My wife's '04 Acura TSX began having issues starting towards the end of last summer. It would turn over just fine, but took 5-10 seconds to actually start running. The problem was intermittent - some days, it would start fine, other days were problems.

The problem went away for the winter, and came roaring back this week in the returning summer heat, leading us to conclude that the problem was heat-related. Several folks online suggested that Hondas and Acuras often have trouble with their fuel pump relays in the heat, and suggested that replacing the fuel pump relay might be a good start to diagnosing the issue.

After finding this, I've taken some diagnostic steps that have resulted in the car not starting at all. I'm wondering if anybody can shed some light. Here's what I've done:

  1. Removed the fuel pump relay, hooked it up to a 12V power source, and checked that the circuit was completed as it should have been. (This was the suggested method for checking that a relay was working.) The relay closed once, but I could not get it to close again.
  2. Put the original fuel pump relay back in (after testing). The car did not start at all at this point.
  3. Purchased an aftermarket relay, which turned out to be incorrect (auto parts store had wrong part in the box). I inserted this, and the car did not start.
  4. Discovered that the part was boxed improperly, went back to the parts store, and had them change it out for the correct part. Popped it in. Still no dice.
  5. Following this document from Acura http://www.tsxclub.com/forums/content/manual_gen1/Engine/Fuel%20Supply%20System.pdf I measured voltage between pin 3 of the fuel pump relay socket and ground. Did the same for pin 1. Got a solid 12v both times, suggesting there is no problem with the main relay, or the fuel pump fuse.

At this point, I'm struggling to think what might be going on. Perhaps I somehow burned out the original relay when testing it (not really sure how that's possible, but seems to fit what's going on). Perhaps the aftermarket fuel pump relay is not a good match and doesn't work well (I've got a used OEM on the way now). Perhaps inserting the incorrect relay that was boxed improperly fried something? Perhaps the fuel pump was the problem all along, was already on its last legs and just happened to die right as I was swapping out the relays (seems awfully coincidental...)?

Any suggestions would be very much appreciated. I'm at a loss for how unplugging a relay and plugging in an aftermarket replacement would leave my car failing to start.

  • Is this issue persistent regardless of whether the car has been sitting for a long period or not?
    – Zaid
    Commented May 31, 2016 at 10:23
  • Yes. It's frequently happened after sitting in front of the house all day. But, just this week, it struggled to start after my wife had spent 5 minutes in Target. Commented May 31, 2016 at 11:37
  • In that case your problem may not be limited to the fuel pump relay in that case. a bad fuel return check valve, for instance, could explain your symptoms.
    – Zaid
    Commented May 31, 2016 at 11:43

1 Answer 1


The power in the in pins 1 and 3 proves that fuse 19 is OK and that the PCM will turn on PGMFI relay #1 and further that said relay will supply power to the fuel pump relay (PGMFI relay #2). It does not prove that the relay is good. My next tests would be apply power to Pin #2 and see if the fuel pump activates. If not repair the pump. If the pump works check to see if the PCM connects pin 4 to ground for two seconds at key on and all the time during cranking.

Fuel pump failures are very rare in this model but anything is possible.

The link you provided is the OEM service manual test series for the fuel pump and its circuit. It goes the long way around to find the fault. But it will get you there if you follow all the steps.

  • Thanks for this. Can I potentially hurt anything if I manually close the circuit between pins 1 and 2 with a wire or something? Seems like the answer should be "no" but I don't want to cause further damage. Also, thanks for the tips on easier troubleshooting steps. I got about as far as I could in the service manual before I started getting lost with some of their suggested diagnostics... Commented May 31, 2016 at 11:40
  • 1
    It is OK to jump 1 and 2. I often check circuits at relays by jumping the load pins to see if the components operate. Pin 4 is the " you could cause damage pin" since it is connected to very expensive controller. Professional techs do not often use the test sequence in the manuals, it takes too long. Commented May 31, 2016 at 15:17
  • @FredWilson +1 on pinpoint tests taking way too much time.
    – Ben
    Commented May 31, 2016 at 18:27
  • Alright. Jumped pins 1 and 2, and the fuel pump runs like a champ. Stuck my multimeter between pin 4 and body ground. I get nothing when the fuel pump shouldn't be running, and I get 5V when the fuel pump should be running (e.g., first 2 seconds of key on). Is this expected behavior? Commented Jun 1, 2016 at 0:45
  • I have not checked the voltage on that circuit in the past. This 5 volts is likely a circuit check voltage. It gets defeated by the 12v coming out of the relay coil. I would check pin 4 with an ohmmeter even with the 5 volts coming out of the PCM. We are looking for a low reading during pump command. Commented Jun 1, 2016 at 2:57

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .