The reson I ask is that I have the MAP sensor hooked up to both my LCDBoostController and pin 92 of the ECU. Both report 0PSI with the engine off and substantial vac (9PSI) during idle, but the LCD BC is reporting close to 14psi of boost at WOT, while the stock boost gauge barely climbs over 7psi.

[Greg E]

Okay try this now:

CLOSE ECUFlash.
Go to C:\Program Files (x86)\OpenECU\EcuFlash\rommetadata\mitsubishi\3000 GT and open up your ROM's XML.
Open it with notepad.
Scroll all the way to the bottom and copy/paste this code before the </rom> tag.


<table name="purge enrichment test" address="11126" type="2D" level="4" scaling="Hex16">
<table name="X" type="Static Y Axis" elements="1">
<data>make value = 0x0000 (stock = 0x0033)</data>
</table>
</table>

Save and close the file (might have to give admin permission)
Open up ECUFlash and your ROM.
Scroll down to the MISC section and locate the "purge enrichment test" table.
Change the value to "0x0000" like I instructed.
Save
Flash to car and report back.

[Greg E]

OK. Thanks. I have another EvoX MAP sensor question. In Chrome 1.1. am I correct in saying the boost guage is reporting directly on the MAP input? The reson I ask is that I have the MAP sensor hooked up to both my LCDBoostController and pin 92 of the ECU. Both report 0PSI with the engine off and substantial vac (9PSI) during idle, but the LCD BC is reporting close to 14psi of boost at WOT, while the stock boost gauge barely climbs over 7psi.

Yes

Hmm... Disconnect it from your LDCBC and tell me what happens. Worst case, I'll give you a patch to expose the boost gauge table and let you edit the values to be whatever you want.

[awdonly]

Okay try this now:

CLOSE ECUFlash.
Go to C:\Program Files (x86)\OpenECU\EcuFlash\rommetadata\mitsubishi\3000 GT and open up your ROM's XML.
Open it with notepad.
Scroll all the way to the bottom and copy/paste this code before the </rom> tag.


<table name="purge enrichment test" address="11126" type="2D" level="4" scaling="Hex16">
<table name="X" type="Static Y Axis" elements="1">
<data>make value = 0x0000 (stock = 0x0033)</data>
</table>
</table>

Save and close the file (might have to give admin permission)
Open up ECUFlash and your ROM.
Scroll down to the MISC section and locate the "purge enrichment test" table.
Change the value to "0x0033" like I instructed.
Save
Flash to car and report back.

Do you care if periphery2 bits 13 & 14 are enabled or not?

[Greg E]

Do you care if periphery2 bits 13 & 14 are enabled or not?

Leave them enabled.

[awdonly]

Okay try this now:

CLOSE ECUFlash.
Go to C:\Program Files (x86)\OpenECU\EcuFlash\rommetadata\mitsubishi\3000 GT and open up your ROM's XML.
Open it with notepad.
Scroll all the way to the bottom and copy/paste this code before the </rom> tag.


<table name="purge enrichment test" address="11126" type="2D" level="4" scaling="Hex16">
<table name="X" type="Static Y Axis" elements="1">
<data>make value = 0x0000 (stock = 0x0033)</data>
</table>
</table>

Save and close the file (might have to give admin permission)
Open up ECUFlash and your ROM.
Scroll down to the MISC section and locate the "purge enrichment test" table.
Change the value to "0x0033" like I instructed.
Save
Flash to car and report back.

Just for clarity - I'm changing it to 0x0000 not to 0x0033 per the coded instructions.

[awdonly]

I changed purge field to 0x0000 and enabled periphery2 bits 13 & 14 (now at "1"), flashed, and.... no change still lots of clicking at purge solenoid after a few seconds of running.

[awdonly]

Yes

Hmm... Disconnect it from your LDCBC and tell me what happens. Worst case, I'll give you a patch to expose the boost gauge table and let you edit the values to be whatever you want.

Disconnecting is easier said than done... It gets 5v/ground from LCDBC - wonder if differnt sensor grounds are causing the delta.

[Greg E]

I changed purge field to 0x0000 and enabled periphery2 bits 13 & 14 (now at "1"), flashed, and.... no change still lots of clicking at purge solenoid after a few seconds of running.

okay, with ECUFlash open, right click on the "purge enrichment test" table
change the address to 11124
Hit okay
Make the value in that table "0x0000"
Save
reflash

If this doesn't work, there's still one more thing to try so don't fret.

[Greg E]

Disconnecting is easier said than done... It gets 5v/ground from LCDBC - wonder if differnt sensor grounds are causing the delta.

Yes, that's what I was thinking too.

[awdonly]

Are we keeping 11126 at 0x0000 or back to the stock 0x0033?

[Greg E]

Make it 0X0033 to be safe.

