Jump to content

mikec

Moderators
  • Posts

    2162
  • Joined

  • Last visited

  • Days Won

    3

mikec last won the day on December 23 2014

mikec had the most liked content!

Profile Information

  • Interests
    Finish line at the Pikes Peak "Race to the Clouds"
  • Location
    Cypress California
  • Gender
    Male

Previous Fields

  • Zip Code
    90630
  • Model
    Starion
  • Type
    ESI-R
  • Model Year
    1988
  • Transmission Type
    Manual
  • Factory Color
    Palermo Gray
  • Interior Color
    Black
  • Status
    On the road

Recent Profile Visitors

3511 profile views

mikec's Achievements

Newbie

Newbie (1/14)

  • Week One Done Rare
  • One Month Later Rare
  • One Year In Rare

Recent Badges

15

Reputation

  1. What you are describing sounds like a classic case of the starter solenoid being weak, have a failed coil, or having lousy voltage from the ignition switch. Basically, the solenoid is an electromagnet with two coils. The two coils work together to "slam" the starter motor pinion gear into the engine flywheel, nudging the flywheel as necessary to get the gears to mesh. If the gear teeth happen to be lined up just right (i.e. side of the teeth hit directly instead of actually meshing the teeth faces) the starter gear needs a LOT of force to rotate the engine & flywheel. Poor electrical wiring & connections in the IGN circuit (battery -> IGN fusible link -> ignition switch input -> START position and ON positions -> factory alarm relay A-40X -> starter solenoid via a black+yellow wire) lead to voltage losses in the wiring which means the solenoids don't get enough input energy to do their jobs. Often when this happens, if you put the car into gear and leave the clutch pedal out while pushing/pulling the car just a little bit (i.e. rotating the engine yourself), that gets the gears to align and the starter "comes to life" because now the solenoids can finish their job. One end of the pushrod inside the solenoid moves the gear assembly to engage the starter to the flywheel, the other end closes a beefy switch to send power to the actual starter motor once the gears are meshed. If the gears never mesh... this switch never closes and the starter motor never runs. You just hear a "click" from the solenoid and gear smacking into the flywheel ring gear. First basic test is to attach a voltmeter to the small wire running to the starter solenoid - this is the wire from the ignition key. When trying to start the engine, this wire should read pretty much full battery voltage. If not, the solenoid is being denied working energy by crappy wiring. A common starter solenoid issue is for one of the two coils to fail. With only 1 of 2 coils working, the solenoid has much less force when trying to mesh the gears... it isn't able to "nudge" the engine/flywheel much. Ergo the starter gear never properly engages and thus the starter motor is never switched ON. If you remove the starter, there are simple tests you can do to check the solenoid. You'll need a 12 power supply or small 12 volt battery. You could use the car's battery but be very careful to avoid short circuits. And if the solenoid coil has failed by internally shorting out, using a car battery for testing will lead to a lot of sparks and likely melted test wires. At least wear gloves and have eye & face protection just in case. With the starter out of the car: (disconnect the battery before taking the starter out!) You'll see a braided wire strap going from the motor body to a stud connection on the solenoid. Undo that strap from the solenoid and bend it out of the way. This keeps the motor from running... which in turn keeps the starter from suddenly jumping around on your workbench! The solenoid post that had the fat wire to the starter motor is the "M" post. The flat connector tab where the car's black+yellow wire snapped into it (this is the wire from the ignition key START position) is the "S" post. The other wire/stud on the solenoid is the "B" (battery) post. The beefy wire directly from the battery "+" post connects to this one. "B" and "M are also the posts of the beefy switch inside the starter solenoid that runs the motor. When the solenoid's beefy switch is ON, "B" is connected to "M" to run the motor. Test procedure: 1. Securely clamp the starter solenoid+motor assembly in a bench vise or some other contraption that'll hold it securely. The assembly will want to move/jump pretty hard so you want it held down. If nothing else, vise-grips or other locking pliers clamped to a mounting ear on the motor and then held to the workbench (or garage floor) by a heavy weight is better than nothing. 2. Undo the braided wire from the "M" post as noted above. 3. Firmly connect the test battery/power supply "-" wire (typically a black wire) to the "M" screw post. 4. Touch the battery/power supply "+" wire (typically the red wire) to the "S" tab for a second or two. The solenoid should firmly snap the starter gear out. Remove the wire and test again... The solenoid action should be quite firm and positive. This tests the main "pull-in" electromagnet coil of the solenoid which is a common solenoid failure. Don't leave power applied for more than a couple seconds to avoid overheating the solenoid coil. When done, remove the battery/power supply "+" wire from the "S" post. 5. If that test is okay, move the battery/power supply "-" wire to the body of the starter (i.e. ground). 