Drag Week 2012 - Project GIRTST 71 Mustang (get it right the second time)

[squirrel]

Looks like a good solution to the heat problem.

[CDMBill]

The trigger wheel (EDIS) project has turned out well. Assembled on the Innovators West damper the wheel looks like this



With the damper installed back on the crank snout I needed to re-set the timing pointer so I could then verify the timing in the EMS with the actual timing based on the spark event. I used a Comp Cams Piston stop tool to mark the damper on both sides of TDC. As the picture shows the stop points were not at the same number of degrees on either side of 0 or TDC on the damper as indicated by the timing pointer. It had been knocked completely out of place by one of the disintegrating P/S belts earlier in the year. It been moved by about 6 degrees. With the pointer aligned, I pre-oiled the motor by spinning the external oil pump, reinstalled all the pulleys and belts and radiator etc. and I fired up the car. It started just fine and after warm-up settled in at idle. I used my ancient CalComp timing light to check the actual timing vs. the 26* I have in the EMS. It showed 30 on the damper under the light.

To adjust for that you change a value known as trigger angle in the software. The EDIS wheel is known as a 36-1 wheel meaning it has 35 teeth and a missing tooth. (you can see the missing tooth location in the first picture) When the sensor sees the skip that tells the EMS where it is relative to TDC so it can tell the coils to fire at the right time. There is no distributor so the only place the psark jumps a gap is at the plug. My wheel is set to have the missing tooth approximatly 7 teeth before TDC and is located at rough 7 o'clock because of the sensor location which is at roughly 4 o'clock.

It was set at 64* degrees previously with the old trigger wheel assembly. To retard the initial timing I added four degrees to the trigger angle to a total of 68* and the timing showed that I now had 26 degrees intital on the damper matching the value in the software timing map.









The pivot arm I fabricated for the GM LS1 sensor allows me to easily set the air gap, mine is set to .029", between the rotating EDIS wheel and the sensor while keeping the relationship between the sensor and the wheel perpendicular.

[Bob Holmes]

Nicely documented Bill, thanks.

[CDMBill]

The EFI update went well this weekend with Scott Clark aka DieselGeek coming out to cool temps in SoCal for the three day thrash.

We made a completly new harness for the injectors, sensors, power supplies etc. This was to ensure a clean slate install after the last five years on the old set-up and to address a persistent issue with voltage spikes at WOT. The other reason is the MS3X supports sequential control over the injectors vs. the batch fire on the old unit and therefore the injector harness has to have eight individual trigger wires. The MS3X also has the ability to drive the 2 four tower watsed spark coils I currently run or eight individual coils which I will eventually transition to when we are ready for individual cylinder spark timing. The wasted spark coils require an out board Igniter and I have used this one form EFI Source for the last five years. It has four op-amps attcahed to massive heat sinks and has been trouble free. The two four tower coils are in essence two coils in each module and they draw 20 amps each.



The other major issue we had to address is that the MS3X is set up natively to drive high impedance low current OEM type injectors in sequential mode, but not the higher current RCI low impedance 96 lb/hr peak and hold injectors I currently have on the car. That meant either getting 8 new high impedance injectors or finding a device to provide additional current and the ability to lessen the 'deadtime' on the RCI's. New high flow injectors high Z injectors would run well over $1000 so we went bloack box device route for $269 which is the VersaFueler from Acceleronics.



The VersaFueler is wired in series between the ECU and the injectors. We directly wired the injector outputs from the ECU harness using new high tech crimp/heat shrink/glued butt connectors. I know you are thinking "butt connectors?" are these guys idiots? Interestingly the thinking on making these connections has evolved and the high end engine managment systems like MOTEC and McLaren specifically prohibit doing the old solder and shrink wrap approach. So using the fancy. $.25 apiece butt connectors with a Klein crimping tool and a heat gun we went to work. Where the harness goes through the firewall we used older GM WeatherPak style connectors. These things are a PITA too but the MSD crimping tool I bought about ten years ago really helps make them secure.