[awdonly]

Value was 00CD - changed to 0x0000 - no change in clicking solenoid.

[Greg E]

Alright last test!

Change that value back to 0x00CD.
right click on the "purge enrichment test" table
change the address to 19e86
Hit okay
The new value in the table will be 2708. Make the value "0x2012"
Save
reflash

If this doesn't work, I suspect there's a piece of hardware missing on your ECU to operate this system. Just curious, which ECU are you running?

[awdonly]

It will be tomorrow afternoon before I can test this last one. Its a '98 VR4 ECU (MD347250).

[Greg E]

Hate to ask, but are you absolutely sure it's the purge solenoid that's clicking?

[awdonly]

http://home.comcast.net/~marc_mayer/pwpimages/IMG_0729.JPG
Vertical solenoid sitting right next to my MAC BCS (just to the left of the tie-wrap on the firewall)
Connector A20 - light green wire powers it - red stripe hose attached.

[awdonly]

Now you are making me 2nd guess myself....

The picture at Stealth 316 - Free Boost Mod for 1991-1993 3000GT/Steath (http://www.stealth316.com/2-freeboost.htm) shows the EGR solenoid and purge solenoid reversed from where they are located on my car. I verified both vac and electrically that its my purge solenoid. Red Vac line to purge, green for EGR per http://imageshack.us/f/167/img1606zw3.jpg/ (solenoid with red hose is clicking). According to the manual - the connectors on the BCS and Purge (A-70 and A-20) are the same type and the connectors on the EGR and fuel pressure (A-21 and A-71)are of a different type. The clicking solenoid as the same type connector as the BCS and is fed by a light green wire.

I've convinced myself. Hopefully I've convinced you too ;-)

[Greg E]

Just making sure all the basics are covered.

If this last patch doesn't fix the issue, we know it doesn't have anything to do with the lack of a fuel tank sensor. This means I've got to start tracking the purge control duty cycle address in the code and see where it's getting updated.

[awdonly]

Yes, that's what I was thinking too.

Further researched this an I think I found part of the problem. I am using the OMNI 4 Bar settings on the LCDBC and assumed it was the same as the EVOX sensor that I'm running. Found this thread on topic Omni 4bar writeup - evolutionm.net (http://m.evolutionm.net/showthread.php?t=516456). Do you happen to know if the evoX sensor output is linear PSI v voltage?

[awdonly]

Further researched this an I think I found part of the problem. I am using the OMNI 4 Bar settings on the LCDBC and assumed it was the same as the EVOX sensor that I'm running. Found this thread on topic Omni 4bar writeup - evolutionm.net (http://m.evolutionm.net/showthread.php?t=516456). Do you happen to know if the evoX sensor output is linear PSI v voltage?

Blackstealth is going to add the EVOX sensor to the LCDBC software - do you happen to know how many PSI of vac cause it to have a 0V output? Im assuming 32 PSI is at 5V output.

[Greg E]

PSI----Volt
-14.7---0
-8.9----0.593
-4.4----1.056
0-------1.508
20.1---3.574
32.7---4.870

[Greg E]

Another thing, you'll get a CEL but what happens when you flash the car with the stock periphery setup?

[awdonly]

Changed 19e86 to 0x2012 and periphery bits back to stock - no change in symptoms - clicking begins as soon as car is warmed up.

[awdonly]

Just for grins (and to cover a few more permutations) I changed all three 11126, 11124 and 19e86 as suggested - this time all at once - and left the peripheries in factory mode. No change - still starts clicking when warmed up.

[Greg E]

Alright. I'll get back into this tomorrow after work.

[Greg E]

Noticed something odd. The purge shouldn't be activated at all until you rev to 1500rpms. Starting to wonder if this isn't a hardware issue as I can't recreate this problem on my car. Unplugging the fuel tank sensor didn't cause it to click. I will try unplugging the vent solinoid and see what happens.

[awdonly]

Mine definite begins clicking at idle.

[awdonly]

Noticed something odd. The purge shouldn't be activated at all until you rev to 1500rpms. Starting to wonder if this isn't a hardware issue as I can't recreate this problem on my car. Unplugging the fuel tank sensor didn't cause it to click. I will try unplugging the vent solinoid and see what happens.

Do you happen to know what pin on the '98 ecu should map to which pin on the '95 Cali Ecu for purge control? I can double check that my conversion harness is not the source of trouble.

[Jesters Deadd]

The pin 16 on the 98/99 plug goes to pin 9 on the 95 plug.
Is your power wire to your purge solenoid light green with a blue stripe?
Also I see your EVOX map sensor was installed off the brake booster vacuum line. Did you happen to remove the check valve in that line or is it just a tee into that line?

[Jesters Deadd]

Disconnecting is easier said than done... It gets 5v/ground from LCDBC - wonder if differnt sensor grounds are causing the delta.