6. Now touch the battery/power supply "+" wire to the "S" tab again. The solenoid should extend the gear again though likely nowhere near as firmly as the pull-in coil test. This is testing the "hold in" coil of the solenoid; this coil keeps the starter gear engaged with the flywheel and keeps the beefy switch ON once the solenoid has fully moved to the engage position. If this coil fails, the starter motor will cycle between ON and OFF as you hold the key. Why? Because, once the solenoid moves via the pull-in coil and closes the beefy switch, the action of the beefy switch happens to disconnect the pull-in coil. So you need this second "hold in" coil to keep the solenoid engaged and the switch ON. On cars with auto transmissions, the "park-neutral safety switch" (aka "Inhibitor Switch" in MItsu's wiring manuals) is added to the wiring. This switch is ON (closed, passing electricity) when the transmission lever is in park or neutral to prevent starting the engine while the transmission is in gear. It's not uncommon for the mounting bolts for this switch to get loose on the vehicle rendering the switch useless. The switch has slots in its mounting holes so it can be adjusted... when the screws get loose the adjustment is lost. Once in a while the switch itself gets crappy: if moving/wigling the shift lever while holding the key in START makes the starter suddenly come to life this is the problem, it's very common on other brands of cars as well. Try wiggling in Neutral as well if Park doesn't work. If your car also shows symptoms of varying intensity interior and/or headlights (rapidly varying as you drive, not varying in sync with engine RPMs) then a lousy ground is also on the possibilities list. There is a beefy wire from the battery "-" post, to a bolt on the body "frame rail" below the battery, and then to the engine block. All three connection points MUST be solid. Corrision is common and must be removed. I like to use "star" washers on these connections. Star washers look like normal washers with thin metal ears facing inwards or outwards; these ears are twisted (like propeller blades) so their edges bite into whatever is sandwiching the washer. Loose starter mounting bolts also result in a lousy ground to the starter motor. Always clean the mating surfaces of the starter & engine when attaching the starter since that is the ground connection. mike c.
  2. The pusher fan does not automatically run when the a/c is on... it turns on only if the refrigerant pressure gets pretty high - indicative of a very hot day. Look for a small connector with black and black+yellow wires on the harness side, it'll be a sensor attached to the a/c plumbing near the receiver/dryer. That's the "high pressure" switch. Unplug the connector and jumper the harness connectors. That should make the pusher fan and engine-side secondary fan engage. The relay triggered by this sensor has blue+black, black+yullow, black, and blue+yellow wires. Relay A-24, one of the relays hanging on that bracket by the ignition coil. The fan configuration on StarQuests has two separate functions: * engine cooling * a/c operation There are separate sensors feeding separate relays for each function. For engine cooling: There are temp sensors on the bottom of the radiator; each sensor triggers a cooling fan relay (located by the air cleaner) to run the fan above it. However, a wire connection from the secondary fan (smaller fan) happens to cross over into the air conditioner's high pressure sensor so it triggers the a/c pusher fan relay & pusher fan as well. Thus, when the water temps get hot enough to trigger the secondary fan (near 3/4 on the dash guage) both the secondary and pusher fan should engage. For a/c operation: The climate control computer has its own relays to run the primary and secondary fans. They are basically in parallel with the cooling fan relays - e.g. either the cooling fan relays OR the climate control relays can feed +12volts to the fans. Sometimes both relays are engaged at the same time... just means the fan has two relays giving it 12 volts... either way, it runs normally. That a/c refrigerant pressure sensor feeds the A-24 relay to kick in the pusher fan only on very hot days (otherwise the a/c could over-cool leading to icing in the evaporator box and thus excessive on-off cycling). Because of that extra "if overheating" wire in the engine cooling function, this sensor will also trigger the secondary engine fan via its cooling fan relay... which should already have been running from the a/c computer's relay controlling that fan anyway. mike c.
