Jump to content
Please check your junk folder for registration emails ×

Durty-Sanchez

General Member
  • Posts

    255
  • Joined

  • Last visited

Posts posted by Durty-Sanchez

  1. As above + a couple of others:

    1. Check battery terminals. Corrosion build up between the electrodes and the connector can give you crap current flow. Take them off and give them a good wash. Re-connect, tighten, then crc to prevent oxidising. This is what causes the majority of my clicking-relay, alarm-wont-work, everything-dies-on-the-rimutakas-infront-of-a-police-car problems.

    2. Remove starter motor earthing plate (connects starter motor electrically to the block). Check for pitting due to high current loads. Pitting causes the contact surface area between the block and the plate to reduce. The result is a higher resistance between the starter and the battery on the ground side, which will require another 20+ amps to get the sucker moving.

    Grind the plate flat and reassemble (30mins all up to do). This will help with starting to no end and may also explain the lack of power to crank the engine even when being jump started.

  2. 24hrs with the battery off should have been enough. The old green & black jumpers and drive trick does clear fault codes and a few other memory locations but to wipe *everything* you can't beat the battery off trick.

    Swapping coils one by one could be a PITA - find some decent ones and swap them all at once, at least that way you can spot any improvements over the old set. I wouldn't rule out the pump just yet... if it's getting on a bit or has a less than ideal power supply then it may struggle under load. At idle/town driving the pump could be supplying enough fuel to do the job. Stick your foot down and all of a sudden the pump starts to struggle to keep up with demand. The resulting lean condition could be your missfire... What did your old plugs look like? When my pump was playing up I noticed they were very white and powdery when I did them + coils, that's what gave me the hint that it was running lean and I had just wasted $500 :(

     judd said:

    it misfires, either not getting spark or not getting fuel into one of the cylinders im assuming because it seems to idle ok run ok at low revs but under load etc just misfires, so possibly rules out the pump because its getting some fuel atleast. im going to try get some second hand ones from wrecker mate swap them out one by one and see how it goes. the car had a flat battery before so would that have been enough to clear ecu memory?? was disconnected for 24 hours while i charged the battery, reset ecu you plug the black jumpers and green jumpers under the steering column and drive constantly above 11km/h or something to that effect??

  3.  judd said:

    ok so knocked a heap of crap out of the exhaust, put it back in, changed the oil took it for a run.. still no change.

    i used to have a turbo integra race car that used to do something similar under acceleration/boost when one of the injector plugs was loose. leads me to think now the problem is either spark or fuel related. spark plugs have just been changed albeit cheap copper plugs not the flash iridium plugs. put some injector cleaner in with the fuel incase of fuel block somewhere. so next step possibly the coil packs?? at what stage of the ej20tt's life do these things start to give you problems?? are they expensive to buy new?? do you have to buy genuine or is there an aftermarket option at repco?? anything else ignition wise that could go wrong that wouldnt throw a cel code? tia

    Coils aren't cheap (110+gst each at the cheapest). You'll need genuine coils unless you convert to aftermarket and mount them elsewhere and run leads to the plugs. Your cars of the age where it probably needs new ones, but it'd suck if you forked out $500 odd and it didn't solve your problem.

    - Does it missfire when you put your foot down? Or does it just feel like it's got no go?

    - Reset your ECU by disconnecting it for 24hrs. Your model still stores timing values in volatile memory so that'll be enough to wipe it clean. I know it's time to do mine when I can see i'm at full boost but have the acceleration of my mothers pulsar.

    - Fuel pumps are another common item to go in these cars around this age. I replaced my coils for the 2nd time before realising it was the pump + crappy wiring that was causing my hesitation/misfires. The wiring to the pump can also be suspect if you have a lot of back seat passengers.

    • Check error codes (if any)
    • Replace oil + coolant + thermostat
    • Has the cambelt/waterpump been done? It's of that age and may not have been seen to yet. Use a genuine subaru thermostat when doing so.
    • Is it a coolant smell or an oil smell?
    • Coil packs.. always fun, but might explain rough idle
    • Check fuel pump wiring - it's amazing what a poorly operating fuel pump can do

  4. Subaru Select Monitor (SSM) is the native Subaru protocol and it owns OBDII.. but it's application specific. There's tonnes of info out there on SSM and SSM2 if you want to build your own, or even better lots of people have already done it and their source code/applications/diagrams are floating around the intertubes. Downside is that unless your ECU/Model has already been documented you'll be doing a bit of exploring yourself...

