Fuel Pump Relay
12-11-2012, 06:59 PM
#1
Chronic Tool Dropper
Lifetime Rennlist
Member
Lifetime Rennlist
Member
Thread Starter
Fuel Pump Relay
Yesterday I decided to ride down to visit some 928 friends for lunch in Anaheim, somewhere around an hour away. The car was sitting for a couple months while I was away on business. Reattached the battery ground and she started right up as always. By the end of our driveway (~1500 ft), the car was surging as it rolled down the hill on closed throttle, like the IAC valve was sticking. So I turned the engine off and tried to restart. Crank but no start. So we pulled it back up the hill to the garage with a tow strap, Karen in the 4Runner.
Diagnosis started with fuel for some reason, mostly because it seemed more likely than ignition with the surging symptom. Fuel pressure gauge was attached and I cranked the car. About 25 PSI on the gauge, maybe half of what we would expect. I can hear the pump running, so my next thought was a failing fuel pump or maybe a failing pressure regulator.
Regulator and dampers tested first. No leaks into the vacuum side, and the Mity-Vac says the diaphragms hold fine. Next diagnostic step would normally be to pinch the return line with the pump running and watch the pressure. For that I need to install a jumper in the relay socket so the pump will run continuously. Disconnect the battery ground, pull the fuel pump relay, and install a jumper between 30 and 87 in the socket. Every 928 owner carries a jumper for stuff like this, right? Reconnect the ground strap and the pump runs. Wander back to the front of the car, ready to pinch the return line and watch the pressure. Pressure is already about 60 PSI, right where it should be. Reached into the same little bag that held the jumper, pulled out a new relay, installed it in the CE panel socket. Crank the car, and it starts. (always encouraging...). Gauge reads mid-40's, right where it should with engine at idle. So I shut down the engine, removed the pressure gauge, checked for leaks, then and restored all the under-hood bits to as-found.
Credit goes to Jim Bailey, whose timeless mantra "relay-relay-relay!" steers a lot of my troubleshooting efforts.
Of course, in hindsight, several telltale symptoms reinforce the need for the new relay. Perhaps the most interesting is watching the MPG display on the digital dash. The display is an estimated consumption number, calculated from speedo readings and the pulse width of the injector signals to the injectors themselves. With the failing relay and the lower fuel pressure, the injectors open longer to get enough fuel into the engine. With a new relay and with fuel pressure restored, the dash indicates 24+ MPG at 70 cruise on level ground. With the old relay, it would show maybe 20 MPG under the same conditions. Will I pick up 10-20% in actual mileage? Not real likely. But we'll see as I burn through this tank of fuel.
So, message to folks who haven't done so already: Replace the critical relays prophylactically. These would be the LH, EZK, and fuel pump relays at minimum. Later cars use a standard 53B relay for each of these duties. The relays are relatively cheap from our regular suppliers for less than $10/each generally. There are some 'aftermarket' replacements from the local discount parts places that I'm not sure I'd trust in the longer term. Conversely, there are some with heavier ratings available from reputable electrical supply sources that might be interesting.
The spares I have in the car toolkit are 40A rated vs the 30A of the factory pieces. I see some 50A ratings on some but question their ability to pass that much current through a 1/4" spade terminal. Nonetheless, if the internal contacts are that large and well braced, maybe they would last longer than what the factory used. If my new ones last another 100k miles, I'll be in my late seventies when I need to start thinking about them again. There's also the chance that by then it will be way too hot for me to drive, and restoring fuel flow will be the last thing on my to-do list.
I hope this little episode helps someone who experiences similar symptoms.
Diagnosis started with fuel for some reason, mostly because it seemed more likely than ignition with the surging symptom. Fuel pressure gauge was attached and I cranked the car. About 25 PSI on the gauge, maybe half of what we would expect. I can hear the pump running, so my next thought was a failing fuel pump or maybe a failing pressure regulator.
Regulator and dampers tested first. No leaks into the vacuum side, and the Mity-Vac says the diaphragms hold fine. Next diagnostic step would normally be to pinch the return line with the pump running and watch the pressure. For that I need to install a jumper in the relay socket so the pump will run continuously. Disconnect the battery ground, pull the fuel pump relay, and install a jumper between 30 and 87 in the socket. Every 928 owner carries a jumper for stuff like this, right? Reconnect the ground strap and the pump runs. Wander back to the front of the car, ready to pinch the return line and watch the pressure. Pressure is already about 60 PSI, right where it should be. Reached into the same little bag that held the jumper, pulled out a new relay, installed it in the CE panel socket. Crank the car, and it starts. (always encouraging...). Gauge reads mid-40's, right where it should with engine at idle. So I shut down the engine, removed the pressure gauge, checked for leaks, then and restored all the under-hood bits to as-found.
