Weak injector pulse

Hello! I've lurked here for a while, but now I might be able to use some advice. I bought a 1991 900s about a year ago with a bad knock in the engine. The engine ran at that time, and I drove it into my driveway where it has sat since then. Now, I've completed the engine swap and it cranks/no start. It will start easily off brake cleaner then stall, but won't hit a lick otherwise. That tells me the spark and compression are adequate. Fuel pressure is good, over 50 psi while cranking, and holds a little over 40 at rest. My plugs stay bone dry. I checked injector pulse with a noid light, and it is there, but very dim. Per Bently manual, the DC voltage reading at an injector should be .6 v during cranking. I only get about .3. I checked ground to the ecm with connector disconnected and can light a test light full bright. Checked a positive input on pin 9 with the ignition on is present. positive input on pin 35 also good. The constant positive feed to the injectors is good. I also provided a ground to pin 18 and verified its presence at each injector with a test light. With the ecm reconnected I then re-measured the voltage to the injector pulse directly at pin 18 and still observed only .3 volts while cranking and flexing the ecm body and connector. Seems to verify the circuit to the injectors is good from the ecm up.

Bad ECM? or is there more to check?
I noticed an oily substance in the loop leading to the ecm connector. the connector itself seems dry anybody seen that?

Thanks for any input.

Not a lot of activity on these older Saabs so glad you stopped lurking and jumped in. Welcome to the forum.

Don't know if your controller is a motronic or an early trionic, your car is to old to be in my WIS and I've never seen the Bently manual, but I've also never seen a .6 volt injector and if the manual specifically states this voltage when tested with a DVM it's not the way I'm used to checking these things. Injectors need more juice and for a Saab this age I would expect it to have the same 12v Bosch injectors and motronic controller as a Porsche and that the manual would provide tolerance for pulse width duration for checking injector signal with an O-Scope.

Though it's just a guess, I would expect the trionic 5? to get injector operating voltage from the same source as a motronic and when the motronics didn't send voltage to all the injectors it was usually a fuel pump relay giving power to the pump but no power to the controller. At 26 years old these controllers go bad as often as relays, so if your voltages are good to the controller and all your injectors are dry you probably need a new controller.

Because this is a vintage vehicle and because bad things happen to old injectors when they sit with old gas, before you go chasing after an engine controller do check the injector coil impedance because measured voltage changes with impedance and that changes with age. Also, if you can safely rig up a collection system, flow test at least one injector to make sure it's not to plugged up to spit gas. New injectors are a lot easier to find than old engine controllers.

Thanks for the info. My controller has a Jetronic label on it. Bently publishers refers to my fuel injection system as Bosch LH2.4. I think the .6 DC volts is meant to be a ball park indicator for people without access a noid light. We don't expect a meter to display 12v DC on a 12 volt pwm in action. I could measure the duty cycle as a percentage with my meter, but due to the noid light operating dimly, I suspect the pwm on time is less of the issue, but the strength of the pulse, meaning its 'on' cycles are below the 12 volts required.

I've gotten some leads on replacement controllers, but will probably do a little more investigation before committing to that. I'd like to spray the injectors manually as suggested. I don't feel this is the issue because I did swap the 4 from my bad engine with no result. But, with ethenol these days and components that have sat for so long, I could accept the possibility of 8 consecutively bad injectors. I've opened carbs recently on some small engines engines that didn't stand a chance after sitting. I'll also open the ecm that I have before I replace it. My understanding of ecms is that outputs are handled by "drivers". I'm curious to see what the component within the computer drives pin 18's output. Perhaps it can be substituted with a solid state relay, or some other digital component. I think it's worth looking into because I have a pulse, it's just weak.

Probably be a week before I get back to it.

If anybody knows what years the ecms are interchangable, that would help.

If your jetronic looks like this

Attachment 16222

you have a good 5 year spread for replacement parts, you also have a kick panel computer and the oily stuff in the wire harness leading to the ECM is probably engine coolant from a leaky heater hose or core at some time. My first go to for used parts http://car-part.com/ should get you hooked up with what you need at an affordable price.

Hello! Sorry it's been so long, but I haven't been able to work on the car much lately. Spraying the injectors by manually providing power and ground proved the injectors weren't at fault. I was satisfied that my presumption that the ecm was the culprit so I opened the ecm housing to look inside. No burnt components or solder joints. The pin for the circuit that triggers the injectors leads directly to a component that looks like a MOSFET transistor. While I attempted to trace the current in and out of that component, the car started and stayed running. I think my probing with a meter lead may have bridged an open that I can't see, most likely where the ecm connector pins meet the board. I plan to solder a small wire from the appropriate leg on the MOSFET to pin 18 on the inside of the ecm, effectively bypassing the printed circuit board. Whenever I get it done, I'll get a picture.

Never been inside a Motronic, have been inside the Trionic 8 and boy howdy did I miss having a digital Pace board repair station to work with. Don't know if the pins meet the ckt board the same way as the newer controller, but if they do, a little heads up on soldering. If the pin connects are covered with something that feels like silicone vac that solder connection clean before adding your new connection. Scratching and scrubbing doesn't get that stuff completely off and you don't want that stuff mixing in with your fresh solder.

Well, I finally set down to solder the suspected leg in the ecm circuit board, but wasn't able to do it because the space in there was just too small for me to do. I'll clarify what I was going to do. Where the MOFSET comes out of the board, I was going to solder a small wire directly to the output leg halfway up from the board to the MOFSET where I could access it, then solder the other end of said wire to the lead rising out of the board to pin 18 in the connector, where it was most accessable. I've done this in other situations with great success on auto climate control heads, and other various modules to restore operation, but this situation was in too tight a space. Since the car hasn't failed to start since it mysteriously fixed itself during troubleshooting, I just reassembled the ecm and plan to obtain a backup one for when the failure reoccurs. It was fun to drive until the clutch master went out. Now she's parked again, but not for long. I already have the part.