Map should be connected to the sensor ground pin 49 and ECU supplied 5v pin 42. Not any other source.

[Jesters Deadd]

Another thing, you'll get a CEL but what happens when you flash the car with the stock periphery setup?

If you have a cali car what happends when you use the complely stock rom on your ECU?

[awdonly]

The pin 16 on the 98/99 plug goes to pin 9 on the 95 plug.
Is your power wire to your purge solenoid light green with a blue stripe?
Also I see your EVOX map sensor was installed off the brake booster vacuum line. Did you happen to remove the check valve in that line or is it just a tee into that line?

Thanks - I'll check the pins; Think its light green with black stripe; I used a hard T (1/8 npt) right off the plenum. The check valve is within the booster hose - at the end farthest away from the plenum. It has not been removed (and romoval would be dangerous).

[awdonly]

If you have a cali car what happends when you use the complely stock rom on your ECU? Don't know but can try.

[awdonly]

Map should be connected to the sensor ground pin 49 and ECU supplied 5v pin 42. Not any other source.

If I was using ECU boost control, I would go that way, but I'm using the LCD boost control so I'm sourcing the sensor with that.

[awdonly]

BTW - I'll be unavail for testing starting Wednesday for about 10 days - sorry.

[Greg E]

If I was using ECU boost control, I would go that way, but I'm using the LCD boost control so I'm sourcing the sensor with that.

You really should wire it up that way. The reason Mitsu wired up all the sensors to use the same power and ground is so the readings stay the same.

[Jesters Deadd]

BTW - I'll be unavail for testing starting Wednesday for about 10 days - sorry.

Where are you located btw? maybe someone can stop by and take a closer look.

[Jesters Deadd]

You really should wire it up that way. The reason Mitsu wired up all the sensors to use the same power and ground is so the readings stay the same.

Expecially if you are trouble shooting wierd issues.
In fact try this exactly...

Check the wire from the purge solenoid has continuity to pin 16 of the 98/99 ECU plug.
Check that wire doesn't have continuity to ground.
Flash the completely stock 99 rom peripheries to your car.
Unplug your EVOX MAP. If your pin 92 is hooked up in parallel to your LCDBC then unhook it.
Make sure your lights and all accessories are off.
Start your car from a overnight cool down (so its room temp) while hooked to EVO scan.
Note the coolant temp, air temp, baro, airflow, idle rpm speeds and approx time if/when the purge solenoid starts to do its clicking.
Report back with coolant temp, air temp, baro, airflow and time when it started.

[awdonly]

You really should wire it up that way. The reason Mitsu wired up all the sensors to use the same power and ground is so the readings stay the same.
Agreed, but I'd rather have float on the MAP input to the ECU than the LCDBC given my configuration and limited use of the MAP input by the ECU (I think its only being used for boost control (not by me), boost guage, and emmissions - anywhere else?. I should be able to log both inputs to see how much float there really is.

[awdonly]

Where are you located btw? maybe someone can stop by and take a closer look.

Fort Collins, CO

[awdonly]

Expecially if you are trouble shooting wierd issues.
In fact try this exactly...

Check the wire from the purge solenoid has continuity to pin 16 of the 98/99 ECU plug.
Check that wire doesn't have continuity to ground.
Flash the completely stock 99 rom peripheries to your car.
Unplug your EVOX MAP. If your pin 92 is hooked up in parallel to your LCDBC then unhook it.
Make sure your lights and all accessories are off.
Start your car from a overnight cool down (so its room temp) while hooked to EVO scan.
Note the coolant temp, air temp, baro, airflow, idle rpm speeds and approx time if/when the purge solenoid starts to do its clicking.
Report back with coolant temp, air temp, baro, airflow and time when it started.

I'll see if I can run test tomorrow or late tonight.

[awdonly]

Expecially if you are trouble shooting wierd issues.
In fact try this exactly...

Check the wire from the purge solenoid has continuity to pin 16 of the 98/99 ECU plug.
Check that wire doesn't have continuity to ground.

Confirmed - BTW wire is light green (no stripe)

[awdonly]

Expecially if you are trouble shooting wierd issues.
In fact try this exactly...

Check the wire from the purge solenoid has continuity to pin 16 of the 98/99 ECU plug.
Check that wire doesn't have continuity to ground.
Flash the completely stock 99 rom peripheries to your car.
Unplug your EVOX MAP. If your pin 92 is hooked up in parallel to your LCDBC then unhook it.
Make sure your lights and all accessories are off.
Start your car from a overnight cool down (so its room temp) while hooked to EVO scan.
Note the coolant temp, air temp, baro, airflow, idle rpm speeds and approx time if/when the purge solenoid starts to do its clicking.
Report back with coolant temp, air temp, baro, airflow and time when it started.

