Re: 1984 Jaguar XJS V12 5.3L HE (11.5:1 CR)
Posted: Wed Mar 21, 2012 1:00 am
Stephen,
Not a lot of major changes, I'm still running the CEI distributor (locked-out vacuum advance) with the VR sensor output running to the Megasquirt. This is not ideal compared to Philip's EDIS setup due to the noise picked up by the VR sensor wiring and the jitter introduced by the mechanical advance, but it has been reliable enough and passes the visual emissions inspection in CA. The software actually does a decent job of generating a stable timing reference from the noise that passes for a timing signal, since adjusting the hysteresis pot appropriately I haven't noticed any missed ignition trigger events.
I don't think I'm going to make any more changes to the setup other than to try to upgrade to the MS/2 Extra code, I've hit the point of diminishing returns if I'm unwilling to make major changes to the ignition system. At this point I'm mostly optimizing timing, cranking/warmup and transient fueling. I was trying for a while to reduce the amount of TPS-based acceleration enrichment by using X-Tau, but the MAP sensor response is too sluggish and causes problems when not completely warmed up.
Here are some things to keep in mind if you're looking to ease into the conversion gently, they all tripped me up the first time:
- Check all the ground and power connections and/or add a new one when you install the MS. I had tons of issues with ground loops or noise due to flaky connections. I cleaned up all the major ground points and used copper anti-sieze on all of them. I also replaced the Lucas alternator with a GM AD230 unit from a Chevrolet truck, which helped quite a bit with voltage fluctuations. A portable oscilloscope is an incredibly handy tool when trying to diagnose these types of issues.
- Make sure the stock ignition system is working properly before you start the conversion if going with fuel only. The centrifugal advance mechanism tends to seize up due to heat-induced lubricant failure, and the vacuum advance capsule tends to leak. The stock ignition is actually not all that bad if in good shape and the vacuum regulator is still working, although I ran mine with the vacuum advance connected straight to the throttle edge tap to improve idle quality until I finished the ignition conversion. Also make sure both coils are good and the connections are clean if you're not running the single coil setup from the XJ-RS.
- The Bosch CTS and ATS response curves are different than GM sensors, you'll need to calibrate accordingly.
- The vacuum line running to the rear of the vehicle for the MAP sensor is often partially clogged and will cause all sorts of fun with transient fueling. Make sure it's clean, I also added a filter at the back with a small bleed hole to keep the line clear.
- The Lucas ECU's on the XJ-S referenced the air intake temperature sensor to +5v. This will need to be changed to ground, I installed a jumper on the engine wiring harness so I could switch between the two systems until I'd gone far enough that it wasn't possible to swap back to the Lucas ECU in the event of a failure.
- The Lucas TPS sensor is junk and often fails open circuit over part of the range. This caused all sorts of issues with the MS-II, but not a lot with the original Lucas ECU.
- The fuel management unit that feeds the trip computer is driven off one of the the "hold" injector transistors in the 6/16CU ECU. If you use the original injectors and drive them viak peak/hold the trip computer's fuel calculations will be way off.
- If you change the fuel rails and injectors, be aware that the hot start vacuum switch or ATS bypass switch is there for a reason. Hot restarts can be difficult due to fuel vapor buildup in the rail, I had to add a significant amount of fuel when the intake air temperature is over 170 degrees F or so, otherwise it was a pain to start. Note that I typically have my A/C compressor disabled to improve fuel economy, which means the fuel cooler isn't functional.
- Make sure both thermostats are working properly and both banks have the same normal operating temperature if only using a single O2 sensor. I was having all sorts of odd issues with rough idling and performance at low speed, and it turned out that one thermostat was stuck closed resulting in a "normal" operating temperature of 240 degrees F for that bank and 205 for the other bank. It ran just fine with the Lucas setup due to the dual O2 sensors though. I currently run only a single O2 sensor for fueling but monitor both, and when the engine is operating properly there is no noticeable difference between them.
Note that the site I've posted the *.msq files on has changed, see the signature in this message for the new URL. I'll try to upload the photos sometime soon.
