Fuel Pump Not Priming

[goldNSX]

Asking for advice and giving advice with respect to a modified car (AEM V1 ECU) is a bit of a crap-shoot. An unmodified car is much easier because everybody has a common understanding of what the hardware arrangement is.

+1
I've once helped a guy out with an aftermarket ECU to start his car which obviouly ran too lean in freezing conditions. That doesn't have to be your problem. But an aftermarket ECU demands ALL of the tuning on your own.
If you say that you have no 12V at the fuel pump resistor did you take into consideration that the fuel pump act up for only 2 seconds if you don't turn the engine over?

Do you still have the Walbro in the fuel tank? Walbors are crap in the longrun, you'd better get a serious one.

But please check the things [MENTION=26435]Old Guy[/MENTION] mentioned because he's older than I am. :wink:

 

[furinax]

Thanks for your input. Of course I understand custom cars will always have some degree of uncertainty.

The 12V pulse does come on when warm at the resistor point (when everything works). I tested that last week. I would not be surprised if the fuel pump resistor was bypassed in which case I would have to take the seat out and check the voltages at certain terminals. The wiring diagrams certainly help and I appreciate you guys sharing your wisdom. In the end it is a learning experience for me. It seems like in the cold all sorts of electrical connections can get finicky.

 

[Old Guy]

I would not be surprised if the fuel pump resistor was bypassed in which case I would have to take the seat out and check the voltages at certain terminals.

I am not an expert on the matter because my new main FI relay remains in a sealed bag somewhere in my stash of critical spares. However, I didn't think you had to remove the seats to get the rear bulkhead covers off. With the seat all the way forward I figured you should have sufficient room to get your fingers under the cover to pop the lower portion off.

 

[furinax]

The fuel pump relay is normally located on the inside back bulkhead.​

This is what I am trying to find. It sounds like it's behind the driver's seat, but hidden behind the firewall if that makes sense.

I hear clicking at the main relay and the aforementioned spot when priming. There is clicking but no fuel prime sound. Checking electrical contacts and everything seems okay (~12V pulse when priming). My current theory is that there is water in the fuel line somewhere which is freezing and doesn't allow fuel to pressurize. I added some Valvoline water remover a few days back and fired it up today. Hopefully we get another cold day (below freezing) so I can test this theory.

If it is still stuck, I bought a M18 Heat Gun since I don't have a garage to run wires from and will start by heating up the ECU, main relay, and the fuel pump relay if I can find it. If all those check out it must be an open circuit in the last piece of wire.

I envy all you guys with a garage!

 

[Old Guy]

You seem to be spinning a bit with your trouble shooting. I suggest that you go back to post #23 and follow the advice [MENTION=31635]emac[/MENTION] provided about testing the voltage on the + supply to the fuel pump where the wire to the pump passes through the rear bulkhead to the fuel tank. The + supply to the pump is on the blk/red wire. Set up a semi permanent connection of a 12 volt light bulb between the blk/red wire and ground. The light bulb should light up whenever the pump is commanded to run by the ECU (steady operation and prime pulse). If the light comes on and the pump does not run or prime, then you have a failed or failing pump or an intermittent electrical connection at the final connection at the tank. If the light does not light up during the key on Prime sequence or during cranking of the starter, then you have an upstream electrical problem which could be the main FI relay, a bad connection at the fuel pump / resistor relay, an intermittent problem with the +12v supply to the main FI relay or there is a problem with the fuel pump control output on the AEM ECU (if you hear a relay clicking this seems .

The heat gun test works in theory; but, is very difficult to carry out in practise. I would do the test using the test light first to try to narrow down whether the failure is on the pump side or the up-stream electrical side. You made the comment

I hear clicking at the main relay and the aforementioned spot when priming. There is clicking but no fuel prime sound.

. During the start / prime sequence the only clicking you should hear is from the Main FI relay. If by 'aforementioned spot' you mean the fuel pump / resistor relay, the fuel pump relay which by-passes the fuel pump resistor does not get energized during the prime / start sequence. The only time you should hear that relay engage is when the ECU switches to the VTEC mode of operation and the chance that you will actually hear the relay click with all the back ground noise are pretty much close to zero. If you are hearing clicking at some other location than the main FI relay, that is something else. Are you sure that you don't have one of those pump booster things?