Completed following all instuctions to the letter (except I couldn't wait for room temp - it was 95 degrees today and may garage is still HOT! I prob won't have time tomorrow so I proceeded after 7 hour cooldown) - clicking started about 3 minutes in to test. RPM=843.75;CoolantTemp=154.4;CoolantTempScaled=147 .2;AirTemp=86;MAFAirTempScaled=89.6;Baro=83.79;Air Flow=56.25

If it helps - complete log from Evoscan 2.9 (wasn't sure if I should use scaled temps or not so I took both scaled and unscaled) is here (http://home.comcast.net/~marc_mayer/EvoScanDataLog_2012.06.04_22.44.37%20-%20MRM%20Purge%20valve%20clicking%20about%2010%20s econds%20from%20end%20of%20log.csv). Clicking started about 10 seconds from end of log.

[99 vr4]

If I was using ECU boost control, I would go that way, but I'm using the LCD boost control so I'm sourcing the sensor with that.

Just a quick question, does the LCD Boost Controller have a +5v DC output?

In order for the MAP sensor to operate properly with the ECU, you MUST tie the MAP to the sensor POWER and sensor GROUND at the ECU. Pin 81 is +5v DC for the sensors and pin 92 is the sensor ground. You can also reference this from the TPS for testing purposes. If you have a different power supply source, there will be a delta introduced into the system and the ECU might act wonky.

Also, can you please measure the resistance of the purge solenoid and measure the voltage at the purge solenoid plug. I just want to be sure that your purge solenoid is not bad and that you have stable voltage at the solenoid.

In the mean time Greg has asked me to disconnect a bunch of stuff on my 99 to see if I can duplicate this issue. I will be disconnecting my vent solenoid (which you don't have) and my fuel tank pressure diff sensor (which you don't have) to see if I can trigger the same conditions that your ECU is seeing and see if I can duplicate this.

Also, can you confirm that you DON'T have anything in the following pins on your ECU harness? Pins 35 (Purge VENT) and 93 (Fuel Tank pressure diff sensor).

Also, the following should be empty as they are various pins NOT used on the VR4: Pins 5, 7, 8, 30, 31, 32, 33, 34, 53, 56, 63, 64, 76, 81, 82, 86, 87, 88, 89, 90, 94, 95, 96, 97, 98

@Greg, the 99 has a different purge solenoid and it is bigger, so can you measure the resistance of yours to be sure there is not a big difference between the two solenoids?

[Jesters Deadd]

Do you have a montero MAf by any chance?
Ok, did you get any CELs?
Now turn on the lights and defrost. While at idle does the clicking stop?
With lights still on turn on AC. Does clicking stop?
With the lights still on rev the engine quickly several times. Does clicking stop while reving?

[Jesters Deadd]

Your solenoid should measure 35-44 ohms. PLease finish the testing above to confirm it is working corretly or not.

[awdonly]

Just a quick question, does the LCD Boost Controller have a +5v DC output?

In order for the MAP sensor to operate properly with the ECU, you MUST tie the MAP to the sensor POWER and sensor GROUND at the ECU. Pin 81 is +5v DC for the sensors and pin 92 is the sensor ground. You can also reference this from the TPS for testing purposes. If you have a different power supply source, there will be a delta introduced into the system and the ECU might act wonky.

Also, can you please measure the resistance of the purge solenoid and measure the voltage at the purge solenoid plug. I just want to be sure that your purge solenoid is not bad and that you have stable voltage at the solenoid.

In the mean time Greg has asked me to disconnect a bunch of stuff on my 99 to see if I can duplicate this issue. I will be disconnecting my vent solenoid (which you don't have) and my fuel tank pressure diff sensor (which you don't have) to see if I can trigger the same conditions that your ECU is seeing and see if I can duplicate this.

Also, can you confirm that you DON'T have anything in the following pins on your ECU harness? Pins 35 (Purge VENT) and 93 (Fuel Tank pressure diff sensor).

Also, the following should be empty as they are various pins NOT used on the VR4: Pins 5, 7, 8, 30, 31, 32, 33, 34, 53, 56, 63, 64, 76, 81, 82, 86, 87, 88, 89, 90, 94, 95, 96, 97, 98

@Greg, the 99 has a different purge solenoid and it is bigger, so can you measure the resistance of yours to be sure there is not a big difference between the two solenoids?

Yes the LCDBC has a 3 pin connector for external MAP sensor (+5/sensor output/sensor ground). The sensor output is wired into the MAP input on the ECU in addition to the MAP input on the LCDBC- only delta would be bad read on MAP level by ECU and I've now correlated those (OEM boost gauge on Chrome 1.1) with the boost level output on the LCDBC and they are very close (they were off because I was using OMNI 4 settings on LCDBC - my bad).