  3. The ARP head studs & nuts are supposed to be installed with an ARP lube or engine oil lube and then torqued to a different (higher if I remember correctly) value than the factory style head bolt torque spec. Without that lube, screw thread friction and friction between the nut & washer & cyl head post will be excessive while torquing the nuts... i.e. the head bolt+nut assembly will actually be too loose due to the extra friction. Which in turn means the head gasket isn't clamped/compressed enough to do its job properly. Got an air compressor? Rig up a connection to a screw-in style compression gauge fitting. Remove the spark plugs and then rotate the engine so the #1 cylinder is at top dead center. Stick the transmission into 4th or 5th gear (stick-shift), set the parking brake and chock the wheels; for auto transmission cars you'll need to attach a beefy wrench to the crank pulley bolt and hold it - keeping the engine from turning in the next step. Then screw the fitting into the #1 spark plug hole and slowly dial up the compressor pressure. Open the radiator cap and look for signs of bubbles in the coolant. No bubbles? Lower the compressor pressure and disconnect the fitting from the spark plug hole, rotate the engine 90 degrees (transmission in neutral for this step of course) so cyl #3 is at top dead center, and test that cyl. Then test cyl 4, then cyl 2 last with the same techniques, rotating the engine 90 degrees between cylinders. If you find one or more cylinders that let compressed air into the coolant you've narrowed down the leak source as block/head/head gasket related. If no cylinder seems to leak, then the problem is more likely in the intake manifold or turbo. mike c.
  4. Anyplace engine coolant flows is a potential leak - not just the head and block parts. Think about the intake manifold, or a poorly installed intake manifold gasket, or leaks inside the turbo, or leaks between the base of the throttle body and intake manifold. If the gasket surfaces of the intake manifold are not flat then the gaskets won't be able to do their job and you'll have leaks. The base of the throttle body and side of the cyl head also have to be dead flat for the same reason. Folks that don't properly remove ALL traces of old gaskets end up with leaks. Improperly torqued cyl head bolts can lead to leaks around the head gasket. With only 90 psi compression pressure something is amiss on the basic engine parts: block, head, head gasket, pistons, cam timing, etc. If the head bolts were installed into dirty holes, or if some oil collected in the bottom of the holes (oil in the bottom of the holes leads to "hydraulic lock" preventing the bolt from turning in as far as it should - as though the hole were too shallow), then the torque reading is false and the head isn't installed correctly. Thus the head gasket won't be working properly. Are you holding the throttle wide open during the compression test? If not, you may get lower than correct readings. And inexpensive compression gauges are often not accurate. mike c.
  5. Unless the TPS voltage is waaaayyy off it won't affect engine startup. The ECU has a different sensor on the throttle body (opposite end of the throttle shaft from the TPS) to detect if the throttle is at idle. However, the ECU doesn't really look at any of that stuff for startup. Remove the throttle cable and try your "snap the throttle" test again. If the TPS output is more consistent you likely have a bad throttle cable. Typically the cable starts jamming/sticking because the main engine ground strap is missing or not making good contact... thus when you run the starter motor the large amperage flow doesn't have the proper path back to the battery and it looks for the next suitable grounds. Those turns out to be the throttle cable and a skinny wire attached to the intake manifold by the EGR valve. The throttle cable heats up and burns like a fuse slowly buring out... this frying "welds" it to the cable cover and makes it sticky. Another potential issue is the cruise control cable which connects to the pedal underneath the dash with a sliding joint (so it has no effect when the cruise is OFF and you're pushing the pedal with your foot)... verify that isn't binding. Folks that take the throttle body apart often have issues with the throttle return spring and don't get it re-installed correctly. If I remember correctly it is supposed to have three turns worth of pre-load... two is pretty hard to do by hand so that third turn is really tough. There is a small electric motor at the rear of the throttle body assembly (covered by the black plastic/rubber piece sticking out) that drives a screw shaft in/out; the end of this shaft has the "idle switch" (also called "nose switch") and in turn presses on the throttle linkage to adjust the idle throttle opening. That assembly is the "Idle Speed Controller" (ISC) unit. If you stand on the passenger side of the car and look over the top of the engine, just to the firewall side of the throttle cable you'll see a 6-sided slotted screw head. That's what adjusts/calibrates the actual throttle position to your car's ISC. Turn it clockwise 1 or 2 full turns to crank the throttle open and try starting as a test. If it makes a difference you'll need to follow - to the letter - the idle speed ISC/MPS/TPS adjustment procedure in the FAQ corresponding to your model year. You'll also see what looks like a carb's idle speed adjust screw at the bottom/center of the throttle body (to the right of the throttle cable when you're viewing from the passenger fender)... this is the "fixed SAS" screw. It's a backup minimum idle adjustment that SHOULD NOT ACTUALLY BE TOUCHING THE THROTTLE LINKAGE in normal conditions. You should be able to turn that screw 1 complete turn before it "touches" the throttle linkage and starts moving stuff. It's purpose is to serve as a backup in case the ISC assembly fails or goes whacko and tries to totally close the throttle plate. mike c.