As a final note, ice in the fuel lines will prevent fuel delivery; but, will not prevent the pump going through its prime cycle. The only ice related problem that would stop the prime cycle is ice inside the pump. If you have a good chunk of ice inside the pump that problem will generally not be intermittent and will generally not go away until you have a protracted period of weather with temperatures well above freezing. Your problem has been going on for a couple of years so I would be inclined to put the ice in gas theory on the bottom shelf. Besides, the NSX has a sealed fuel system making it relatively immune to accumulating moisture in the fuel system unless you are consistently getting water in your gas from your fuel supplier. Are you by any chance running E85 which seems to be able to suck water out of the air and form nifty sludge like mixes during cold snaps?

 

[furinax]

Thanks for your wisdom, Old Guy. I was able to do a few electrical measurements in the cold today after taking off the casing of the main relay. No LED light installed yet, but I have been breaking wires left and right and resoldering them so I may as well install one at one point.

On prime cycle:
pin 3 and pin 8 have 11.6 V as you would expect
continuity between pin 2 and GROUND
pin 7 to the fuel pump gets NO voltage
shorting pins 4 and 8 causes a weird groan by the back of the car and the Fuel Pump primes as it should! Success
Heating the relay with no case doesn't fix the problem
Swapping the relay doesn't fix the problem

Should I have tested pin 5?
What is the "ST. Switch" ??

 

[Old Guy]

On prime cycle:
pin 3 and pin 8 have 11.6 V as you would expect
continuity between pin 2 and GROUND
pin 7 to the fuel pump gets NO voltage
shorting pins 4 and 8 causes a weird groan by the back of the car and the Fuel Pump primes as it should! Success
Heating the relay with no case doesn't fix the problem
Swapping the relay doesn't fix the problem

Should I have tested pin 5?
What is the "ST. Switch" ??

What year is your car? The picture suggests 2002+; but, it is always useful to confirm since some early cars have been fitted with 2002+ bodywork.

The reason this is important is that at some point Honda changed the main FI relay. Truth be told, it does not look like they changed the relay, more like they just changed the pin numbering in the service manual. It is not clear why they would do this, maybe just to mess with our heads. I have attached a photo of the pin numbering from the 1991 service manual and a photo of the pin numbering from the 1997+ service manual.

Pin numbering from the 1991 manual

View attachment 158700


Pin numbering from the 1997+ manual

View attachment 158699

The fact that you are referring to pin 8 makes me think you are using the 1991 pin designations; but, we need to confirm.

Using the 1991 pin numbering:
- pin 3 is the +12v supply to the ECU and fuel injectors from the main relay and should have +12v when the ignition switch is in the run position (whether or not the engine is running).
- pin 8 is the control signal for the portion of the relay that controls the fuel pump. When the ignition switch is in the run position; but the fuel pump / engine is not operating, this pin will float at +12v. In order to run the fuel pump (and the engine) the ECU connects pin #8 to ground which then powers up the relay coil between pin 4 and 8 which closes the relay contact which powers up the fuel pump. When the pump is running the voltage on pin #8 will be close to 0 volts. The voltage will be determined by the value for Vce saturated for the fuel pump control transistor in the ECU.
- pin 7 to the fuel pump only gets +12v when the fuel pump is running, so the engine must be running. Pin 7 will also get +12v during the 2 second prime sequence for the pump.
- shorting pin 4 to pin 8. Did you do this with the relay plugged in? If so all you did was apply a short across the relay coil that operates the contact for the fuel pump. This would prevent that relay coil from operating and energizing the fuel pump. With a short across pins 4 and 8, the fuel pump should most definitely not have operated. You either made a mistake in which pins you applied the short to or your vehicle wiring has been altered. Using the 1991 pin numbering, if you shorted pin 4 to pin 8 and left that short in place while you attempted to start the car you run an exceedingly high risk of frying the transistor in the ECU which controls fuel pump operation.
- ST Switch is the start or cranking signal from the ignition switch. The ECU will not cause the fuel pump to operate in continuous mode until it detects that the engine is running - typically a threshold above a couple of hundred RPM. In order to get the engine running during cranking the ECU and the FI relay get a start / cranking signal from the ignition switch which tells it to forces the fuel pump on even though the engine speed may be below the run threshold.

 

[furinax]

Mine is a 1992. I'm fairly sure I have the pins right, though I did have my doubts. I'll check again tonight what the voltages are at the pins. It's fairly simple but with an aftermarket system you gotta observe what's going on. No more shorting pins on my side since I wouldn't risk damaging components.

