Correct Function of Fuel Pump Relay Signal? (1 Viewer)

[cryorunner]

On my US 1997 LX450, the FPR signal from the ECU (open collector, to pull down and activate the coil on the fuel pump relay) is always pulled low, which causes the fuel pump to always be in low speed mode (~ 9V, powered through the fuel pump resistor). This makes the car always noticeably lean, especially with the supercharger. If I force the fuel pump into high speed mode (jumper to that leg on fuel pump relay or bypass the fuel pump resistor, full voltage to the pump) everything feels like it should be.

The FPR signal looks more like a short, at all times. I’d expect to at least see at least something like a 0.6V junction voltage from the open collector on a DMM in diode mode, but just looks like a short. I think I’ve isolated it to the ECU itself, not any short in the wiring leading to the relay pulling it to ground.

I couldn’t figure out exactly when the switch from low to high speed should happen from the FSM, but its diagnostics say I should see it after 60 seconds “while racing the engine.” I’ve driven it around and had this thing wide open to red-line on the road and never saw the FPR signal change or fuel pump voltage go above 9V (low speed mode).

Can someone tell me for sure when I should be seeing the relay change from low to high speed mode? Or what kind of readings (e.g. diode/junction voltage) I should see at the ECU on that pin when everything is disconnected? I’m reluctant to just spend the money for a new ECU if I could just have the wrong expectations.

Additional pertinent info: the fuel pump relay was swapped out because the old one was so rusted inside it was rattling, and was frozen in its normal position (high speed mode). It had probably been that way for a long time based on how thoroughly nasty the insides were. But there was enough debris rattling around inside that it’s possible that the coil was shorted out at some point (though found as open circuit), maybe explaining why the open collector output for FPR could have been fried, if that’s really the case.

Any info greatly appreciated!

 

[LandLocked93]

For God's sake, what are your symptoms?
Alot of stuff posted, but no cause/effect.

 

[jht3]

For God's sake, what are your symptoms?
Alot of stuff posted, but no cause/effect.

did you miss the first paragraph? he's running lean due to the fuel pump staying in low mode

OP, you are right about the EWD and FSM not being much help w/ diagnostic values. FSM shows tests for the FPR itself, VSV, and circuit opening relay. But I dont see anything about when the ECM should ground the relay. you say you see ECM pin always "pulled low", which to me is the same as grounded; which by my reading of the EWD should put the fuel pump in high mode. but you say this keeps your fuel pump in low, through the resistor. you verified the circuit opening relay is working properly?

are you driving around with a multimeter connected, looking for voltage at the fuel pump? looking for a ground on the ECM pin? curious on your methods. do you have a wideband O2 to monitoring lean/rich conditions? or are you basing your lean conditions on the short/long fuel trims?

 

[DHwreckage]

do you have a wideband installed and how lean are you/when are you lean (closed loop or open loop)? In closed loop the ECU wants the motor to be at 14.7 which would be lean for forced induction but the S/C was designed so you shouldn't really be boosting until open loop which is generally rich. It is my understanding the FPR (fuel pump relay) in stock form should be varying the voltage to the pump depending on what the motor needs, when you start upgrading fuel pumps and injectors etc. for forced induction that's when people start moving to a constant voltage at the fuel pump but regardless the ECU will always try to be at 14.7 when it is closed loop by adjusting fuel trim regardless of bigger pumps or bigger injectors.

 

[cryorunner]

are you driving around with a multimeter connected, looking for voltage at the fuel pump? looking for a ground on the ECM pin? curious on your methods. do you have a wideband O2 to monitoring lean/rich conditions? or are you basing your lean conditions on the short/long fuel trims?

