1984 Jaguar XJS V12 5.3L HE (11.5:1 CR)

Post Reply
macaddict
MegaSquirt Newbie
Posts: 32
Joined: Tue Feb 17, 2009 3:42 pm
Location: Burlingame, CA

1984 Jaguar XJS V12 5.3L HE (11.5:1 CR)

Post by macaddict »

Hi all,

I just recently converted my 1984 Jaguar XJ-S to Megasquirt after getting tired of tweaking with the original Lucas 6CU ECU. I wanted to leave the original wiring as stock as possible so I could go back to the original Lucas ECU if I had to (smog checks, fried Megasquirt, etc.) so I ended up cannibalizing a junk Lucas ECU for the case and connector, and grafting the Megasquirt-II v3.57 onto it:

http://crosstalk.atomz.com/~ross/jagstu ... 999857.jpg
http://crosstalk.atomz.com/~ross/jagstu ... 066320.jpg
http://crosstalk.atomz.com/~ross/jagstuff/megasquirt/ (tuning files)

The Megasquirt was purchased fully assembled from DIY Auto Tune, along with an adapter harness and some GM coolant temperature sensors that I didn't end up needing (used all the original sensors, just had to change the wiring for the IAT sensor so that it referenced to ground and not TPS vref.) The modified ECU case tucks into the original location in the trunk, and aside from a serial cable snaking out of it and heading into the passenger compartment it looks completely stock as it's mostly hidden by the carpeting in the trunk. The stock injector resistors are bypassed and the Megasquirt is running in peak and hold mode with a 1.1ms open time and the pwm duty cycle set at 21%. The firing mode is set to 4 squirts per cycle alternating, with both cylinder banks running off of one VE map and the passenger side 02 sensor. I connected a WB 02 sensor to the driver's side O2 sensor bung and used it for tuning, but haven't bothered to switch to dual VE table support with two O2 sensors as the difference in AFR between banks when working properly is negligible.

Right now the system is handling fuel injection only with the timing signal coming from the output of the stock Delco HEI ignition. I have the ignition output of the Megasquirt connected to the full load signal line coming from the vacuum switch mounted in the engine compartment, that way I can drive a MSD ignition box or possibly another HEI module to control ignition timing at some point.

A future project may be to move the Megasquirt ECU to a location underneath the dash (there's a nice spot underneath an access panel in the engine bay where a vacuum reservoir currently resides) to shorten the cable runs and remove the need for a vacuum pipe running the length of the car, but that would make the installation decidedly non-stock.

Overall, I'd have to say this is the best modification I've done so far on this car. The throttle response has greatly improved, the engine starts and idles better than it ever has, and it pulls better at WOT as the original ECU was running way too rich at full load. It can also break the wheels loose on dry pavement, which is something it never could do before no matter how well it was running. It feels like a completely different car, thanks Megasquirt!


- Ross.
1984 Jaguar XJ-S HE (Rosie III) - Megasquirt-II v3.57 fuel+spark
http://crosstalk.splunk.com/~rdykes/jagstuff
Philip Lochner
Helpful Squirter
Posts: 45
Joined: Thu Feb 16, 2006 5:18 am
Location: Centurion, South Africa

Re: 1984 Jaguar XJS V12 5.3L HE (11.5:1 CR)

Post by Philip Lochner »

Well done Ross!!!

Could you post your MSQ and tell us which fw you are running please? I have a friend who also has an 11.5:1 HE who would be interested in your MSQ.
Kind regards
Philip
My MS projects: viewtopic.php?t=20086
macaddict
MegaSquirt Newbie
Posts: 32
Joined: Tue Feb 17, 2009 3:42 pm
Location: Burlingame, CA

Re: 1984 Jaguar XJS V12 5.3L HE (11.5:1 CR)

Post by macaddict »

Philip,

Sure thing, I'm running the MS2-MicroSquirt BG 2.888 code and the msq files can be found at the link below:

http://crosstalk.atomz.com/~ross/jagstuff/megasquirt/

I'm running the latest version of the file in that directory, I'll post new ones as I continue to tweak the VE tables. Right now it's running fine but the VE table is still a little rich in places (mostly below 55 kPa and 4000 RPM.) I also have it biased a bit rich at idle to help idle quality (on my engine it's about 13.8 - 14.1 when warm) and I have it set to go too rich at WOT near redline to help keep the rear end from breaking loose during the 1-2 shift. The first night out after the install I managed to get it sideways. =)