 

[furinax]

OK the weather finally dropped below 30F tonight and I got some measurements. Comparing it with the measurements when the weather was normal, here are the results:

1. Definitely the NA1 relay. I can tell since they have the numbering printed on the circuit board.
2. Pin #7 does not get 12V on the priming cycle.
3. The problem seems to be with pin #8 . It stays at 12V and doesn't drop to ground when cold.

I can only assume that my AEM somehow has a faulty ground at below freezing temps. The regular ground on pin 7 reads fine in all cases. I know AEM v1 is supposed to run just fine in the cold, so maybe mine is just getting old. What I don't understand is that something is clicking in the cold, but it is just a softer click than the normal relay operation. The pins all experience some voltage change when going through the priming cycle in normal + stuck conditions, but you don't get that big grounding action on pin 7. Hopefully I still have many more years on my AEM. I would prefer to stay boosted.

I didn't short the relay anymore, but from what I recall when I first did it, shorting (3 and 7) or (4 and 8) caused the same growling sound behind the firewall. But only shorting 4 and 8 briefly would cause the prime to activate. I will not speculate but this is just what I observed.

 

[Old Guy]

2. Pin <a href="http://www.nsxprime.com/forum/usertag.php?do=list&action=hash&hash=7" target="_blank">#7 does not get 12V on the priming cycle.

Just to confirm. The prime cycle only lasts about 2 seconds shortly after you turn the ignition key to the run position. Pin #7 only goes to 12v during that 2 second period. It will be at ground potential otherwise until the engine starts. If you have +12v on pin 5 and you don't get 12v on pin #7 when the prime pulse 'should' occur, then you either have a problem in the fuel pump portion of the main relay (a solder joint fracture that opens up when cold?) or you are failing to get the ground signal from the ECU - your 'next' observation.

3. The problem seems to be with pin <a href="http://www.nsxprime.com/forum/usertag.php?do=list&action=hash&hash=8" target="_blank">#8 . It stays at 12V and doesn't drop to ground when cold.

Three possible causes.

1st - intermittent wiring connection between pin #8 and the ECU. Does your AEM use a short adapter harness to interface between the AEM and the vehicle ECU connectors? If so, check the connector at the ECU and the adapter harness to see if there is a wiring problem or damage to the pins in the plugs. A temperature related failure of this type would be rare; but, not impossible.

2nd - there is temperature sensitive problem internal to the AEM. Not likely an electronic device problem; but, could be the same type of solder joint cold fracture problem that the main relay can suffer from. If this were an OEM ECU, I would suggest watching the check engine light to make sure that the ECU is successfully completing the power on self test (POST) - light goes on then goes off. If the AEM uses the check engine light function, make sure it does the POST successfully. If the AEM does not utilize the check engine light function, is there any way to confirm that the AEM has powered up? There must be a serial port on the AEM? Can you use that port to connect a laptop and confirm that the AEM is powering up when the temperature is cold?** If the AEM is not powering up (option #3 below) when cold then there may be an upstream power supply problem to the AEM. If the AEM does the POST; but, is not generating the ground connection to pin #8 , then it appears that there is some type of temperature sensitive problem internal to the AEM. If AEM still supports your series 1 sending it to them for diagnosis might be an option; but, since your problem is intermittent and seems to be temperature sensitive replicating the failure on a test bench would likely have a low probability of success.

3rd - The AEM is not powering up. You previously confirmed that pin #3 on the main relay gets 12v on power up. Pin #3 is the ECU power up so it does not seem to be a supply problem to the AEM unless there is a wiring harness problem between the main relay and the ECU.

** I have another car with an aftermarket ECU and the interconnecting software has a dashboard with a series of status indicators. One of those status indicators is a fuel pump output status. If the fuel pump status indicator lights up during the prime cycle; but, the pump does not run that means that the software is doing its 'thing'. In such case there might be a temperature sensitive connection problem on the AEM board for the fuel pump output transistor or there might be a wiring problem. If the AEM software has that feature that would go a long way towards narrowing the location of the problem.