    Edit - I believe that the likes of OpenECU support SSM(2) as well as OBD(II) protocols so a tactrix cable may be a good investment. Read up a bit on it fist though to be sure

  5. pffft genuine... Actually that's not a bad idea now that supercheap ditched their fram filters (I liked the sandpaper like grip on them) for ryco. Also means I don't need to search every repco/supercheap in the region for 20mm crush washers ;D

    Cheers guys

  6. Fuel pump relay doesn't initiate sometimes ... its fairly old so I figure it's just getting stuck (comes right after a few minutes). New one will do the trick - unless it's the alarm relays!

    In cabin fan relay is also gone (I assume). The main power wire to the fan went some years back somewhere between the relay and the fan (so under the dash somewhere). I ran a new wire and it was sweet for a year but eventually it stopped again. Power gets to the relay but not out of it so it's my main suspect... could be running too much current or shorting out somewhere, potentially I need to add some resistance to the fan if it's drawing too much current or find where the power wire's shorting out.

    Stoked about not needing the de-gas!

  7. So.. will be attempting a heater core swap in a couple of weeks (and attending to blown relays/wiring that plauges the legacy). Aside from de-gassing the A/C, are there any other things I should be preparing for? New seals? Car is a 94 TT Legacy...

    Thanks!

  8.  familyman said:

    Fantasticly written there Sanchez. Thought of freelancing for NZPC mag? They need some good writers I hear.

    Thanks, although I'm sure I've seen the editor posting tranny porn of himself on /b/ :-\

    If Pdizzle reads this, soz bro - happy to accept a job offer though!

  9. Background - Your turbo has a outlet from the compressor side that supplies pressurised air that is fed in to the wastegate diaphram. In the diaphram is a spring, which will begin to move once a certain pressure is reached, and eventually is fully pushed in when full boost is reached.

    Every system mentioned simply tricks the above system in to doing the same s*** but at different times.

    Standard - The manufacturer put's a t-piece in between the turbo and wastegate and plumbs that to a normally closed solenoid. So essentially it's the same setup as mentioned above but with extra piping. Now the ECU is pre-programmed to open this solenoid rapidly under certain situations. The effect of this solenoid opening is that it creates a hole in the piping, a hole from which pressurised air escapes, thus delaying the wastegate diaphram spring from being fully pushed in (the other side of the solenoid is the atmosphere). The idea behind opening the hole rapidly is that you can assume that if the hole were open for half as much as it were closed the hole looks to only be 1/2 the size of the full size hole. Now if the hole were open 3/4 of time and closed 1/4 the time the hole now appears to be bigger (3/4 the size of the full hole). The larger the hole the more pressure is "bled" from the system, so the more pressure is removed and unable to open the wastegate. That's a rough guide of pneumatic control.

    Electronic - This replicates the above except unlike the factory ECU you get to control it and do the settings. Other than that it's almost exactly the same as the standard method of control.

    Boost-T - This is the analouge version of the above two. The T-piece is adjustable so you can select the size of the hole you want, which then delays the time it takes to create the amount of pressure to open the diaphram. Its simple and effective, but lacks the ability to adjust the hole like the above methods do because to adjust the unit you need to stop, pop the hood, and tinker with the device manually. You can use this in conjunction with the standard boost control to simply "up" the boost while retaining some form of electronic boost control.

  10.  boostcut said:

    what are the factors that lead you to believe its not plugs? have you even looked at them? also, just looking at the plug doesnt always mean its fine. it sounds to me very much like an ignition problem, as a high gear and high boost means load on the ignition system. maybe you could spend 5 minutes checking things and save some money? but what do i know.... i didnt do 6 years of study to be an ATech mechanic for nothing

    If spark plugs blow out/fuel pump dies/AFM dies I can still see what RPM my engine is doing... how else can I coast down a hill and still see the tacho working?? I didn't spend 10 seconds working that one for nothing ;) Tacho dying at the same time suggests that it's cam/crank angle sensor, ECU, or dud connection somewhere. Cam/crank sensor signal is piped through the ECU, which then outputs the tacho display signal.