I also need to find a way to interface with the trip computer as it can't deal with the lack of a "hold" injector signal. Possibilities include modifying the code to drive one of the spare outputs low during the entire injection event, building an interface circuit to provide a square wave signal, or tapping into the flyback damping circuit if that's active the entire time the injector is open. I plan on replacing the injectors and rail with saturated coil 17 lb. Delphi disc injectors and custom fabricated rail at some point, which would solve the problem as there would be no PWM involved.

Your setup was what inspired me to give it a shot, although I took the lazy way out and stuck with the Lucas ignition system to keep the California smog check inspectors at bay. I want to use the existing 4-pin HEI module to interface the VR sensor in the distributor to the MS-II and drive a second 4-pin HEI module to fire the coil (ideally mounted next to the second coil in front of the radiator where nobody will notice it.) If it works, it'll allow me to switch back to the original ECU by connecting the coil directly to the first HEI module. I was considering leaving the mechanical advance in place (since I just rebuilt the darn thing!) and just using the MS-II to emulate the vacuum advance. Not sure if that'll cause problems though as the mechanical advance kicking in during acceleration could affect the MS-II's interpretation of engine RPM. I guess I'll find out this weekend!


Cheers,

- Ross.
1984 Jaguar XJ-S HE (Rosie III) - Megasquirt-II v3.57 fuel+spark
http://crosstalk.splunk.com/~rdykes/jagstuff
Philip Lochner
Helpful Squirter
Posts: 45
Joined: Thu Feb 16, 2006 5:18 am
Location: Centurion, South Africa

Re: 1984 Jaguar XJS V12 5.3L HE (11.5:1 CR)

Post by Philip Lochner »

macaddict wrote:I was considering leaving the mechanical advance in place (since I just rebuilt the darn thing!) and just using the MS-II to emulate the vacuum advance.
My recommendation is that you MUST leave the mechanical advance in place. Otherwise you may find the spark jumping to the wrong plug terminal much too quickly as the width of the rotor now has to make up for both mechanical and vacuum advance. I did the sums once (forgotten now) and found that due to the narrow rotor tip, the amount of vacuum advance was quite limited.

The idea then is to set the dissy such that the MS controlled spark will occur at a time (using the timing offset) when the trailing edge of the rotor lines up with the terminal at your point of least advance leaving the remaining width of the rotor for maximised MS controlled vacuum advance. This might mean getting hold of another dissy cap that you drill a hole into so you can see the alignment of the rotor with your timing light.
Kind regards
Philip
My MS projects: viewtopic.php?t=20086
macaddict
MegaSquirt Newbie
Posts: 32
Joined: Tue Feb 17, 2009 3:42 pm
Location: Burlingame, CA

Re: 1984 Jaguar XJS V12 5.3L HE (11.5:1 CR)

Post by macaddict »

Ah, good to know. I was wondering what the probability of the spark jumping to the wrong terminal was, now I know it's fairly high! I do have an old cap and rotor lying around I can drill into, I'll try that and see what the rotor to cap terminal relationship looks like with the stock timing settings and determine how much I'll have to adjust it. I know the HE vacuum advance capsule provided more advance than the pre-HE capsule, so I hope there's enough room in there for the timing advance the MS-II will need to add.
1984 Jaguar XJ-S HE (Rosie III) - Megasquirt-II v3.57 fuel+spark
http://crosstalk.splunk.com/~rdykes/jagstuff
StevenD57
MegaSquirt Newbie
Posts: 30
Joined: Fri Oct 14, 2005 12:14 pm
Location: Phoenix, AZ
Contact:

Re: 1984 Jaguar XJS V12 5.3L HE (11.5:1 CR)

Post by StevenD57 »

Are you planning on any additional modifications like crankfired ignition or going to a more modern idle control system (PWM or stepper motor IAC)?
Going to crankfired ignition would solve your distributor cap & spark jump problems. I know from my frustrations with the crummy factory AAV idle control thing I would switch to one of these other idle control mechanisms in a moment when I megasquirt one of my V12 cars.
Jaguar E-Types, modified Merkur XR4Ti, Jensen-Healey
Linux = no virii or spyware, EVER
macaddict
MegaSquirt Newbie
Posts: 32
Joined: Tue Feb 17, 2009 3:42 pm
Location: Burlingame, CA