 

[Mac Attack]

Just to confirm. The prime cycle only lasts about 2 seconds shortly after you turn the ignition key to the run position. Pin #7 only goes to 12v during that 2 second period. It will be at ground potential otherwise until the engine starts. If you have +12v on pin 5 and you don't get 12v on pin #7 when the prime pulse 'should' occur, then you either have a problem in the fuel pump portion of the main relay (a solder joint fracture that opens up when cold?) or you are failing to get the ground signal from the ECU - your 'next' observation.

There's one more possibility - Perhaps his AEM calibration file does not include a prime sequence. This is an adjustable parameter. FYI - I have mine set for the OEM duration of three seconds.
[MENTION=31854]furinax[/MENTION]
I can check your AEM cal file to see if yours is programmed for a fuel pump prime sequence if you want to email it to me.

 

[Old Guy]

There's one more possibility - Perhaps his AEM calibration file does not include a prime sequence. This is an adjustable parameter. FYI - I have mine set for the OEM duration of three seconds.

That is a good point. I had presumed that failure to prime was also accompanied by failure to start, i.e. pump not running in either the prime sequence or when the key is switched to start. However, in reviewing the thread details that was never explicitly spelled out. If in fact the engine is starting when the key is in the start / crank position, then this may purely be a configuration issue - which would have existed from the date that the software was last configured.

 

[furinax]

I think my best bet is to check the AEM using the serial port. I wonder if I have a computer with a serial port still!!! I could check the harness but that seems like a low failure thing to me. All the wires I have checked have been good- and I only damage things by moving them around.

There is a 2s prime cycle when the key gets turned on. This prime cycle is what is not working when I said "fuel pump not priming". It's very loud and audible- both the clicking of the relay and the fuel pump running.

Thanks so much for your help. I should take a picture to show you guys what I'm dealing with back there before I put it all back together. We may not get many more cold nights here so I may just drop the issue if it's internal to the AEM.

 

[Mac Attack]

I have to admit I'm a little confused.

From what I understand, this thread went from not priming at all to now only priming when it is warm outside after replacing a lot of stuff?

Is that correct? If so, can you please list all of the parts you have replaced?

 

[Old Guy]

I think my best bet is to check the AEM using the serial port. I wonder if I have a computer with a serial port still!!! I could check the harness but that seems like a low failure thing to me. All the wires I have checked have been good- and I only damage things by moving them around.

If your laptop was made after 2000, you have one - USB (Universal Serial Bus) port. My understanding is that the AEM series 1 did not support communications via USB. Depending on the vintage of the AEM, it might use an RS 232 serial hardware interface (most commonly a 9 pin plug that sort of looks like a VGA plug; but, does not have to look like that). You can buy USB to RS 232 adapters that work well. The adapters based upon the FTDI chip seem to work the best. You will have to install a software driver to use the FTDI adapter.

Is your AEM series 1 of the same vintage as the one in this link?

https://www.plex-tuning.com/instructions/SDM_Instructions/aem_series_1.htm

If so, definitely looks like RS232.

 

[furinax]

2nd - there is temperature sensitive problem internal to the AEM.

Updating this thread because there was a fb post about this, not because I found some sort of closure or source cause. My experience leads me to believe that this second reason Old Guy mentioned is what is wrong. When temps drop below 36 F, the pin #8 on a '92 main relay doesn't drop properly during the prime phase. This causes a "start click" and "stop click" (Relay sounds) but not the whirring prime sound. I still blame the AEM v1 since others seem to also have this issue, but it is not very well documented and the v1 isn't much supported anymore.

The result is not too bad because this only happens in the winter, and I only have summer tires. When the car is left outside in the sun for a while or if it was still warm from being driven recently (a few hours) then the fuel pump still primes. It wasn't a bad main relay, it wasn't ice in the fuel hose, and it wasn't a bad fuel pump. I also had a bad starter (two in fact!) and a faulty ignition switch, but those lead to different symptoms which were easy to replace. So I apologize for the confusion I may have caused for the unrelated issues and I thank you guys for your patience in dealing with this.

All the problems become non-issues when I bought a house with a garage. This leads me to understand that you need a garage to own such a classic car as this. Not only is it a requirement for classic car insurance, but my garage never goes down below 50F so the car never has this issue. Also, the battery is on a tender now so the parasitic drain and the cold temps won't cause funky electrical issues or needing to jump start, which can in turn damage the electronics. On top of that being out of the sun and elements as well as having privacy to work on the car, it's just better overall. At the time the garage was not an option for me, and it may not be for some potential buyers. Hopefully this is a good lesson.