ECU conversion harness was purchased from BlackStealth on 3si who has been making these for awhile. I have no reason to believe it was made incorrectly, but will ask him to verify the pins in question are not used.

Solenoid measures 43.6 ohms.

[Jesters Deadd]

Yes the LCDBC has a 3 pin connector for external MAP sensor (+5/sensor output/sensor ground). The sensor output is wired into the MAP input on the ECU in addition to the MAP input on the LCDBC- only delta would be bad read on MAP level by ECU and I've now correlated those (OEM boost gauge on Chrome 1.1) with the boost level output on the LCDBC and they are very close (they were off because I was using OMNI 4 settings on LCDBC - my bad).

ECU conversion harness was purchased from BlackStealth on 3si who has been making these for awhile. I have no reason to believe it was made incorrectly, but will ask him to verify the pins in question are not used.

Solenoid measures 43.6 ohms.
I doubt it is the conversion harness.
What power circuit is your LCD boost controller hooked to?

[awdonly]

Switched Fuse 12 (15A) - chose this circuit because low loading - it no longer powers fuel pump.

[awdonly]

Do you have a montero MAf by any chance?
Ok, did you get any CELs?
Now turn on the lights and defrost. While at idle does the clicking stop?
With lights still on turn on AC. Does clicking stop?
With the lights still on rev the engine quickly several times. Does clicking stop while reving?

Stock MAF (evo MAF ready to install but limiting variables). Yes I've constantly had CELs since testing included keeping all peripheries as stock - and stock '99 code is also the same. I can double-check what is stored after dinner and should be able to run test you requested tonight. Unfortunately that will be the last input from me on this for 10 days. Anything else you want me to squeeze in tonight?

[Greg E]

What power circuit is your LCD boost controller hooked to?

BINGO! 150 points to Adam.

Turned on my headlights, fog lights, AC, rear defrost, stereo and 4 ways. Engine warmed up, purge solenoid started clicking! Now to figure out which address is the load switch and I'll have something for you to test again awdonly! Thank you Adam!

Few other notes:

*The purge solenoid on my car shows a resistance of 38.9ohms. Lower than what awdonly is showing HOWEVER the manual shows a spec standard value of 36-44ohms for ALL year cars (91-99). So even though they may be different part numbers, they have the same specs.

*On the 98/99 cars the purge is mounted onto the intake plenum instead of on the firewall. So when it's active on my car, I can't hear it. Had to touch it with my hand in order to notice it was active.

*The EGR solenoid on the 98/99s was moved down into the position of the purge solenoid on the 91-97 cars. That tricked me earlier....

*The vent system even though it functions with the purge system, it's a completely and totally separate system. The purge does NOT use the fuel tank sensor for operation. Only a few condition bits activate the purge which I'm currently testing. The only thing the two systems share is periphery 2 bit 13 and NONE of the peripheries disable the system. Only the OBD checks for them. The purpose of the vent solenoid is just to check the operation of the fuel tank sensor. Since this isn't present on the 91-97 cars, periphery 2 bit 13 needs to be switched off to keep the CE light off. This will NOT effect the purge system in anyway.

Going back into my little hole for a bit but I hope to have something for you to try later awdonly. BTW, what's your real name? I appreciate your help with this too!

[BlackStealth]

I just checked the list of empty pins, and they match to what I have as empty pin also. The very early conversion harnesses did have pin 75 (Muffler mode changeover) connected, but I have left that out, as I'm sure it's useless for a 99 ECU.

One thing I want to point out is that signal ground on the ECU is not the same signal ground used on the LCD BC. How much of an impact this has, I don't know. Sharing the sensor could potentially cause a problem. Does Evoscan show correct readings for the MAP sensor?

[Jesters Deadd]

Stock MAF (evo MAF ready to install but limiting variables). Yes I've constantly had CELs since testing included keeping all peripheries as stock - and stock '99 code is also the same. I can double-check what is stored after dinner and should be able to run test you requested tonight. Unfortunately that will be the last input from me on this for 10 days. Anything else you want me to squeeze in tonight?

OK.
Start car turn off all electric load. Ie lights, defrost, brake light, moving mirrors, etc.
Once the solenoid starts clicking test pin 52 and see if it has 12v going to it.
Remove the wire and see if the solenoid still clicks when the car is running at idle.
The thought is you have something tripping your electric load switch, pin 52 with 12v power.
When your elec load switch is tripped the solenoid will activate.

[awdonly]

BINGO! 150 points to Adam.

Turned on my headlights, fog lights, AC, rear defrost, stereo and 4 ways. Engine warmed up, purge solenoid started clicking! Now to figure out which address is the load switch and I'll have something for you to test again awdonly! Thank you Adam!