  6. A stock StarQuest ECU with a missing or bad/incorrect ECU coolant temp sensor (the 2-prong one in the intake manifold by the oil dipstick) will not run correctly. At cold engine startup the "choke mode" fast idle and extra fuel enrichment could be missing so the engine will be hard to start and will have a lumpy idle. When the engine warms up, without a proper reading from the sensor the ECU thinks "engine is still cold" and you'll get too much fuel leading to bucking (feels almost like detonation) any time you try to accelerate faster than grandma pace. Do not mess with this sensor for a new temp display unless you've replaced the stock ECU with some other system. And if you do have a stock ECU, make sure the sensor works (should read below 300 ohms on a fully warmed up engine) and that the wiring is okay. On a stock StarQuest ECU setup, all sensors are pretty much required and must be the correct ones - close is not good enough. No jury rigging either. Sensors by the thermostat: lowest one with vacuum nipples is for the emissions system, it enables the evap canister cleanout and some other stuff. lowest electrical one drives the dash temp gauge. On 88-later cars it'll be a 2-prong sensor; the 2nd prong is an on/off switch that trips when the coolant temp is warm. This switch enables overdrive on automatic transmission cars and tells the climate control computer "there is heat in the coolant, enable the interior heater" when it senses some heat in the coolant. On 87-earlier cars the temp sensor is 1 wire and a different sensor (under the dash) taps into the coolant flow at the heater core for the climate computer and overdrive functions. Highest one, a smaller electrical sensor, is an on/off switch... when the coolant temps are dangerously high this switches OFF to disable the air conditioning compressor. This sensor is a common failure point; just grounding the wire that normally connects to it will let the A/C function (even if the engine is about to overheat). Replacing this sensor with one for an aftermarket temp display is common. mike c.
  7. Time for some very basic checks. First, a compression check. That'll verify the engine mechanicals are basically "in time" with each other. Second, a real ignition timing check. Remember, the timing mark on the crank pully will line up with the "T" mark TWICE for each flash of the timing gun. You may have the timing 1 complete engine revolution incorrect (i.e. cylinder #1 is getting spark when cyl #4 is actually ready for spark). Third, how do you know it is actually getting fuel? A few things come to mind: The press-fit roll pin between the camshaft, cam drive sprocket, and distributor drive shaft (that stubby shaft extending the camshaft fowards) might be busted or missing... so the camshaft, distributor, and cam sprocket could be way out of whack. Can you slip a wire through the hole in the end of the stubby shaft, through the cam sprocket, and through the camshaft? You should be able to if the pin is installed and those three parts are lined up correctly. How old is the gas in the tank? Old gas degrades into a varnish-like crud that 1) doesn't burn very well and 2) gums up the injectors. Any ECU error codes? Just turn the key ON and look for error codes (procedure in the FAQ). You should have code 1. If you have any others, especially a airflow sensor code, fix those issues. StarQuests will not tolerate any air leaks between the air filter canister (including leaks between the lid and canister or between the lid and airflow sensor) all the way to the throttle body. The system is based on measuring airflow to compute fuel flow... so anything that lets air into the system after the airflow sensor means you'll have too little fuel. Right after trying to start for 20 seconds or so, quickly remove the EGR valve. It's the flying saucer shaped piece bolted to the front/bottom of the intake manifold below the throttle body. Two bolts, a ground wire, a metal clip holding the connector from the distributor, and a thin metal gasket (looks like a shim)... with the EGR off you have openings into the intake manifold. If gas dribbles out those holes your injectors or fuel pressure regulator are dumping gas into the intake manifold --> engine is flooded. Figure out why. Replace EGR valve. Remove the spark plugs. With a big wrench, turn the crankshaft pulley bolt (clockwise only as you face the engine!), transmission in neutral, until the "v" notch in the rear-most pulley lip lines up with the "T" on the timing case cover. Now either the #1 or #4 cylinder is at top dead center. Remove the valve cover. Look at the rocker arms for the #1 cylinder and the #4 cylinder. One of them should have both the intake and exhaust rocker arms riding on the "base circle" of the camshaft - i.e. not on a lobe. If it's the #4 cylinder, rotate the crank pulley bolt 1 full turn back to the "T" mark. Now the #1 cylinder rockers should both be on the base circle and the valves closed. Now remove the distributor cap. Look at where the metal contact of the rotor is aimed, it should line up with the post for the #1 cylinder spark plug wire on the cap. (hold the cap as though you were about to re-install it with the screws lined up) If not, the distributor timing is wrong. Also look at that roll pin (or the hole for it) tying the distributor drive shaft, cam sprocket, and camshaft together: the pin/hole should be at the 12 o'clock position of the cam sprocket. If not, the main crank to camshaft timing is incorrect. Don't worry about the timing chain "plated links" described in the factory service manuals; those are used when assembling the engine. They line up when you put the engine together but once the engine turns they won't line up again until something like 17 engine revolutions - the chain length means the chain does not make 1 complete revolution for each 1 or 2 engine revolutions. The StarQuest ECU wants to see pulses from the ignition system before it commands fuel flow; this is a safety item. These pulses are the same pulses that drive the dash tachometer. Does it wiggle off zero RPM while trying to start? If not, find out why. mike c.
  8. What check engine light? There isn't one on the dash/instrument panel. There is a test port in the top of the glove box where you can connect a voltmeter or a 12volt LED to check/monitor ECU error codes. There are only a few error codes though - not much help. And the symptoms you've described won't trigger ECU error codes anyways. Possibilities: * Bad fuel pump. Some aftermarket ones are pretty crappy. Old/dying fuel pumps tend to work okay when physically cold and then crap out as they get hotter... thermal expansion causes spinning parts to rub and slow down until the pump totally jams. * lousy wiring in the fuel pump circuit including a too-old "ECI" fusible link. Or a poor ground at the fuel pump itself. Both deny the fuel pump proper working electical energy. * There are small filters at the fuel pump inlet and at the end of the fuel pick-up tube. These could be clogged - often happens on cars that sit for a while and get rust on the fuel tank inner walls from the trapped humidity. * gas cap vent function busted so a big vacuum builds up in the gas tank as the fuel pump sucks out fuel... eventually the pump can't overcome the vacuum and the fuel flow rate drops. If driving the car with the gas cap removed works better, get a new gas cap. There is no "gas cap" ECU error code either, nor a "evaporative emissions system failure" error code on StarQuests. * Any chance the car is actually running over-rich instead of loosing fuel? Issues with the fuel pressure regulator or a blocked fuel return hose lead to excessive fuel going through the injectors. At high power the over-rich can lead to a bucking sensation and a general lack of power. The coolant temp sensor used by the ECU (different sensor from the one that drives the dash temp gauge) is a common failure point; when it goes bad the ECU thinks the engine is stone cold and runs it over-rich. Usually there is a bucking/hesitation when trying to accelerate hard but the engine will run okay when babied if this sensor fails. It's the 2-prong sensor in the intake manifold, next to the oil dipstick, with a squarish connector. The sensor is mounted vertically with the prongs pointing upwards. * weak spark can cause these symptoms, especially if it is from a dying ignition coil. Like the fuel pump, dying ignition coils often work okay when cold and get progressively weaker as they heat up. Incorrect spark plugs or too-large plug gaps can also cause problems. StarQuests like old-style basic plugs... not the expensive platinum plugs or other long-life ideas. Many of those fancy and expensive plugs can't take the abuse of a turbo engine anyway. * wrong injectors... lots of different injector flow rates were used on Mitsu cars over the years. The 87s use a low-flow-rate "primary" injector with a black plastic band, and a higher flowing "secondary" injector with a blue or green band. Unfortunately a lot of other Mitsu injectors have the black color... and Mitsu Japan was shipping incorrect flow rate injectors as spares for many years - a situation they never corrected. Those medium-flow black banded injectors won't work in 87-later StarQuests. If you look carefully at the injectors you should see faint letter+number codes stamped into them just below the plastic ring. The black banded injector should have an "L" letter, the blue/green one should be an "M" label. Ignore the numbers, they don't matter... just the letters. Also, while the engine is idling, unplug the connector on the