I fucking hate black box testing.
It's just such a fuck around, taking a wild guess at what could have gone wrong by inferring from some ambiguous readings or results. I do this shit at work all the time. You might get some weird trend in your results, so you go through it step by step for each of the million different things that could have caused it when in the end it could just have been pure chance. 90% of the time you never find what went wrong because you can never repeat it. Biological systems are like that, at the end of day you decide on some theory as to what it was and blather on in the CAPA how you came up with that idea trying to sound like it wasn't pulled out of the air.
Cars aren't much better. They shouldn't be complicated, you've got bits that move when shit gets burnt when some bit fires a spark. Except each of those bits relies on 30 other bits to do the right thing and every now and then something doesn't. Then the fun starts.
Mind you it helps when you do a bit of research or read a service manual, but the worst thing about owning a Skyline is that you own a Skyline. I'd sell the R32 for a Nissan Crew cause at least then I wouldn't have fat Systems Administrators waving at me because he bought an R33 and is now part of the special community. Its also annoying when to look for a solution to a problem you have to trawl the forums. Millions of threads for people asking how to get more power from their automatic R33 all featuring the keyword you are searching for. Cunts.
To be honest, I like to grab a spanner and see whats going on myself, except its taken me forever to fault find in this skyline because I've had other (more?) broken cars to take care of. In the 6 or so months since I last drove it regularly I've ghetto-ed up some more solid mounts for the intercooler (because I initially thought that it might have been an air leak causing it to lean out), rewired half the electrics under the dash because the immobiliser failed to disengage, pulled off the fuel rail in a failed attempt to remove the fuel pressure reg (which wasn't faulty anyway), hooked up an oil pressure gauge to test fuel pressures to test the reg and cleaned the air flow meter more times than I can remember and heaps of other shit.
As it turned out all those sensors work, even the TPS which seemed to be a perfect explanation for the problems (and easiest to fix) was fine. AFM voltages were fine too. Everything is linear and everything is about right for base voltages. So nothing should be causing such a complete shut down of the motor. Except I never considered the CAS, clearly I've been around cars built in the 80s too much lately to even consider it. Anyway, that's my current theory, ie the CAS is shagged out. It explains why it shuts down, mostly. Why it failed at 3000 or so RPM I have no idea.
The good thing about having the PowerFC is that all the sensor inputs it needs and ECU outputs it produces are displayed on the hand controller. The CAS though isn't something you can really easily display, it outputs 2 wave forms; one for each cylinder's TDC and one which is segments between the TDCs. It basically allows the ECU to know where the motor is at (precisely) and how quickly its spinning. You can't really view that in real time, at least the raw form of it. There are of course ways to intercept what the CAS is saying the to ECU, but I don't really care what its saying, I just want to know if its fucked. The bad thing about the PowerFC is that it has no failure modes. If a sensor fails completely, goes out of range or whatever it has no idea what to do. A standard ECU has limp-home modes for half the things that can fail and error codes for everything. A CAS thats failing generates a code 11 or whatever and you just nip down to Nissan and pick up a new one for $1500. The PowerFC just keeps running like nothing is wrong even if its getting signals from the CAS that the motor has gone from 3000rpm to 0rpm or decided to skip a few cylinders to save time. The motor makes a fucking horrible noise and flames drop out the back end. You have to crawl it back up to your driveway as a few of the neighbours disapproving faces appear from behind curtains.
So yeah, next step is to find a new (or rather second-hand) CAS and hope that fixes things. If not, I'm running out of things that could be broken.
Also, speaking of disapproving neighbours, I'm apparently not the pariah I thought I was tearing up the street in cars with abbreviated exhausts leaving behind engine oil, cakes of mud or whatever else happens to fall off the car. I got invited to one of the neighbours' birthday lunch. I mean I've lived there for 2 years now, but surely being woken at 2am by someone parking their car would shit you off somewhat. I've also never spoken to them in that whole time except for the minimum that etiquette requires when another neighbour Thunderbird 2 launches his car backwards out a driveway and into a tree.
I think I need to be busy that day...