Few other notes:

*The purge solenoid on my car shows a resistance of 38.9ohms. Lower than what awdonly is showing HOWEVER the manual shows a spec standard value of 36-44ohms for ALL year cars (91-99). So even though they may be different part numbers, they have the same specs.

*On the 98/99 cars the purge is mounted onto the intake plenum instead of on the firewall. So when it's active on my car, I can't hear it. Had to touch it with my hand in order to notice it was active.

*The EGR solenoid on the 98/99s was moved down into the position of the purge solenoid on the 91-97 cars. That tricked me earlier....

*The vent system even though it functions with the purge system, it's a completely and totally separate system. The purge does NOT use the fuel tank sensor for operation. Only a few condition bits activate the purge which I'm currently testing. The only thing the two systems share is periphery 2 bit 13 and NONE of the peripheries disable the system. Only the OBD checks for them. The purpose of the vent solenoid is just to check the operation of the fuel tank sensor. Since this isn't present on the 91-97 cars, periphery 2 bit 13 needs to be switched off to keep the CE light off. This will NOT effect the purge system in anyway.

Going back into my little hole for a bit but I hope to have something for you to try later awdonly. BTW, what's your real name? I appreciate your help with this too!

No problem - I'm having fun. I wonder why I'm getting clicking with 0 accessories on.... My name is Marc.

[99 vr4]

I just checked the list of empty pins, and they match to what I have as empty pin also. The very early conversion harnesses did have pin 75 (Muffler mode changeover) connected, but I have left that out, as I'm sure it's useless for a 99 ECU.

One thing I want to point out is that signal ground on the ECU is not the same signal ground used on the LCD BC. How much of an impact this has, I don't know. Sharing the sensor could potentially cause a problem. Does Evoscan show correct readings for the MAP sensor?

Mitsubishi threshold engineered this car. With that in mind, yes there is a potential of a different voltage potential between the grounds. For instance, if you measured voltage between sensor ground and your LDBC ground, you could have +/- 0.2 volts . . . . that means that all of the inputs would be offset +/- 0.2 volts. For the MAP, that might mean several pounds . . .

[awdonly]

I just checked the list of empty pins, and they match to what I have as empty pin also. The very early conversion harnesses did have pin 75 (Muffler mode changeover) connected, but I have left that out, as I'm sure it's useless for a 99 ECU.

One thing I want to point out is that signal ground on the ECU is not the same signal ground used on the LCD BC. How much of an impact this has, I don't know. Sharing the sensor could potentially cause a problem. Does Evoscan show correct readings for the MAP sensor?

Thanks. Haven't datalogged MAP output yet - Is it possible to datalog MAP data from ECU and MAP data from LCDBC separately?

[99 vr4]

OK.
Start car turn off all electric load. Ie lights, defrost, brake light, moving mirrors, etc.
Once the solenoid starts clicking test pin 52 and see if it has 12v going to it.
Remove the wire and see if the solenoid still clicks when the car is running at idle.
The thought is you have something tripping your electric load switch, pin 52 with 12v power.
When your elec load switch is tripped the solenoid will activate.

Hmmmmm . . . Adam, why does THAT solenoid activate with the electric load switch connected. Should it? I thought the purpose of the electric load switch was so that the ECU knew when their was a large draw on the system and then it could compensate accordingly . . .

[Greg E]

I just checked the list of empty pins, and they match to what I have as empty pin also. The very early conversion harnesses did have pin 75 (Muffler mode changeover) connected, but I have left that out, as I'm sure it's useless for a 99 ECU.

Actually it's not useless. ;)

That's another disassembly discussion for another time though.


One thing I want to point out is that signal ground on the ECU is not the same signal ground used on the LCD BC. How much of an impact this has, I don't know. Sharing the sensor could potentially cause a problem. Does Evoscan show correct readings for the MAP sensor?

EVOScan does show the correct readings for the EVOX MAP sensor if you're using my data.xml definition file! Marc noted something about coolant temp scaled which I know is not in my file but it is in the data.xml file that came with EVOScan. I don't think Marc has EVOScan setup right.

[99 vr4]

Actually it's not useless. ;)

That's another disassembly discussion for another time though.



EVOScan does show the correct readings for the EVOX MAP sensor if you're using my data.xml definition file! Marc noted something about coolant temp scaled which I know is not in my file but it is in the data.xml file that came with EVOScan. I don't think Marc has EVOScan setup right.

Marc is using EvoScan 2.9 . . . he would be a GREAT beta tester for my new 2.9 file . . .

[Greg E]

Marc is using EvoScan 2.9 . . . he would be a GREAT beta tester for my new 2.9 file . . .

:lol:

I have much to go thru and correct in that file though. That's still #3 on the list.

