Project #7 from Jason: 12 valve VR6 Half Build.

[Jason]

I got to thinking about this turbo on a bike and wouldn't this be a rather large turbo for a bike? What is spool like? Laggy? I guess on a bike you are rev happy with it and are way out there but what is the initial spool like? I know this turbo is too large for what I would put on a 1.8T 4 cyl application.. The largest I would suggest on that is a 3076R and they really seem to like the GT28RS although fully spooled is getting up there. I could have gone larger on my motor but spool gets out there for what I would like to set my redline to. Factory redline is 7000 i think and my current NA chip lifts it 7200 which is probably what I will tune the standalone to. I do have heavy duty springs and titanium retainers so I really could rev higher without a problem but valve float in a motor with this much money and a daily driver seems too risky for what I want..

[Jason]

Got some more of my components for this build and got started on somethings..

Innovate MTX-L Widband
MegaSquirt 3 v3.57 Standalone
MegaSquirt 3x expansion
4 bar MAP sensor with second sensor for barometric correction
Boost solenoid
Harnesses


Nothin like pulling apart a brand new $650 standalone to modify it.



New MapDaddy 4 bar with barometric correction


Remove the 2.5 bar map sensor that comes in the standalone



Solder in the 4 bar

I got the board a little hot with my big *** tip I had on my iron but its ok :/

[Jason]

Next was add a jumper wire from JS5 to the barometric correction sensor



Notice JS5 next to the emtpy hole.. This is under the white colored board.


Put it back together... What a close fit


I decided to tear down my spare factory ECU so I could create a Plug and Play solution.

Next was add a jumper wire from JS5 to the barometric correction sensor



Notice JS5 next to the emtpy hole.. This is under the white colored board.


Put it back together... What a close fit

[Jason]

Next I moved on to control... I have not fully decided how I want to control power to the new components. I can wire the 12v lead to the factory 12v pin and run the ECU off the factory relay or I can use my own main ECU relay and power the ECU with it and altimately power a block for anything that needs to be ECU switched.. Also since I will have my factory fuel pump plus the 044, I can decide if I wanna run both off of the factory relay or use a seperate relay for the bosch. Anyways I wired this up for the time being and am leaning towards using it..


1 relay for ECU and one for 044 pump.
1 block is IGN switched +12v and the other(has 044 relay) is ECU switched +12v

To explain, the IGN switched block powers the ECU relay turning on the ECU through the brown (ground) wire and jumps power to the ECU switched block. The ECU block has the power for the pump relay and the ground wire from this would go to the ECU as the ECU turns on the pump by grounding out the relay. This is why the relay must be ECU switched and not given a constant +12v as it may back feed into the ECU keeping it powered up. Also notice the fuse block to provide the constant +12v to the relays. The extra red +12v wire coming off the pump relay goes to the pump + side.

Make sure all connectors have dielectric greese


Back to the harness, I put a hole in the side of the factory case for the wires to it in


Make it so I can remove the wires from the case if I want (prob will to feed the harness through firewall)


Start soldering the wires to the correct pins

[Jason]

And eventually they are all attached other than grounds which I will prob pull to the battery and the boost solenoid signal



In the last pic, you will notice extra wires rolled up. These wires are not used in my setup so I will tie them up inside the car. The MS3 and MS3x have seperate harnesses so I split them each into two strands (used wires vsnon used). I then joined the two harnesses into one so I will have one strand rolled up in the car that does not get used and one run to the rain tray where it will plug into the factory harness at the factory ecu location.
MS3 harness (non covered wires are not used)


MS3x harness joined with MS3.. All unuesed wires rolled up in one cover and all used wires in the other cover.



And that is where it sits. I have already gone through and setup the ecu for the car so it should hopefully be put it in the car, plug it in and start+run the car.. It will need to be tuned still though.. I am also considering getting 6 EGT sensors before the turbo setup is run so I can make sure all my cylinders are equal before tuning based on the AFR.. On the VR6 motor, the back 3 cylinders tend to run hotter so running sequential fuel, I can dump a little extra fuel into those cyls to cool them and equalize them all out. Most people claim 5 to 10 % more fuel is required in these cylinders. The MegaSquirt system would then be able to connect with the EGT sensors either directly with some mods I believe or by using a module for the EGT sensors that then would communicate with the MegaSquirt over CAN-H and CAN-L channels.

[Jason]

So more parts are here.


Started putting the standalone in the car. Cut a hole in the rain tray area to run the harness through.


Ran the grounds over to the battery


These wrapped wires are ground and the rest of the harness to the bay is visible towards the top of the pic


Ran the harness for the wideband O2, a wire to hook up one of the widebands outputs to the factory O2 plug, and had to add a wire I fforgot to put in the MS3 harness to switch on the factory grounds. Also notice the new home made gromets the plug the hole


My friend Kyle climbed under my LOW ### car to swap my O2 sensor... Nice of him


Temp pod for the O2 guage


Temp ground hookup. Thinking about putting them on a terminal block and runing one or two large grounds to the battery from the terminal block.