    There's an outside chance that *something* is shorting out the ECU - possibly in the solenoid box or on the ECU itself. Check for ECU power at all of its power sources. One good way might be to plug in the diagnostic connectors and go for a drive. If the problem is repeatable (and you have no error codes) then it will flash a steady pulse at you. If you lose that pulse for the duration of your power loss then you know the ECU has ceased executing its firmware.

    I believe the ECU requires both cam and crank signals to work in order to synch them on startup. After that it only cares about one of them, but there's still a chance that one or both of them doesn't like heavy acceleration so dies. Same goes for a bunch of other connections, but it's just a matter of ruling all the causes out one by one until you've got it. You're on the right track it seems so keep poking around, you'll find it eventually! Do you have a spare ECU you can test? If not PM me and I can send you a spare one to try out.

  11. You'd need to pick a FET and get it's on/off gate voltage, then pick a RC pair (or variable resistor region) that best suits... etc. I'd just get the jaycar one, it comes in a cool box and everything!

     pappu']

    [quote name='Durty-Sanchez said:

    A simple RC circuit would do the trick here...

    1. Push button to charge cap and start pump.

    2. Release button to let cap discharge through (variable) resistor.

    3. Cap discharges over time period t, at which point the voltage has dropped so much that the pump goes off

    If you use a variable resistor on the discharge circuit you can alter how long it takes to discharge the cap (which is how long the pump stays on for). Just hook the cap voltage up to a FET load driver of some kind and it'd work a treat

    /quote]

    wud u know what sort of circuit i wud need to say do a 3 or 4 secont time lapse?

    what about a 10 sec one.

    how do i put the FET in there lol.. not too clued up on that one.. but m sure need it to drive the pump relay right?

  12. A simple RC circuit would do the trick here...

    1. Push button to charge cap and start pump.

    2. Release button to let cap discharge through (variable) resistor.

    3. Cap discharges over time period t, at which point the voltage has dropped so much that the pump goes off

    If you use a variable resistor on the discharge circuit you can alter how long it takes to discharge the cap (which is how long the pump stays on for). Just hook the cap voltage up to a FET load driver of some kind and it'd work a treat

  13. Pass, its been many years since my time as an alarm installer... but I seem to recall that subies are wired to negatively switch, where as most alrams are geared to positively switch. Don't quote me on it, just something that hangs in the back of my mind that *if* you ran in to you could say "oh hey... that guy said that would happen"

  14. Do It Myself? Depending on what your alarm does you can reverse wire the door lock actuators so they work (from memory subies are wired to pulse negative rather than positive, but most alarms are set up to pulse positive to the actuators)... dont quote me on that though... g'luck!

  15. 1. If by wireless entry you mean remote-unlocking then no, so long as your car already has central locking. Subie's do some odd things though with voltage transitions to get stuff to work... been a while so can't recall exactly. If you mean to completely open the door then yes you will need something to push it open.

    2. I don't know of any, but it's not exactly rocket science. Nothing a test light or multimeter cant tell you.

    3. At $150 to get it installed vs how ever many hours it will take you it's not such a bad price to pay. Having said that though you'll learn a bit about how it all works. Even the best alarm can be crap and offer little protection if it's poorly installed. Try to immobilise the critical things (computer, fuel pump, AFM if you can) and make it as confusing as hell.

  16. [metro man forums] Agreed, infact i'm fairly sure the whole reason I bought my legacy was because it had a sunroof. The ladies love it too...

    Sedan roof swaps are possible, mitsi lancers get it done all the time because they rust something awful. AFAIK the wiring's all self contained aside from power (not sure where it's fused?). Leaks are managed by the built in guttering, which if you get a roof cut is part of the structure. It's probably not all that impossible or beyond budget to do so long as you can find a complete roof intact

  17. Try the fuse box under the bonnet. Also, check if your dash lights come on when you turn the lights on, it might help in finding if the fault is in the cabin or under the hood

  • Who's Online   0 Members, 0 Anonymous, 135 Guests (See full list)

    • There are no registered users currently online
×
×
  • Create New...