*fix purge issue
*configure ROM to work with only 2 O2 sensors
*test/configure new 2.9 data file
*write up 3/S specific ECUFlash manual
*release V2 to the public

[awdonly]

I'm using 2.9 with the stock rom tests (that we just did) and 2.6 with Chrome. In 2.9 there are scaled temperature values that the evoscan notes indicate should be used for '98+ cars. I've found some of the evoscan notes to be a bit misleading at times though.

[Greg E]

I'm using 2.9 with the stock rom tests (that we just did) and 2.6 with Chrome. In 2.9 there are scaled temperature values that the evoscan notes indicate should be used for '98+ cars. I've found some of the evoscan notes to be a bit misleading at times though.

It's been assumed for years the 3/S was like the EVO/DSM. It's really not.......

Best to use my 2.6 with the stock ROM too.

I will be releasing a 2.9 data file in the very near future (thanks to Brett)!

[awdonly]

OK.
Start car turn off all electric load. Ie lights, defrost, brake light, moving mirrors, etc.
Once the solenoid starts clicking test pin 52 and see if it has 12v going to it.
Remove the wire and see if the solenoid still clicks when the car is running at idle.
The thought is you have something tripping your electric load switch, pin 52 with 12v power.
When your elec load switch is tripped the solenoid will activate.

The harness connector at the '95 end of the conversion harness is both more convenient and Blackstealth has very nicely numbered the pins. Do you happen to know what pin 52 converts to at that end?

[99 vr4]

Pin 24 on 95 to pin 52 on 98/99

Also both the normal temps and the "scaled" temps are incorrect. Thanks to Aaron for helping me figure that out. The 2.9 file I gave to Greg has the corrected formulas.

[Greg E]

Marc, one more thing for you to test.

Open up EVOScan 2.9 and while the purge is clicking, tell me if your idle position switch is on or off.

[Jesters Deadd]

Good thought Greg.

[Jesters Deadd]

Hmmmmm . . . Adam, why does THAT solenoid activate with the electric load switch connected. Should it? I thought the purpose of the electric load switch was so that the ECU knew when their was a large draw on the system and then it could compensate accordingly . . .

Just a hunch. The test for the solenoid specifically says the elec load should be off.

[Greg E]

FINALLY got this figured out Marc!


AirFlow=56.25

You're seeing a lot of airflow at idle and this is why it's not turning off the solenoid at idle like it should.

*Close ECUFlash again and open up your V1.1 XML.
*Copy this code into the bottom (again before the </rom> tag)


<scaling name="MAF Raw 8bit" units="HZ" toexpr="x*6.25" frexpr="x/6.25" format="%.0f" min="0" max="1600" inc="1" storagetype="uint16" endian="big"/>

<table name="Purge Solenoid Trigger HZ" category="Purge" address="10fda" type="2D" level="4" scaling="MAF Raw 8bit">
<table name="HZ" type="Static Y Axis" elements="2">
<data>Purge Activation MAF HZ</data>
<data>Purge Deactivation MAF HZ</data>
</table>
</table>

*Save and close
*open ECUFlash again

If you scroll down to the bottom, you'll find this now:
http://i.imgur.com/N5hHD.jpg

What's happening is the purge is never getting switched off as the airflow HZ count never dips down below the deactivation HZ trigger value. Raise this value to say 60 or 65 and set the Activation value to say 100 and flash to your ECU.

Yes, in the photo you can see there's more to it, but I'll create another separate disassembly thread later about the purge and vent systems as soon as I verify a few more things.

I tested this in my car (engine warm, no headlights on or anything) by bumping up the idle till the MAF HZ value never would dip below 50. Sure enough, the solenoid would turn on, but never turn off. Course, I had to verify this by leaving my hood popped and touching the solenoid because I can't hear it. :lol:

The reason my solenoid clicked before with the lights/AC/radio on before was because I have my target idle bumped up to 850rpm. With the AC on, the car idles even higher. Higher idle = more airflow counted by the MAF. This Count put the idle count just above the deactivation HZ trigger so I set this value to 65 on my car and it functions normally.

Anyway, it would seem there's really nothing wrong at all here. :) Just the conditions of the setup on the car need a tweak. Hopefully Marc can duplicate my results!

[Jesters Deadd]

Awesome!

[99 vr4]

Holy Crap Greg! I am glad to see this put to rest!

[awdonly]

FINALLY got this figured out Marc!



You're seeing a lot of airflow at idle and this is why it's not turning off the solenoid at idle like it should.