Kyle hooking up my relay/terminal pieces to the factory relay panel... I decided to get rid of the IGN terminal block and just use one of the legs on the ECU switched block. I only needed two things hooked to the IGN switched power so we just put them on the same screw and cut out an entire block. The fuse block and single terminal block are now 3M tapped to the bottom of the factory fuse panel and holding fine..Can't see them unless you are bent over looking under there.

[Jason]

Plugged in the MS3 for a temp time to try it out. Found it didn't work right off the bat..


When I started this, I made a spreadsheet to show which wire of the factory plug/harness went to each wire of the MegaSquirt/MegaSquirt3x harnesses.


The first issue I ran into was getting the correct reading from my crank sensor. After figuring out the crank sensor wiring of the harness I got which was different than others usually wire these in vw's, I was able to get some signal input. There was still a problem with it but it wasn't too bad of an issue once we started to work it out. Everyone I know of running the MS on their VW has the crank sensor set to falling edge. This is in relation to the sensor counting the teeth on the crank sensor wheel. In the following pic you can see two long teeth (One really long and one semi.. About half of the really long one) which indicates that I needed to change the MS3 to change the software to look at the raising edge rather than the falling edge.


Once that was resolved, the motor fired right up however the idle was wandering but I had not yet set that up in the software. I quickly set the tune to control the idle and realized I need to either modify my MS3 by adding a TIP120 and a diode into the PWM FIDLE (Idle) circuit or switch over to using the Idle output on the MS3x. The current plan is to switch over to the MS3x Idle output and give that a shot tuning the MS to use PWM Warmup Only output to control the Idle valve. With the generation of my car, the MS3x output is hit or miss whether it will just work or not so I need to test it and if it does not work then I will mod the FIDLE circuit and switch back to using that. Either way I can resolve it and that is next step. Following the Idle will be some tunning on the NA motor and then I move to building the new motor setup followed by the turbo setup

[sportbike]

A guy I know has a '53 beetle with a turbocharged 2.0 watercooled motor in it and recently started having MS crank trigger issues with it as well. Not sure if he got it figured out.

[Jason]

The tooth logger definetly helped diagnose my problem with it. If his was already working, might wanna make sure the sensor is not going out. They like to stop reading when they get warm when they are bad. The cars will start and run unitl it gets up to running temps and then it dies until it cools down.

[sportbike]

He said he tried two other sensors. Now he is kind of out of ideas, unless it is the MS box itself. He had a wild ### idea that the crank may have become magnetized, but I've never experiences that.

[breakneckmot]

Absolutely awesome build!

[its_34]

Nice work! The MS wiring harness looks like a nightmare but you seem to have a methodical approach. Good notes go a long way to diagnosing issues down the road.

Does the MS read direct from a lambda sensor or does it take an output from the Innovate?

[Jason]

First off, I will answer the lamda question. The Innovate wide band was put in place of my factory sensor. When running the wideband, it has a harness that actually goes to a plug on the guage. The wideband controller is built into the guage. Now the nice thing about this was it has 2 analog outputs but they are off the back of the controller. I have one wired to the MS harness and wired the other to the factory harness where the factory sensor plugged in. I actually clipped the wires from the back of the sensor and then wired the wideband output (one of the red wires) to the signal wire of the factory sensor plug. This means the wideband will provide both the factory ECU and the MS. This has made troubleshooting more efficient and reasonable. I just swap the factory plug and hook up the vacuum line and plug in the MS.. If you notice the pink wire in the pic with the MS, that is the lamda wire for the MS which then has a tad bit of a loop hanging down in the pic of kyle plugging into the relay panel.

Now on to the update.... I switched from using the MS3 to control the idle valve to using the MS3x Idle control and realized I had hooked up ms3/28 but not to vw pin 23. I ran ms3/28 over to the the ign switch leg of the terminal block off the relay panel. After switching to the MS3x Idle control, it still was not working out. That night I tried all kinds of PWM setting but couldn't seem to get an idle at all. Last night we were at it again and right off I re calibrated the TPS, Changed my required fuel value, and make a complete change to the fuel ve table. The coolant temp was 137 at this point but I decided to turn the key anyways and sure enough it fired right up and idled.

Now on to the next wave of steps. My idle is a tad high (100 - 150 RPM) so I need to figure out whats up with that. On to the next step, it acts like there is a fuel cut at 2200RPM when just reving it. Idle is sitting at a pretty steady 14.4 with minimal fluctuation of up to maybe 14.7. The final minor issue currently is the AFR guage in MS is .4 higher than the innovate AFR guage but I have a pretty sure cause for this and I just need to fix it. In most MS installs, they say to ground the MS to the engine. I grounded it at the battery which I believe is still ok. The issue is that I grounded the Innovate under the dash which they tend to show the difference when not grounded at the engine. I will actually move the ground to the battery where the MS is but highly considering the switch to a ground terminal block by the battery with a single short 12GA wire to the battery. Either way I believe that grounding the Innovate at the same location as the MS will take care of the output difference. Finally after these things are worked out, it will be time to tune the cruise followed by WOT tunning. and then the fine tunning will follow it. I will post some pictures later once I get them online.