Credit goes to Jim Bailey, whose timeless mantra "relay-relay-relay!" steers a lot of my troubleshooting efforts.
Of course, in hindsight, several telltale symptoms reinforce the need for the new relay. Perhaps the most interesting is watching the MPG display on the digital dash. The display is an estimated consumption number, calculated from speedo readings and the pulse width of the injector signals to the injectors themselves. With the failing relay and the lower fuel pressure, the injectors open longer to get enough fuel into the engine. With a new relay and with fuel pressure restored, the dash indicates 24+ MPG at 70 cruise on level ground. With the old relay, it would show maybe 20 MPG under the same conditions. Will I pick up 10-20% in actual mileage? Not real likely. But we'll see as I burn through this tank of fuel.
So, message to folks who haven't done so already: Replace the critical relays prophylactically. These would be the LH, EZK, and fuel pump relays at minimum. Later cars use a standard 53B relay for each of these duties. The relays are relatively cheap from our regular suppliers for less than $10/each generally. There are some 'aftermarket' replacements from the local discount parts places that I'm not sure I'd trust in the longer term. Conversely, there are some with heavier ratings available from reputable electrical supply sources that might be interesting.
The spares I have in the car toolkit are 40A rated vs the 30A of the factory pieces. I see some 50A ratings on some but question their ability to pass that much current through a 1/4" spade terminal. Nonetheless, if the internal contacts are that large and well braced, maybe they would last longer than what the factory used. If my new ones last another 100k miles, I'll be in my late seventies when I need to start thinking about them again. There's also the chance that by then it will be way too hot for me to drive, and restoring fuel flow will be the last thing on my to-do list.
I hope this little episode helps someone who experiences similar symptoms.
12-12-2012, 01:23 PM
#4
Rennlist Member
12-12-2012, 02:19 PM
#5
Chronic Tool Dropper
Lifetime Rennlist
Member
Lifetime Rennlist
Member
Thread Starter
PM is too close to PMS to fit the exact intent.
12-12-2012, 04:10 PM
#7
Rennlist Member
Trending Topics
12-12-2012, 07:41 PM
#8
Addict
Rennlist Member
Rennlist Member
A little more than two years ago, I was chasing down a really bizarre problem with my '90 S4 after a top-end refresh, where, when warm, I'd start the car and it would run for 75 seconds and then stop. If started cold, it would run for a little more than 10 minutes and then stop. Upon restart warm... 75 seconds, then stop.
You can read where I first noticed the 75 second aspect of the problem at post #117 in the long thread "Yet another top-end refresh (WAS: Trying to fix irregular/poor idle ...)". The dozen or so posts before that include some more descriptions of the problem, and a few posts suggest possible relay problems.
What's interesting, though, is that (see post #144) when I finally solved the problem, I (unfortunately) did two things at once:
At that time, I assumed (for some reason; perhaps the almost uncanny repeatability of the 75 second delay before the engine would stall) that it was the PEMs or just fooling with the LH (jiggling wires around, reseating connectors, etc.) that fixed the problem, but in retrospect, I'll bet it was the fuel pump relay.
I wish I would have saved the old relay (maybe I did, but I doubt it's marked in any way that would uniquely identify it), because I'd really like to test it and see if I could isolate it as the cause.
To this day, the problem has never reappeared.
You can read where I first noticed the 75 second aspect of the problem at post #117 in the long thread "Yet another top-end refresh (WAS: Trying to fix irregular/poor idle ...)". The dozen or so posts before that include some more descriptions of the problem, and a few posts suggest possible relay problems.
What's interesting, though, is that (see post #144) when I finally solved the problem, I (unfortunately) did two things at once:
- put PEMs in my LH (because I was going to use my SharkTuner to assist with further diagnostics); and
- replaced the fuel pump relay.
At that time, I assumed (for some reason; perhaps the almost uncanny repeatability of the 75 second delay before the engine would stall) that it was the PEMs or just fooling with the LH (jiggling wires around, reseating connectors, etc.) that fixed the problem, but in retrospect, I'll bet it was the fuel pump relay.
I wish I would have saved the old relay (maybe I did, but I doubt it's marked in any way that would uniquely identify it), because I'd really like to test it and see if I could isolate it as the cause.
To this day, the problem has never reappeared.