We added the tooth gear reading wheel speed sensors which are US made Cherry P/N CH414N from DigiKey to the passenger fornt and back wheel. They read the wheel studs off the back side of the hub and the axle respectively which keeps them out of harms way as much as possible as well as making the senosr mount fabrication really easy.







For sequential EFI you also have to add a cam postion sensor. We used the same Cherry sensor and fabbed up a mount so that it would read a tab we bolted to the cam gear on the belt drive. We learned the hard way that the tab needs to be made of pretty stout material as the first one broke off during the first WOT test. The second tab is cut from a scrap piece of 1x1x.125 angle iron.

Even more interesting was that when the first tab failed the ECM switched over also seamlessly to semi-sequential mode while we were going about 85 on the test course aka the toll road allowing us to drive home. We had done a slower pull up to around 130 without incident but the acceleration rate at WOT was the difference. The analytics in the MS3X are awseome as we could see the 36-1 trigger wheel signal was still on the laptop screen, but the once in 720* cam sensor signal had disappeared. Fortunatly the tab didn't take anything with it when it left the engine compartment.



The other really neat addtion in the MS3X is the SD card datalogging feature. The ECU has a card slot in front next to the USB port (sorry for the crap picture) and offers the ability to start and stop the datalogging with a simple momentary switch. It will also drive an LED which pulses when you are successfully in datalogging mode and then blinks quickly to confirm the end of the log. We made a little extension harness as I wasn't yet sure where to mount the button and the LED. I was able to log all three test runs at Irwindale Saturday night without having to have the laptop in the car. Between rounds we reviewed the logs, compared notes on hwat we saw and heard and made tuning adjustmenrts.



I sacrificed my glove compartment to house the MS3X, the VersaFueler box and to have a place to mount the external MAP sensor we are using to measure crankcase vacuum as well as the relays and wiring busses for ground, 5V+ and 12V+ that are on the underside of the mounting plate. This alsp makes connecting the laptop easy, aswell as accessing the SD card and the other ports on the ECU. The harness is long enough so the mounting palte can be pulled out for service or testing if/when something fails.

The best part of this was that the car started on the first try and other than running out of gas while doing the initial tune and having to coast nto a gas station we had no issues. The mysterious voltage spikes that had caused trouble before are gome and I attribute this to having the ECU and sensors on a complety isolated circuit form all the nosisy stuff like injectors, injector drivers, coil drivers, coils, fans, pumps etc.

[milner351]

Well done! Great progress in a weekend. Sounds like the newest generation of MS is worth the price.

[TheSilverBuick]

Great write up Bill! I really really really really like using back of the stud or screw head for the wheel sensor.

[Scott Liggett]

You dont need a glove box. Your big stang has all kinds of places yo store junk.

Cool to see it run again. And, hang out again.

[CDMBill]

Great write up Bill! I really really really really like using back of the stud or screw head for the wheel sensor.

I can't take credit for the idea. The guy who is doing fabrication and wiring on the LSR Charger Daytona da 'Geek has been tuning and wrenching on did it on that car and we copied the concept.

Funny story: I thought it might work well to read the signal off the 10 bolts that attach the rotor to the hat and after putting it in place and spinning it around we got nothing. So it suddenly dawns on us, are those fancy tri-wing fasteners steel? Nope they're non-ferous stainless, back to the 5 wheel studs which actually allows us to bury the sensor a bit more out of the way of road debris.

[hauen]

Hmmm, with adding two more wheel speed sensors, you could build an MSGPIO and create your own ABS and traction control too! Don't know where you'd put yet another box though.

[CDMBill]

ABS would require an ABS pump, distribution system and other stuff to pulse the brakes, something I wouldn't do on this car. As to traction control, well that idea idea is just silly and wouldn't work and we would never consider it and please don't bring it up again.

It would be as if we could control launch timing or pull fuel and spark like a two step or manage AFR's during WOT at each each RPM window or gear change down track. Of course we can't do those those things.

You are a smart guy. I aren't.