Yep, driving around with two multimeters on the dash, one looking at the voltage going to the fuel pump (pin 3 of the fuel pump relay) and one looking at the voltage on the FPR (pin 1). As long as the circuit opening relay is closed, I was expecting to see FPR alternate between ~12V (pulled up through the coil) and ~0 V (maybe 0.6 V, pulled low from through the open collector output from the ECM). I was expecting the see these change at some point while driving, but always see ~9 V to the fuel pump even full throttle and high revs, which I could have attributed to a bad new relay, but the FPR signal never changes either. And the only other ways I could see that happening is if there was no power from the circuit opening relay (but since pin 5 is common power and coil I'd have no fuel pump at all) or the coil was broken so there was nothing to pull the open collector FPR up (relay tests fine on the bench though).

On the high/low speed vs. FPR being "pulled low" this was very confusing for me at first too, because I see two conflicting diagrams from Toyota. The diagram in my paper Lexus FSM shows that the "normal" position for the relay (no power on the relay coil, FPR not being pulled low by the open collector at the ECM) gives you high speed mode (direct 12V from pin 3 of the relay). But a diagram from the LC EWD I saw on the internet shows the exact oposite - normal position going through the resistor. Bench testing the new relay agrees with my FSM diagram, high speed if FPR is high (no voltage through the relay coil) and low speed if FPR pulled low.

As to why I went hunting for this, it "just didn't feel right" under power after swapping the relay, just not quite the right power off the bottom and noticeable pinging at the top RPMs. Just didn't feel like the last 60 kmiles since I put the supercharger in. Thinking it felt lean at the top I decided to sniff around, started by looking at what the fuel pump was being asked to do and shocked to see it always in low speed mode all the time. Put in my old relay (frozen at high speed mode) back in and none of this. To me it just couldn't be right that the designers would want it in low speed mode at full throttle, but was trying to see if anyone here knew for sure what conditions guaranteed high speed mode so I could feel confident it was the open collector FRP output my ECM that wasn't acting right.

 

[jht3]

On the high/low speed vs. FPR being "pulled low" this was very confusing for me at first too, because I see two conflicting diagrams from Toyota. The diagram in my paper Lexus FSM shows that the "normal" position for the relay (no power on the relay coil, FPR not being pulled low by the open collector at the ECM) gives you high speed mode (direct 12V from pin 3 of the relay). But a diagram from the LC EWD I saw on the internet shows the exact oposite - normal position going through the resistor. Bench testing the new relay agrees with my FSM diagram, high speed if FPR is high (no voltage through the relay coil) and low speed if FPR pulled low.

i missed earlier that you said you had a Lexus. I was looking at the LC EWD b/c that's what I have on my computer. But looking at the 97 LX EWD, it shows the same schematic for the ECM and FP Relay. Relay pin 1 is connected to the ECM, which is the collector of a NPN transistor. the emitter is grounded. they don't show what is happening to the base. my elementary understanding is the transistor is being used as a switch, based on the conditions Toyota doesn't specify.

the relay tests out using this from the FSM?

As long as the circuit opening relay is closed, I was expecting to see FPR alternate between ~12V (pulled up through the coil) and ~0 V (maybe 0.6 V, pulled low from through the open collector output from the ECM).

hmmm. i dont think the voltage the fuel pump sees is changing, but instead the current is being changed by the 0.73 ohm resistor. no?

 

[cryorunner]

the relay tests out using this from the FSM?

Yep, my relay behaves as described in this diagnostic test, and also as shown in the schematic earlier in the diagnostic section of the Lexus FSM. But it’s worth noting for anyone trying to troubleshoot this system that both the Toyota and Lexus Electrical Wiring Diagram schematics show the opposite behavior between pin 2 or 3 being the normal connection to pin 5. The actual behavior shown in this diagnostic (pin 5 and 3 connected when unpowered, bypassing the resistor) is the correct one, but it was confusing at first trying to start from the wiring diagrams.

hmmm. i dont think the voltage the fuel pump sees is changing, but instead the current is being changed by the 0.73 ohm resistor. no?

