1985-1998 [Subscribe to Daily Digest] |
 | [Main 9000 Bulletin Board | BBFAQ |
Prev by Date | Next by Date | Post Followup ]
Member Login / Signup - Members see fewer ads. - Latest Member Gallery Photos
Some CPS info - long  1 Saabers Like This Post!
Posted by Ari [Email]  (#2847) [Profile/Gallery]
(more from Ari) on Sat, 4 Dec 2010 14:22:13
Members do not see ads below this line. - Help Keep This Site Online - Signup
|
I just had a no-start issue with my '92 9000 N/A, so I want to pass on some information.
Note that this ONLY applies to the '91 to '93 DI-equipped cars. In 1994, the Crank Position Sensor (CPS) went to an entirely different technology (more on that, below.)
Issue - the car ran fine, but after I shut it down, it refused to start. Cranked fine, no tach bounce, didn't hear the fuel pump. Did some quick checks, and narrowed it down to the Crank Position Sensor (or so I thought). The clues were no tach bounce, fuel pump not running while cranking, and when I read out the LH flash codes, it was Engine Speed Signal Missing.
With these cars, you should see the Tach move a little when cranking. Another test is to take a voltmeter and pop open the test connector - under the hood, LH side, right by the windshield. Look for the round connector. You want pin 5. Pin 6 is blank, and it counts counterclockwise, so 5 is the first one clockwise from the blank. With a meter set to DC Volts, expect to see 6-8 volts when cranking. Bad is 0 or 12. You're really looking for pulses, which are hard to see with a digital volt meter - that's why you should see the Tach needle move just a little when cranking. It it doesn't move at all - problem.
This Does Not Mean the Sensor is Bad. Probably, but read on. The sensor is a hall-effect device right behind the crank pulley. A shutter wheel on the crank pulley passes through a slot in the device, which produces a pulse. It's a solid state device, with three leads - power, ground, and signal (out). The CPS goes to the DI Controller box (plastic), which is in the aquarium, right by the LH controller (metal). The DI Controller plays with the signal some, sending it to the DI cassette and the Tach. So the signal you see on the test connector and the Tach are from the DI controller, not directly from the sensor.
How do you know if the CPS is OK? Ohm checks won't work. The best you can do is tap into the signal as it goes from the DI Controller to the sensor (remember - the DI Controller is the plastic box in the aquarium - not the DI Cassette on top of the engine!) If you really, really, really know what you're doing, disconnect the connector from the DI Controller, ground pin 15 on the harness connector, apply 5 volts DC or so to pin 6, and monitor pin 23 with a meter or an oscilloscope. As you crank the engine, you should see the signal on pin 23 pulse. If so, the CPS is OK. If no change, the CPS is bad. The other way is to let the DI controller supply power - leave the connector on the box, and using a pin, poke through the wire to pin 23 (green/red - note, there are TWO green/red wires, so ohm out pin 23 to the wires to figure out which is right). Then put a meter or scope on the pin - just don't let it touch ground. Again, this is not for the faint of heart.
I expected a bad CPS, and I was wrong. It worked fine. That meant my controller was bad (!). As you may guess, I have things like an oscilloscope and a passable electronics lab. If all you've got is a $20 Radio Shack meter, and soldering is something you do to copper plumbing, look for help. So I opened up the controller (very easy - four screws.) The following won't make any sense unless you've got an electronics background. It turns out the problem was a corroded component - the timing crystal for the microprocessor. The crystal wasn't oscillating, and the processor wasn't cycling. It sits on a little foam pad, and that probably sucked up water (the box is not hermetically sealed.) I cleaned it best I could, and got it to cycle. When that happened the car would start and run. I knew this was temporary (it was badly corroded), so I mail-ordered a new crystal (4 in fact - they're cheap $0.75) from Digikey. Soldered in the fresh one, and the car runs great. Again - if you're not familiar with soldering delicate circuit boards, find a professional.
So it turned out to be the controller, and I was able to fix it. However, when I first started troubleshooting, I suspected the CPS. For this car, not only is it $300, but some of our fine sponsors don't carry it! After some research, I found that the sensing element itself is available, if you're willing to splice it in. You'll also need to re-use the mounting hardware from the bad one. So while I had an order open to Digikey, I ordered a part number 4AV16F Hall Effect Vane Sensor. $28, or one-tenth of a new one (albeit without connector and mounting hardware). Why did I do this? Because the sensors do fail. If I didn't buy it, the one in the car would die next week. By buying this one for a spare, the one in the car will last until the sun goes out.
Note of warning - I'm pretty sure the part number sensor I ordered will fit. But since I did not actually pull the old (perfectly good) sensor and see if the parts would match without playing around, your mileage may vary. However, this sensor is advertised to shops as being the splice-in replacement for the Saab Hall Effect CPS - both 900 and 9000.
Note about splicing - the sensor signal is a relatively robust digital signal, and the current levels are low. No worries about wire size when splicing. Soldering the wires is a good idea (it's a corrosive area), and heat shrink tubing is recommended.
Summary - I suspect DI Controllers may fail due to corrosion on the timing crystal. A simple fix for those with soldering skills. Although I confirmed it with a scope, a simple visual inspection of the crystal showed the problem. (The crystal is the silvery metal can). If the CPS is failed (not uncommon), you can get a splice in unit for one-tenth the cost of a new sensor (Honeywell 4AV16F)
So what about the CPS for '94 and up? That is a variable reluctance sensor that plugs into the block near the transmission, on the side facing the radiator. Much easier to access, and a lot more rugged (it's the same technology used to measure shaft speeds in jet engines). It's also easy to test- pull the connector off the Trionic module, and with an ohm meter, measure resistance between pins 41 and 67 on the harness connector. It should be 485 to 600 ohms. If not, the sensor is bad. If you're off a dozen ohms or so is no big deal - a bad sensor will likely be an open circuit, short circuit, or a couple of thousand ohms.
posted by 75.11.5...
, Sat, 4 Dec 2010 14:22:13 <-- Viewing This Message
1
, Tue, 31 May 2016 23:46:08, Sun, 5 Dec 2010 06:50:27
No Site Registration is Required to Post - Site Membership is optional (Member Features List), but helps to keep the site online
for all Saabers. If the site helps you, please consider helping the site by becoming a member.
 |
 |
 |
 |
 |