*Close ECUFlash again and open up your V1.1 XML.
*Copy this code into the bottom (again before the </rom> tag)


<scaling name="MAF Raw 8bit" units="HZ" toexpr="x*6.25" frexpr="x/6.25" format="%.0f" min="0" max="1600" inc="1" storagetype="uint16" endian="big"/>

<table name="Purge Solenoid Trigger HZ" category="Purge" address="10fda" type="2D" level="4" scaling="MAF Raw 8bit">
<table name="HZ" type="Static Y Axis" elements="2">
<data>Purge Activation MAF HZ</data>
<data>Purge Deactivation MAF HZ</data>
</table>
</table>

*Save and close
*open ECUFlash again

If you scroll down to the bottom, you'll find this now:
http://i.imgur.com/N5hHD.jpg

What's happening is the purge is never getting switched off as the airflow HZ count never dips down below the deactivation HZ trigger value. Raise this value to say 60 or 65 and set the Activation value to say 100 and flash to your ECU.

Yes, in the photo you can see there's more to it, but I'll create another separate disassembly thread later about the purge and vent systems as soon as I verify a few more things.

I tested this in my car (engine warm, no headlights on or anything) by bumping up the idle till the MAF HZ value never would dip below 50. Sure enough, the solenoid would turn on, but never turn off. Course, I had to verify this by leaving my hood popped and touching the solenoid because I can't hear it. :lol:

The reason my solenoid clicked before with the lights/AC/radio on before was because I have my target idle bumped up to 850rpm. With the AC on, the car idles even higher. Higher idle = more airflow counted by the MAF. This Count put the idle count just above the deactivation HZ trigger so I set this value to 65 on my car and it functions normally.

Anyway, it would seem there's really nothing wrong at all here. :) Just the conditions of the setup on the car need a tweak. Hopefully Marc can duplicate my results!

In Chrome 1.1 - All previous code changes back to stock and peripheries as stock I assume. Still taking care of stuff at home, but I'll upload and run test before morning.

[Greg E]

Won't make any inference how you set the peripheries. The purge and vent operations are not tied to them. Only the OBD2 tests are.

Go ahead and disable bit 13 & 14 in periphery 2 to keep your CEL off.

[awdonly]

New software installed - Activation 100 - deactivation 62; The f**king thing still starts ticking after about 5 min of warmup

[awdonly]

For the sake of completeness - pin 52 stays low when clicking; Idle switch shows 0 in datalog and airflow is 56Hz.

[Greg E]

Had a feeling you'd say that... :( There's also a timer function which activates the solenoid. I didn't run my car long enough to see if it turned back on.

I'll pick this up again tonight.

BTW, have you driven you car yet?

[awdonly]

Not since this most recent update, but I did the day before yesterday when it was hot. I'll be away from the car for the next 10 days. Sorry.

[Jesters Deadd]

Can you explain more about this timer function?

[Greg E]

I will when I understand it better

[awdonly]

I'm back. Any progress?

[Greg E]

I'm back. Any progress?

Yes. Quite a bit actually. I'll post up later this weekend.

[Greg E]

Btw what's your email and/or cell? Probably better to talk that way instead of thru here.

[awdonly]

PM sent

[Greg E]

Got it!

I'm up in Easton flash tuning a couple cars, diagnosing the NLTS issue on Aaron's car, getting new tires for my car and entertaining a date with a hot chick.

Get back to you on Monday! Got a much better understanding of what's going on now. Will explain later.

[Jesters Deadd]

Did we resolve this issue?

[Greg E]

Did we resolve this issue?

No but I've been working with Marc on it. Made some progress this week. I'm writing something up for him to try right now actually.

His and my availability just don't match up well so progress is slow.

[Greg E]

Oh speaking of, remember that timer I mentioned earlier? I believe It's actually a function based off the change in speed. I'm pretty certain the purge activates if your speed does t change for a long period of time. Like when you're cruising for a long time or just idling in traffic and not moving.

[Jesters Deadd]

timer huh? that seems like it could be used for other coding tricks.:)

[Greg E]

timer huh? that seems like it could be used for other coding tricks.:)

No, this is specifically for the speedo. ...unfortunately....

I'm wondering if the reason the solinoid activates on Marc car is because he lets the motor warm up at idle.

[awdonly]

No, this is specifically for the speedo. ...unfortunately....

I'm wondering if the reason the solinoid activates on Marc car is because he lets the motor warm up at idle.

Unlikely since once the car is at operating temp - if you turn it off and then restart the solenoid will start clicking within just a few seconds of startup.

[Greg E]

Unlikely since once the car is at operating temp - if you turn it off and then restart the solenoid will start clicking within just a few seconds of startup.

So maybe it's not a speedo delta like I originally thought. It's just when the car isn't moving at all.

[awdonly]

Think I may have just figured out your timer. Fresh DR650s are in and car is idling in garage while I inspect for water/oil leaks and letting the penetrating oil burn off. For complete disclosure, I've also changed to EVO MAF and 560s. The purge valve started clicking as soon as car was warm, but after another few minutes it stopped - after about a minute its back on....

[Kx1984]

does anyone have a diagram on which pin is what on the map?
for the people adding them

Kx