T4 stock cooling--Crank trigger/CNP or COP using Megasquirt

[Piledriver]

That'll do it just fine.

You MIGHT be able to mount a VR/bracket in the top lip of the fan housing area (thin area that block space between the fan housing proper and the breather chimney)

[supaninja]

I'm goint to mock the crank/case/shroud up tomorrow and check that out. I have most of next week off so time is on my side.

[Piledriver]

Almost forgot... The Square is on the road, got it inspected this AM and spent several hours of quality time helping Megatune and then TS itself (after I registered it) massage the VE map.
Snorts pretty good for an almost stock 1.8.
(OK, 1.8 with 2L Porsche heads Yes heads are temporary)

I think I managed to finally seat the rings, did wonders for idle A/F ratio, I couldn't get it to idle decent over 11:1 until then.

TS works a LOT better/faster once registered... Unregistered TS is not just feature limited, it's handicapped in speed as well.
Between the live VE autotune and the faster updates etc I think I save the $40 worth of time just today.

[supaninja]

Dude thats awesome your finally got your rig back on the road. I still need to register my TS, thats on the list. I have a question for you Pile, where did you mount your IAT? If I can't mount mine on the end of the uni filter then i'll mount it to the plenum, just want to minimize heat soak. Anyways, you got a sweet setup and i'm glad it works, another type 4 in a type 3 success story.

[Piledriver]

Dude thats awesome your finally got your rig back on the road. I still need to register my TS, thats on the list. I have a question for you Pile, where did you mount your IAT? If I can't mount mine on the end of the uni filter then i'll mount it to the plenum, just want to minimize heat soak. Anyways, you got a sweet setup and i'm glad it works, another type 4 in a type 3 success story.

At the moment it's crammed wires first in the flex conduit I ran the ign cabling through.

It's about 1" from the air cleaner, in the engine compartment, so I imagine it's correct.

The next step in my setup is sealing the engine compartment so I can ditch the bellows.

[supaninja]

quick mockup of the motor, looks like a winner.



On a side note I'm going to have to get either a 411 or a 914 fan, or one of those bolt on timing plates for a bus. there is only one notch on my fan and its way too far to the left. Aside from that everything jives nicely. I'm going to drill and tap a couple of holes at 270deg on the back side of the shroud, then thread some button heads from the inside. I'll shim the sensor toward the block a couple of mm's. Its completely safe from the belt, and clears the motor bar and oil cooler too.

[Piledriver]

You know... You can put the marks in yourself... and know for sure they are in the right place.

[supaninja]

thats probably what i'll do.

[Piledriver]

After Supaninja drove it the other day as he passed through (great meeting you!)
...I finally Got Busy with this, more than just inspired by RonW
(First is of my "old" coil off HD twincam , Delphi with ion sense output, 2 coils per pack.)

HD-twincamcoil.JPG

The funny thing was the bolt spacing was the same for the (Delphi) LS2 coils, just needed some longer bolts and spacers, followed Rons lead here: (shown on a spare manifold for decent pics)

LS2-sideview.JPG

LS2-coils-2L-manifold.JPG

I simply added 4x 1K resistors off 5v to the outputs of the VB94xs in my driver box, std dwell, 4.9ms, I turned the accell compensation off for the moment. (std dwell works here as the VB94x circuit with the pullup "inverts" the signal, if you were driving direct logic, you would want "inverted"

She ran fine as soon as I remembered to plug the driver box back in after flashing to 3.1-pre-rc1 FW.

It still prefers to idle ~11:1, but I in a quick map session with autotune and MLV I found I can lean the rest of the map out some now in light load areas where anything >12.5:1 caused lean misfires before....The NGK triple electrode plugs are going back in, as the std coppers made made absolutely no difference.

[supaninja]

It was awesome driving your squareback, just the motivation I needed, thanks! My GSXR COPs were waiting for me when I got home from the trip. I was originally going to us COPs from a Ford, but I never like how big the tops were and how the stalks are made of a super soft/floppy rubber. The gsxr ones are a nice hard plastic stalk with a small coil on top, doesn't look like they will need to clamped down. They were cheap...real cheap too, they use the same connectors as my Ford COPs and VR sensor. Now they are made for the smaller 14mm plugs. Shouldn't be an issue, the 14mm plugs in my honda use the same thread size/length as a type 4, I just need to verify the heat range and make sure they are resistor plugs. I probably got 3 sets of good one step colder iridium plugs laying around from when it was turbod.

[Piledriver]

Was a blast seeing you too!

I forgot to mention--- the plug wires I used on the LS2 coils were the same Harley Davidson resistor wires I was using... The center pin/depth is the same, and the boot stretches fine, they're IIRC <$3 and come in singles in two lengths, the ones I have are almost a foot long, will be replacing them with the shorter ~7" type, eventually.

[Piledriver]

Bit of "missing" documentation---
The pin function is "black boxed" in the Vanagon wiring diagrams.

From an email I sent to DIYAutotune so Matt could update his page (edited for brevity and clarity)...
(Re:Vanagon hall sensor connector wiring)

Late Vanagon:(no vac can or advance mechanism)
pin1 ---Br/W gnd (from MS db37, preferably)
Pin2 --- Gn/W 12v hall output (needs pullup---active low)
pin3 ---R/W power/12v (I used the old ignition power wire as it was handy)

The switch is active low... goes high/open with the "tooth" in the sensor.
It pulls LOW when it is NOT blocked by the trigger wheel.

[Piledriver]

Minor update---

Still doing dailiy 80 mile commute, worked last 3 weeks 6 days/week, and weather has been garbage on days off.

Lost a voltage reg (solid state type) and full fielded alt while doing 70... BRIGHT headlights... killed the JAW, but it probably only toasted the LM317 used to drive the sensors heater, have some in stock, just need time to fix. Fortunately it was a lifetime warranty item from Oreillys, so I'm probably only out a buck and a little time to fix the JAW.

The MS was unfazed, although I did need to turn WBO2 closed loop back off. (the maps were pretty much nailed anyway using autotune on the commutes)
IIRC the ms2.2 and v3 boards feed unregulated 12v to the injector fet driver, which is a different chip than the 1.1 board== fet driver uses 5v. (I'm using external jbperformance 4 channel pwm board anyway)

Actually have a full 2 days off this week with a decent forecast, home to get the bn4 install finished and some of the other minor things like get the new rubber installed on both doors and trunk at least.

The MS continues to simply work.

[miniman82]

The MS continues to simply work.

But it's a garbage ECU, it'll never work in the real world!

[Piledriver]

It continues to run... and sequential DOES matter. A lot.

Here's an unspectacular "dyno" RWHP/Tq graph, courtesy of VDR 5.71.
(just copied/pasted the time and RPM columns from a MS burst log loaded in openoffice calc)
I still need to add a short ram tube, the TB is too short, should help the bottom end>midrange
I haven't figured out how to get VDR to plot <2500 rpm.

3rd gear pull,(1.26) 4.125 final, 24.25" tires, 2250 vehicle weight plus my 245 lb lardass, 20sqft frontal area (guess) and .37Cd. (reasonable estimate, a notch would be .34)

vdr-type3-1.8l.png