Re: 1984 Jaguar XJS V12 5.3L HE (11.5:1 CR)

Post by macaddict »

Oh god, yes. =) The stock idle control mechanism on the v12's is a piece of crap, and even when the ignition system is working properly (and right now it isn't) there's quite a bit of jitter in the timing signal. I have a Ford PWM idle valve lying around somewhere I was going to use , I just need to find it, build a mounting plate and run a wire back to the ECU (or maybe just use the "B" bank O2 sensor wire as I'm not running dual VE tables.)

I think at this point I'd like to go with a crank sensor for the timing reference but leave the distributor and legacy advance mechanisms in place to allow me to revert to the stock ECU without too much hassle. As much as I'd love to move to coil packs it probably wouldn't pass the visual emissions inspection in CA. I'm running an electric fan, so there should be plenty of room for a crank trigger wheel and sensor.
1984 Jaguar XJ-S HE (Rosie III) - Megasquirt-II v3.57 fuel+spark
http://crosstalk.splunk.com/~rdykes/jagstuff
macaddict
MegaSquirt Newbie
Posts: 32
Joined: Tue Feb 17, 2009 3:42 pm
Location: Burlingame, CA

Re: 1984 Jaguar XJS V12 5.3L HE (11.5:1 CR)

Post by macaddict »

Hmm, so it seems that much of the timing jitter and the occasional misfire I've been seeing may be bad plug wires. When I hook up a timing light it looks like four of the plugs (A2 - A5) on the A bank are firing inconsistently. Fiddling with the plug wires affects the stability of the timing and the misfiring, and the resistance of those four cables varies considerably when you stretch them so it looks like Lucas has struck again. Kind of annoying as the wires are only a few years and about 10,000 miles old. Ordered a spool of Taylor Cable bulk plug wire and the boots/terminals from Summit Racing, we'll see if that solves my immediate problem.

I think I'll ask Philip how he had his crank trigger wheel fabricated, it looks really slick. I'd like to do something similar as my original idea of using a Marelli crank damper and sensor won't work due to the trigger wheel only having three teeth (good thing I looked into it before buying one!) Apparently that setup used two sensors, one triggering off the flywheel for timing and one off the crank damper for cylinder synchronization.
1984 Jaguar XJ-S HE (Rosie III) - Megasquirt-II v3.57 fuel+spark
http://crosstalk.splunk.com/~rdykes/jagstuff
StevenD57
MegaSquirt Newbie
Posts: 30
Joined: Fri Oct 14, 2005 12:14 pm
Location: Phoenix, AZ
Contact:

Re: 1984 Jaguar XJS V12 5.3L HE (11.5:1 CR)

Post by StevenD57 »

Yes, but that three eared thing on the back of the Marelli damper unbolts and can thus be replaced with a 36-1 toothed wheel from a late nineties Ford Taurus.
The three ear Marelli timing wheel has the same thickness as the center portion of the Taurus 36-1 EDIS wheel which on most of those Taurus dampers I have found unbolts as well although some late model Taurus dampers had it cast in. The Marelli three eared thing is held on with three 1/4 bolts.
Jaguar E-Types, modified Merkur XR4Ti, Jensen-Healey
Linux = no virii or spyware, EVER
macaddict
MegaSquirt Newbie
Posts: 32
Joined: Tue Feb 17, 2009 3:42 pm
Location: Burlingame, CA

Re: 1984 Jaguar XJS V12 5.3L HE (11.5:1 CR)

Post by macaddict »

Ah, you mean from the 3.0L Vulcan pushrod engine:

http://www.napaonline.com/MasterPages/N ... ancer+Assy

Now all I need to do is get my hands on a Marelli V12 damper, a rebuilt one is under $600 which is probably worth it as it wouldn't hurt to replace the original one.
1984 Jaguar XJ-S HE (Rosie III) - Megasquirt-II v3.57 fuel+spark
http://crosstalk.splunk.com/~rdykes/jagstuff
Post Reply