blue/green banded injector. The engine should continue to idle just fine; if it dies the injector connectors are backwards. Only the black banded primary is used at engine idle. * Incorrect method used to set ignition timing. StarQuests have a simple electronic ignition system; timing is not controlled by the main ECU. The ignition module does get 2 signals from the ECU that slightly modify how it works... but the ignition system can operate even with the ECU unplugged - it's that independent of the ECU. One of those 2 signals is a signal sent by the ECU when the engine is stone-cold or the ECU thinks the vehicle is operating at a high altitude; when present this signal makes the ignition module advance the timing 5 degrees beyond what the centrifugal weights and vacuum advancer mechanism call for. If somebody sets/checks ignition timing on a cold engine, or when the engine is in fact at high altitude, the timing probably got set to the 10 deg BTDC service manual specification when it should have been set to around 15 deg BTDC thanks to the extra 5 degrees of advance. There is a 2-wire connector hanging from the wire bundle between the air filter and fender... unplug the protective cap on this puppy and jumper the wires together. This forces the ignition system to "warmed-up engine at low altitudes" removing the extra 5 deg timing advance - now you can properly set the ignition timing to 10 deg BTDC. That's the correct procedure - using that jumper first. StarQuests with incorrect ignition timing - especially retarded a few degrees - have significant power loss. A car with the timing tuned/set while the engine was cold will end up with only 5 deg BTDC once a little heat is in the engine and will be a total dog. * A partially plugged exhaust from old/damaged catalytic converters will slowly choke off the engine; it'll happen faster at higher power settings. Take the car for a drive without "mashing the gas" - just around town at 40 to 50 MPH until you feel the power loss is significant. Pull over (engine still running) and quickly pop the hood. Look at the exhaust end of the turbo. If it's glowing at all then you can bet there is an exhaust blockage. Usually the "pre-cat" (first catalytic converter --> the one bolted to the turbo) plugs first. If the vehicle ever had a "blown turbo" or other major issue then oil got into the pre-cat and that's a pretty fast way to kill it. Has the mechanic connected a fuel pressure gauge to the engine? That will tell you if the issue is fuel delivery (wiring, fuel pump, blocked passages, gas cap, etc.) or someting in the engine itself (ignition timing, plugged exhaust, etc.) There is an allen-head plug at the very top/rear of the throttle body where a fitting can be inserted (instead of using the "special service tool" funky block shown in Mitsu's service manuals). Use a couple washers between the driver side hood hinge and hood to lift that corner of the hood for the gauge's hose and you can tape the gauge to the wipers will you go for a test drive. You should see 36 to 38 psi at idle or while driving; the pressure should increase in sync with turbo boost pressure (i.e. 41 to 43 psi when the turbo is creating 5 psi boost). If the pressure drops off while you drive then it's a fuel delivery issue. mike c.
  9. Tractor tire = bull ramp = china shop Somebody way underbuilt that ramp! First time I watched the vid and saw that tire bash through the ramp totally cracked me up. Days later I'm still chuckling thinking about it as I type this! I also liked the "catch" setup at the end of the run... and how poorly it did. Some tires missed it... they set up additional things to (try to) slow down that tractor tire. mike c.
  10. new stock injectors... very difficult to find. If you get the part numbers from the FAQ or the scanned PDF service manual floating around the internet you might find a dealer willing to order them. Be warned though that Mitsu Japan had been sending the wrong injector for the primary... basically sending H/I/J/K (medium flow) injectors instead of the lower flowing L injector. Those injectors in 87-later cars don't work properly - just too rich. So a good used set is about the only option I'm aware of if you want to stick with Mitsu injectors. Trilogy and a few others sell modified injectrors from a different manufacturer. These can work in StarQuests - some folks have them and love them, others didn't have much luck. These injectors are modified by having their upper portions machined to a smaller diameter to fit the stock StarQuest "injector cover" (i.e. the top of the throttle body assembly) and to take Mitsu-style o-rings... or a modified injector cover (the factory openings for the injectors gets drilled to accept larger diameter injector inlets) is sometimes provided. These injectors are a tad taller than Mitsu injectors too - so longer bolts for the injector cover and some spacer pieces are included. These kits are the only simple "buy it, install it, and just drive" alternative. The thing is... almost all fuel injectors available aftermarket are designed for port fuel injection applications. This