[sportbike]

I've seen offsets from Innovate controllers from their digital outputs to the analog outputs as well. Seems they are fairly sensitive to loading on the analog outputs. Assuming the digital out is as "good as it gets", I've seen where folks tweak the transfer function to make the analog numbers match the digital.

[its_34]

I've seen offsets from Innovate controllers from their digital outputs to the analog outputs as well. Seems they are fairly sensitive to loading on the analog outputs. Assuming the digital out is as "good as it gets", I've seen where folks tweak the transfer function to make the analog numbers match the digital.

It would be nice to answer to the "which is right?" question. I have a Innovate now, but haven't run it yet. In a previous car, I had an AEM unit and it read .2 to .5 richer than the dyno sensor while tuning. I took the conservative route and followed the dyno.

[Jason]

My Innovate was reading .6 lower than the MS but it is correct to follow the Innovate Guage.. Imoved the ground for the Innovate to the same location as the MS tonight and now it is only .1-.2 lower than the MS. There is still another wire I could try moving and I am probably going to move all the grounds from the battery to the engine anyways so we will see. Even with a .1 to .2 difference, I can live with that.

While trying to work out the cutout issues at ~2200 RPM I was also tuning Idle and trying to get it down near 850 as its sitting ~1000. I got the AFR to sit at a solid 14.4 and pretty stable at about 1000 with adjusting ign timing and the fuel map but could not get it down to the 850. After getting the pretty stable idle and the AFR where I want it, I realized that for some reason the tune had lost the setting for my 4 bar map sensor that I had modified the MS with. I reset the map sensor calibration which then meant I had to go change my fuel map and such again but it did clear up my response issue and the fact I wasnt reaching the expected values in the map but I thought that may have been because we were not driving. Now the map sensor is correct again and the response is great and the idle is looking good still at ~1000 RPM though.

I moved on to capturing some logs of the cutout problem which I thought maybe some timing adjustments may help however they did not. After diagnosing the logs, I may still have issues with my crank sensor it seems.

This first picture shows the log right before a cut.


Notice the key points of this issue outlined by Blue Boxes.


Now a shot as a cut happens.


Again the key points outlined in Blue


If you notice the Lost Sync count increased. Also looking at this, the SecL number increased. With both of these numbers increasing, I know the cut was not an ECU reset or at least can assume that. This indicated a problem with the crank sensor and you will notice in Blue again that the error reason is code 2. Code 2 actually works out to "missing tooth at wrong time".

I guess next will be working on figuring out this issue now. I am going to start with adjusting a POT on the MS main board but we will see what it is.

[sportbike]

I haven't really followed MS for awhile, but VW beetle with the 2.0L turbo engine in it guy noted that he ended up tweaking the pots for the crank sensor to get his to quit dropping the signal. Not really sure what he actually did, or what the pots even do exactly.
Screw is, give them all a tweak and see what it does...

[sportbike]

As for the idle, I'd check to be sure you don;t have any vacuum leaks or a crack in a TB boot or similar.
Assuming you are using some sort of ISC motor, is it fully closed and the idle is still too high? Maybe the thottle plate is hanging open a little?

[Jason]

Idle is dead perfect on the factory ECU. My idle valve is set to PWM Warmup only so it opens the tb 0% at 160 deg F or higher. I just need to work through it still.

In regards to the crank sensor, could play with one of the POTs on the JP1 jumper but that still may not fix it. Today I have spent a little bit of time and modified the MS to what should correct the problem. At idle the tooth logger is perfect and there is no issue with it.

So today the mod started by getting some resistors.


and tear down again..


locate JP1



Take off jumper


And solder resistor as a jumper between pins 1 and 2



Next is to try it and make sure that resolves the sync loss at ~2200 RPM..

I am also thinking about ordering a knock module to mod my MS3 with so that I can hook up my knock sensors and tune for those before I start running the new motor with boost.

[sportbike]

here's what the VW guy I referred to did to his. Not sure any of it is relevant to your application, although the diode swap may be useful.

"Replaced D24 (1N4001 diode) with a faster recovery (1N4148 diode) turned pots both ccw then turned them cw until I got a clean lower rpm signal. Test ran and had a really bad stutter 4-5500 rpm .changed plugs from Super 8 copper to iridium ( both resistor type) and then it revved smooth the whole rpm range.The ckp sensor has a .010 shim (I believe when the crank was magnafluxed they didn't fully demagnetize it and it has a magnetic field causing interference). The AFR issue ended up being a shorted relay that was dropping voltage to the pump intermittently."