Sure the current is being reduced, but since the fuel pump resistor and fuel pump in series set up a voltage divider, a quick diagnostic while running was to look at that voltage running fuel pump (e.g. tapping into pin 3 of the relay or other side of the resistor - line that goes to power the fuel pump). While probing there low speed mode shows ~ 9 V and high-speed mode ~13.5 V (full battery/alternator voltage). I was able to jam a sharp probe tip past the rubber boots around the wires entering the connectors.

Sorry for the delay, but thanks for the help!

 

[cryorunner]

Follow up to answer my own questions, just in case someone else ends up trying to troubleshoot something similar in the future. My ECU/ECM did have a burned FPR output, so it wasn’t clear how things were really supposed to work until I finally got a good ECU to compare it to.

First, getting the FPR signal to switch from low to high speed fuel pump mode is pretty easy to accomplish without even having to drive the car, once it’s clear when it’s supposed to switch.

When starting and for a second afterwards should be high speed mode. FPR (connected to a working relay coil) and the power to the fuel pump will be ~13.5 V (battery/alternator voltage, lower while cranking).

During idle after starting the FPR line will be pulled down to ~0.1 V and the fuel pump will see ~9 V, being reduced through the resistor in series.

To make it switch to high speed again while in park, you need to quickly jab FULL throttle. You’ll see the indications above again for a second or so as the engine winds down. This won’t happen if you simply creep it up to redline while parked, it needs the throttle open. The FSM says to “race” the engine, but I didn’t really know what they meant. You’ll also hear the fuel pump relay click, which I think was mentioned in another thread, but I didn’t believe would be possible to hear over the engine running - at least I could when it was out in the open during testing.


As for what the normal readings on the FPR pin on the ECU should be, it should definitely not be a short like mine was. Mine had burned 1/4 of the power MOSFET array controlling the FPR. The other 3/4 control the Circuit Opening Relay (line FC), EGR, and FPU - at least in my - 60650 version, all on IC606 - so this could maybe help in diagnosing issues with control to any of those open collector outputs too. There is no inherent protection for these open collector outputs (as drawn, technically open drain I guess), so something to look out for if you’re having trouble after a potential short like I had.

With the ECU disconnected you should see high impedance (100s of kohms) between FPU (pin 13 on the 22-pin connector) and ground (pin 26 on the 26-pin). Checking with a DMM in diode mode you should see a 0.5 - 0.6 V junction from the MOSFET body diode when measuring reversed, with red on ground and black on FPU. Mine had obviously overheated, shattered the IC package, and presented a dead short to FPR leaving me in low speed all the time.

Hopefully maybe someday this helps someone else trying to figure out problems with the fuel pump relay control or other related open collector outputs from the ECU and maybe avoid having to buy a new ECU just for testing.

 

[jht3]

i was wondering what happened with this. thanks for closing it out.

so you purchased a used ECM? any chance of repairing your failed one?

 

[cryorunner]

so you purchased a used ECM? any chance of repairing your failed one?

Yeah once I got my head around the wiring and how that fuel pump circuit works, I just couldn't see any other cause than a short in the ECU, so broke down and bought one used to confirm that FPR really does switch pretty often.

But I'm WAY too stubborn not to waste time trying fix the old one, and mine is a lot less rusty than the used one I bought (probably overpaid too, trying to buy from a known source).

At least in my 89661-60650 ECM, the quad-MOSFET package at IC606 is an NEC part number uPA1501H (u as in lower-case Greek mu, "micro"). It removed pretty easily, but that FPR pin 11 had gotten very hot during the damage, enough to have burned the conformal coating over the solder, and resulted in that one pad being lifted on removal. Should be easy to reconnect to the trace though.

Big question is what all these "new" uPA1501H chips listed actually are. Seems like everything anymore is either board pulls, sanded and remarked, or otherwise counterfeit. I'm keeping my fingers crossed that what I ordered will be real, and that I pull off actually fixing it. I'm definitely dumb enough to try.