means they expect to be mounted in the intake manifold right next to where it bolts to the cylinder head. The injector spray pattern is a narrow cone such that the cone diameter is similar to the diameter of the intake valve so majority of the fuel gets sprayed right into the cylinders - and not along the walls of the intake manifold or cyl head passages. The spray velocity is pretty high too - like the "jet" setting on your garden hose gun. On throttle body systems like StarQuests the injector flow pattern is totally different since spraying a cone downwards from the throttle body mounted cone-style injector would just paint the intake manifold with gas. Little would mix with the airflow. Instead, throttle body injectors spray a nearly flat pattern - almost like spraying sideways from the injectors. The resulting flow pattern looks like an open umbrella. The spray velocity is lower too (since the same amount of gas is spread out over a wide plume instead of being concentrated into a narrow cone) so it's much easier for the airflow to mix with the gas and carry fuel to the cylinders. Since the aftermarket injectors "adapted" to the StarQuest throttle body generally have a cone spray pattern the air + gas mixing is poor at low airflow rates (e.g. engine idling) leading to a lumpy idle. At high power the airflow is strong enough to overpower the cone pattern. There are three main types of failures on StarQuest injectors: 1: the electrical coil goes bad. The ECU can't "fire" the injector and the injector is basically junk. 2: the "pintle valve" inside the injector gets gummed up with varnish (caused by old/dirty gas) and sticks. When this happens the injector may be stuck closed --> no fuel flow or it may be stuck such that fuel always flows regardless of the ECU commands --> floods the engine and causes black smoke if you re-start the engine a few minutes after shutting it off (7-11 or Starbucks type of trips) Cleaning the injector with spray carb cleaner and a 9-volt battery (to power the coil, moving the pintle) will often cure this failure type. 3: the o-rings inside the injector go bad - like any rubber part they eventually get very hard and fail to seal against the fuel pressure. When this happens you'll have gas OUTSIDE the injector, typically oozing/spraying between the plastic and metal parts. The plastic part of Mitsu injectors is NOT part of the fuel seal... fuel is supposed to remain in the center channel of the injector. When the o-rings fail gas squeezes outside the center channel and fills up the injector body - i.e. where the electric coil is and eventually fuel pressure shoves the gas past the plastic parts and you have an external fuel leak. Nothing you can do to properly fix such injectors either. In theory you could open the injector and replace the o-rings but enough stuff gets damaged trying to open the injector that no company is willing to do this. Not profitable. You have to remove the plastic (by busting it apart) without busting the electrical wires to the coil, then "unbend" the curled metal lip of the round central can-shaped part of the injector (where the coil lives) to free the inlet pipe and the flat metal disk grabbed by the curl. You can't see this disk nor the curl normally because they're covered by the plastic. Then you can lift off the inlet pipe + disk, remove the coil, and see a couple o-rings. Replace them (after chipping out the old ones - they'll act like they are glued to everything), re-curl the metal, and mold new plastic. Easy-peasy? Not hardly. If your car has un-fixable injectors, the options basically boil down to: 1: take your chances on finding a good set of used injectors for sale. Who knows how long they'll last though. 2: try the aftermarket ones - with the less than ideal spray pattern. They may be good enough for you - as I noted above many folks have daily driver StarQuests running okay with the aftermarket injectors and others have had bad luck with them. 3: jettison the stock throttle body setup and replace it with some kind of port fuel injection: new manifold, new injectors, new fuel computer, etc. Not for the faint of heart and may not be legal in some smog-testing areas. There are no complete "just buy this kit and install it" setups out there to my knowledge. 4: Replace failed stock injectors with other Mitsu style throttle body injectors that may have different flow rates (much more choice, much more chance of finding new old-stock injectors) and adjusting the car to work with them: e.g. new fuel computer like the Megasquirt that you can program to work with whatever injectors you use, or some scab-on gizmo that alters the stock airflow sensor to stock ECU data stream to compensate for the different injectors. Not impossible but no simple & direct "buy this, install it, and just drive" solution... everything needs some tuning. Obviously #1 and #2 are the easiest options. Injector failures, and the cost of new injectors, airflow sensors, and a few other parts, lead to many StarQuests getting traded in on "easier" cars long ago... or getting junked... or getting engine swaps. Combined with dealer mechanics that often had poor skills with these cars... only dedicated fans of the car keep them! mike c.