[hauen]

that idea idea is just silly and wouldn't work and we would never consider it and please don't bring it up again

Forget I ever mentioned it. ;)

You are a smart guy. I aren't

I'm the guy who's having my ass handed to me by a gas tank replacement while you're doing golly-gee-whiz-hi-tech-bang-zoom stuff. I think my reputation as the dumbest guy in the room is safe.

[CDMBill]

Here are a few more photos that show some of the other parts that encompass the EFI system. I mentioned the cam sensor, this shows the little bracket we fabbed and the relationship to the tab we added to the cam gear of the belt drive. I remain nervous about the balance of the cam gear which most people say isn't material but I will add an additional counterweight on the opposite side anyway unless someone talks me out of it.





I also mentioned the four tower coils which come from Pantera EFI. They also make the IGN-1A COP coils many of you have read about. These are in essence two IGN's in each module. Spark energy has yet to be an issue and as you'll gather, these operate in wasted spark mode meaning that each pair of coils fires twice in the firing order. For example, that would be lighting off the first cylinder in the firing order and harmlessly firing the fifth hole at the same time while the piston is nearing the top of the exhuast stroke. I mounted them where the distributor would normally be to keep the plug wires easy to access for spark plug changes and to keep them off the valve covers to make checking lash easy too.





While doing the conversion we simply bundled wires like these sensor leads for the Manififold Air Temp (inlet Air Temp or MAT), Coolant Temp (CLT) and Throttle Postion Sensor (TPS) and Cam Position Sensor (CAM). I'll be going back now that everything works and has been debugged and wrapping them up and making them as neat as possible while still being servicable and easy to remove should more major repairs ever become necessary. The MAT sensor attaches to the air cleaner base and needs more individual length so the air cleaner can be moved aside without disconnecting the MAT sensor. If you run the car with that sensor disconnected the default mode is -40* and it'll run rich as hell. Ask me how I know. You can also see the dual four barrel throttle bodies under the air cleaner base.



Easily the worst part about EFI is all the wires. While we don't use all of them, the back of the MS3X box has two 37 pin connectors and most of those connections have to go through the firewall to injectors, ignition and sensors. You have to be able to remove both the ECU, injector driver, coil driver and the harnesses for use on an engine dyno and for on the road diagnosis and repair. So in my case you end up with a wad of WeatherPak connectors for each set of functions. Here you see the two injector bundles (5 wire), the sensor bundle (5 wire) wheel sensor (3 wire) and a bunch of others including the currently unused nitrous fogger wiring.



Pre-clean-up this is what things look like behind the the ECU and injector driver box. The forest of wires includes the two AFR guages, all the other bundles shown above plus the fuel pressure guage/sensor, EMS power, injector power, other sensors etc.



This will never be really pretty but everything is labeled and will be tie wrapped and secured out of the way of big feet. Stock bodied cars with roll cages can be a bear to work in and around. I use split wrap and tape where the wiring bundles go through the fire wall grommets and cold shrink wrap tape for the wire looms to keep out moisture and avoid the risk of exposing wires to bare meal..

[Bob Holmes]

As to traction control, well that idea idea is just silly and wouldn't work and we would never consider it and please don't bring it up again.

It would be as if we could control launch timing or pull fuel and spark like a two step or manage AFR's during WOT at each each RPM window or gear change down track. Of course we can't do those those things.

You are a smart guy. I aren't.

[SuperBuickGuy]

ABS would require an ABS pump, ditribution system and other stuff to pulse the brakes, something I wouldn't do on this car. As to traction control, well that idea idea is just silly and wouldn't work and we would never consider it and please don't bring it up again.

It would be as if we could control launch timing or pull fuel and spark like a two step or manage AFR's during WOT at each each RPM window or gear change down track. Of course we can't do those those things.

You are a smart guy. I aren't.

it's not like you know anyone who could program such a thing and even worse, that'd be something only a lawyer would do (look for exceptions to rules)

[CDMBill]

I don't know what you guys are talking about. And even if I did writing about it would violate a tuner/client privledge.