It's just such a fuck around, taking a wild guess at what could have gone wrong by inferring from some ambiguous readings or results. I do this shit at work all the time. You might get some weird trend in your results, so you go through it step by step for each of the million different things that could have caused it when in the end it could just have been pure chance. 90% of the time you never find what went wrong because you can never repeat it. Biological systems are like that, at the end of day you decide on some theory as to what it was and blather on in the CAPA how you came up with that idea trying to sound like it wasn't pulled out of the air.
Cars aren't much better. They shouldn't be complicated, you've got bits that move when shit gets burnt when some bit fires a spark. Except each of those bits relies on 30 other bits to do the right thing and every now and then something doesn't. Then the fun starts.
Mind you it helps when you do a bit of research or read a service manual, but the worst thing about owning a Skyline is that you own a Skyline. I'd sell the R32 for a Nissan Crew cause at least then I wouldn't have fat Systems Administrators waving at me because he bought an R33 and is now part of the special community. Its also annoying when to look for a solution to a problem you have to trawl the forums. Millions of threads for people asking how to get more power from their automatic R33 all featuring the keyword you are searching for. Cunts.
To be honest, I like to grab a spanner and see whats going on myself, except its taken me forever to fault find in this skyline because I've had other (more?) broken cars to take care of. In the 6 or so months since I last drove it regularly I've ghetto-ed up some more solid mounts for the intercooler (because I initially thought that it might have been an air leak causing it to lean out), rewired half the electrics under the dash because the immobiliser failed to disengage, pulled off the fuel rail in a failed attempt to remove the fuel pressure reg (which wasn't faulty anyway), hooked up an oil pressure gauge to test fuel pressures to test the reg and cleaned the air flow meter more times than I can remember and heaps of other shit.
As it turned out all those sensors work, even the TPS which seemed to be a perfect explanation for the problems (and easiest to fix) was fine. AFM voltages were fine too. Everything is linear and everything is about right for base voltages. So nothing should be causing such a complete shut down of the motor. Except I never considered the CAS, clearly I've been around cars built in the 80s too much lately to even consider it. Anyway, that's my current theory, ie the CAS is shagged out. It explains why it shuts down, mostly. Why it failed at 3000 or so RPM I have no idea.
The good thing about having the PowerFC is that all the sensor inputs it needs and ECU outputs it produces are displayed on the hand controller. The CAS though isn't something you can really easily display, it outputs 2 wave forms; one for each cylinder's TDC and one which is segments between the TDCs. It basically allows the ECU to know where the motor is at (precisely) and how quickly its spinning. You can't really view that in real time, at least the raw form of it. There are of course ways to intercept what the CAS is saying the to ECU, but I don't really care what its saying, I just want to know if its fucked. The bad thing about the PowerFC is that it has no failure modes. If a sensor fails completely, goes out of range or whatever it has no idea what to do. A standard ECU has limp-home modes for half the things that can fail and error codes for everything. A CAS thats failing generates a code 11 or whatever and you just nip down to Nissan and pick up a new one for $1500. The PowerFC just keeps running like nothing is wrong even if its getting signals from the CAS that the motor has gone from 3000rpm to 0rpm or decided to skip a few cylinders to save time. The motor makes a fucking horrible noise and flames drop out the back end. You have to crawl it back up to your driveway as a few of the neighbours disapproving faces appear from behind curtains.
So yeah, next step is to find a new (or rather second-hand) CAS and hope that fixes things. If not, I'm running out of things that could be broken.
Also, speaking of disapproving neighbours, I'm apparently not the pariah I thought I was tearing up the street in cars with abbreviated exhausts leaving behind engine oil, cakes of mud or whatever else happens to fall off the car. I got invited to one of the neighbours' birthday lunch. I mean I've lived there for 2 years now, but surely being woken at 2am by someone parking their car would shit you off somewhat. I've also never spoken to them in that whole time except for the minimum that etiquette requires when another neighbour Thunderbird 2 launches his car backwards out a driveway and into a tree.
I think I need to be busy that day...
2008-07-04 03:29:33 ( 0 Comments )





