  11. Good morning to you too! Or good evening now I guess. mike c.
  12. Nobody knows for sure (outside of Mitsu engineers anyway) on exactly what was tuned with the bypass. Was it to adjust air:fuel ratios at idle where the ECU is in open-loop mode? That's my guess, and what many others thought, since idle smog tests are common. This would allow fine-tuning of idle emissions to compensate for car-to-car variations in injector flow rates, airflow sensor readings, fuel pressures, etc. Once the ECU goes into closed loop (engine warmed up and running at a fairly steady part-throttle RPM) the ECU closed-loop logic will basically get the emissions as good as possible no matter what the bypass is set to... rendering it redundant. At full-throttle accel (ECU in open loop again) the bypass might have some effect but no state/county/city smog test I know about tests under this type of load - only the initial (and massive) vehicle smog certification testing done by the manufacturer before the new model goes on sale looks at wide-open-throttle emissions. I never really examined the amount of bypass air on my 88... and my car has the HKS turbo compressor bypass kit which replaces the accordion hose. Ergo I don't have the 1" port on the accordion hose so my bypass system is capped on the air filter lid. And I have no problems passing California emissions either on the "road dyno" method currently in use or the old style "at idle" sniffer test. mike c.
  13. US Spec intercooled StarQuest 87-88-89 fuel systems are virtually identical and almost all parts will interchange. The only significant differences are the ECU itself (87 is different from 88-later but 88-later will work in 87 cars just fine if you swap the cases; 87 ECUs though are a big downgrade for 88-later cars) and the turbo wastegate actuator. The ribbed "accordion hose" and the lid to the air filter canister are different on 88-later cars; the lid has an extra 1" port for a fine-tuning bypass hose that connects to the 88-later accordion hose. You'll have to make a cap for this port to install the air filter canister+lid on your 87 unless you get the 88's accordion hose too. No big deal to cap this port. The 88-later airflow sensor itself will work fine on your 87. mike c.
  14. US spec 87-88-89 "wideobdy" cars with intercoolers used the same injectors. The FAQ has a listing of the various injectors used on various Mitsu vehicles. The list includes the "letter code" which is the best way to identify what injector - correct or not - is actually installed on a car. If you look closely at the injectors you'll see a faint stamping just under the plastic band - a number+letter combination. The letter is the key, the numbers don't have any known meaning. On US-spec StarQuests, you'll find basically find 4 different injectors based on production date & intercooled or not: * the 1983 models had very unique injectors - they look nothing like the injectors for other model years, have their own wires (instead of using a connector), etc. And they fit only the 1983 style throttle body I believe. * 84 to 86 model year cars used medium flow rate injectors, 850cc or something in that ballpark. They have a black plastic band. Letter code is H or J. * 87-later intercooled cars (all US spec models I believe) came with the lower flowing "primary" injector (580cc ballpark) letter code L with a black plastic band, and the higher flowing "secondary" injector (1080cc ballpark) letter code M and a blue or green plastic band. The other letters fit Galants and other Mitsu engined cars and won't work in a stock StarQuest. They'll physically fit into the throttle body... but have totally incorrect flow rates and the StarQuest ECU doesn't learn/adapt much unlike more modern ECUs so the engine won't run very well with incorrect injector flow rates. Note that the 87-later Mitsu factory service manuals use the terms "small" and "large" for the injectors, not primary and secondary. Small/large refer to the flow rates... not the physical injector size as all 84-later injectors are physically the same size. Only the size of the spray tip opening varies - that's what determines the flow rate. The parts list uses "front" and "rear" for the injectors. Front=primary=small mike c.
  15. You could have easily pinched a wire when re-installing the battery tray... and blown the "battery circuit" fusible link. It's one of the red ones in the 4-link box next to the battery. (Link #3, in between the brown and green links) That's what powers low-amperage stuff in the car that does not depend on the ignition switch. The headlights have their own fusible link... but the pop-up motors are on the "battery circuit" link. Check that link, follow the wires going in/out of the fusible link box, and that fat square 3-wire connector several inches below the box. That connector is a known trouble spot on StarQuests - it gets corrosion buildup by being close to the battery... bump it/shake it/disturb it and it's likely to quit. mike c.
×
×
  • Create New...