Cheers!
- Ross
Not a lot of major changes, I'm still running the CEI distributor (locked-out vacuum advance) with the VR sensor output running to the Megasquirt. This is not ideal compared to Philip's EDIS setup due to the noise picked up by the VR sensor wiring and the jitter introduced by the mechanical advance, but it has been reliable enough and passes the visual emissions inspection in CA. The software actually does a decent job of generating a stable timing reference from the noise that passes for a timing signal, since adjusting the hysteresis pot appropriately I haven't noticed any missed ignition trigger events.
I don't think I'm going to make any more changes to the setup other than to try to upgrade to the MS/2 Extra code, I've hit the point of diminishing returns if I'm unwilling to make major changes to the ignition system. At this point I'm mostly optimizing timing, cranking/warmup and transient fueling. I was trying for a while to reduce the amount of TPS-based acceleration enrichment by using X-Tau, but the MAP sensor response is too sluggish and causes problems when not completely warmed up.
Here are some things to keep in mind if you're looking to ease into the conversion gently, they all tripped me up the first time:
- Check all the ground and power connections and/or add a new one when you install the MS. I had tons of issues with ground loops or noise due to flaky connections. I cleaned up all the major ground points and used copper anti-sieze on all of them. I also replaced the Lucas alternator with a GM AD230 unit from a Chevrolet truck, which helped quite a bit with voltage fluctuations. A portable oscilloscope is an incredibly handy tool when trying to diagnose these types of issues.
- Make sure the stock ignition system is working properly before you start the conversion if going with fuel only. The centrifugal advance mechanism tends to seize up due to heat-induced lubricant failure, and the vacuum advance capsule tends to leak. The stock ignition is actually not all that bad if in good shape and the vacuum regulator is still working, although I ran mine with the vacuum advance connected straight to the throttle edge tap to improve idle quality until I finished the ignition conversion. Also make sure both coils are good and the connections are clean if you're not running the single coil setup from the XJ-RS.
- The Bosch CTS and ATS response curves are different than GM sensors, you'll need to calibrate accordingly.
- The vacuum line running to the rear of the vehicle for the MAP sensor is often partially clogged and will cause all sorts of fun with transient fueling. Make sure it's clean, I also added a filter at the back with a small bleed hole to keep the line clear.
- The Lucas ECU's on the XJ-S referenced the air intake temperature sensor to +5v. This will need to be changed to ground, I installed a jumper on the engine wiring harness so I could switch between the two systems until I'd gone far enough that it wasn't possible to swap back to the Lucas ECU in the event of a failure.
- The Lucas TPS sensor is junk and often fails open circuit over part of the range. This caused all sorts of issues with the MS-II, but not a lot with the original Lucas ECU.
- The fuel management unit that feeds the trip computer is driven off one of the the "hold" injector transistors in the 6/16CU ECU. If you use the original injectors and drive them viak peak/hold the trip computer's fuel calculations will be way off.
- If you change the fuel rails and injectors, be aware that the hot start vacuum switch or ATS bypass switch is there for a reason. Hot restarts can be difficult due to fuel vapor buildup in the rail, I had to add a significant amount of fuel when the intake air temperature is over 170 degrees F or so, otherwise it was a pain to start. Note that I typically have my A/C compressor disabled to improve fuel economy, which means the fuel cooler isn't functional.
- Make sure both thermostats are working properly and both banks have the same normal operating temperature if only using a single O2 sensor. I was having all sorts of odd issues with rough idling and performance at low speed, and it turned out that one thermostat was stuck closed resulting in a "normal" operating temperature of 240 degrees F for that bank and 205 for the other bank. It ran just fine with the Lucas setup due to the dual O2 sensors though. I currently run only a single O2 sensor for fueling but monitor both, and when the engine is operating properly there is no noticeable difference between them.
Note that the site I've posted the *.msq files on has changed, see the signature in this message for the new URL. I'll try to upload the photos sometime soon.
Cheers!
- Ross