Caterham/Duratec; fuel pump operation?

[Austin David]

Background: 2021 Caterham 360S, Duratec motor.  I'm swapping out the electronics (ECU) and intake with an ITB kit.  Before I unplugged the original ECU, it ran perfectly.

 

How does the fuel pump work?

 

- I can hear it prime when I turn on the ignition.  I can also feel the relay engage.  Removing the relay, fuse OR unplugging the inertial switch will prevent this.  So I feel like the FP is engaging when the ignition is on, and is continuing to receive power.

- I do NOT hear the pump running when I try to start the car.

- I cannot find any direct connection from the original ECU to the fuel pump, and as noted I think the FP is controlled by the ignition, not the ECU.  Based on the fact that the relay closes when the ignition is on, and it stays closed.

 

In the old (2015) manual there are 4 connections shown to the fuel tank; two black (B10 / B11), one GB148 which terminates at the fuel gauge, and one YG168 which terminates at the inertial; other side of inertial is marked YG167.  Confirmed, inertial has two identical-looking yellow / green wires.  As above, with ignition on / ECU disconnected, I can repeatedly prime the fuel pump by toggling that inertial switch.

 

I have been able to "milk" the pump to hold an idle, by unplugging/plugging the inertial every second or so, to engage the FP priming.  Otherwise the FP doesn't seem to want to run consistently, and (without fuel) the engine won't idle.

 

Based on all the above I feel like the FP isn't running enough, but I'm not sure how to trick it into running more.  What am I missing?

[JohnCh]

The fuel pump has a digital control module located on the side of the tank that is controlled by the ECU, which is separate from the relay.  For safety reasons, it cuts power to the pump if the engine stops turning, and the ignition is still on.  When you say you swapped out the ECU, is it still the MBE and did you get it from Caterham or another source?  If the ECU either doesn't have the provision for the controller or if it is programmed to use a different pin than your loom expects, this could explain the problem.  

[Austin David]

Ah yeah this definitely explains the problem.  I'd like to drive that extra control module with my new ECU, any idea where I can track down that connection?  

 

I'm using a whole new ECU, the MBE is disconnected entirely.  The new ME ECU has low side signalling or I can tinker to get the FP running.  Worst case I could bypass the controller entirely and run the relay directly from my new ECU.

[JohnCh]

The controller is on the driver's side of the tank and, on my car, was attached to the boot floor. 

 

 

Keep in mind that because the factory uses a returnless setup, the controller adjusts the fuel pump to achieve expected fuel pressure.  I removed this from my car and went with a return-style system with a fuel pressure regulator, so I have no idea what happens with the factory pump if you bypass the controller.  SBD programmed my ECU to control a second relay that replaced the controller.

 

I took a look at my build notes.  On my car ('22 420R) the signal wire to the pump controller is gray/green, and I think it might be pin 33 on the ECU plug, but you'd have to confirm.  

 

 

[Austin David]

Yep, pin 33 is the signal wire to the controller.  Did some more research, that controller wants some sort of duty cycle.  I can sort of trick it into running if I cycle that pin, but nothing I can obviously make my new ECU simulate.  The new ECU will happily just run a fuel pump via relay, so I'll look into what's required to just bypass this controller entirely, and let the ECU drive the fuel pump as nature intended. 

 

Any idea why that controller is in use?  We already have the inertial, and in some cases an external cutoff.  What does this controller provide that we wouldn't get from an ECU-managed relay?

[JohnCh]

From the research I did earlier, it's a PWM controller used to set the correct fuel pressure based on various sensor input.  In the Ford application, one of the sensors is a fuel pressure sensor.  That makes sense: cycle the pump on and off to maintain a target fuel pressure based on specific and immediate needs.  The Caterham implementation does not use a fuel pressure sensor, so it is not operating in a feedback mode as Ford intended.  It's possible the factory map has a table that adjusts the pump based on certain parameters like engine rpm and temperature, but that's speculation on my part.  If you want to research the Ford application, the part number is AU5A-9D370-FC.

 

 

 

[Austin David]

https://www.fordflex.net/forums/viewtopic.php?t=20842

 

Confirmed the "bypass" method works.  My limited research says basically the same thing: it's used to modulate pressure, based on sensors we don't use, or can cut fuel based again on sensors we don't use.  Odd.

 

With the bypass in place, the fuel pump runs when the relay is on (ignition on).  this might be my best option, I don't see a straightforward way to engage the fuel controller without inventing something worse than a bypass...

 

Thanks for the tips, @JohnCh

[MV8]

I'm trying to follow the logic. You don't have a fordflex or oem ecu, but because this guy could not tell any difference with his current pump condition and no wideband, it's ok?

Edited September 2 by MV8

[Austin David]

My car has a Ford fuel control module on a Ford engine.  I could speculate why it was included on this car, but mostly I'm trying to puzzle out what I miss out on by ditching it entirely.  In theory (hard to test) the Caterham ECU would use it to modulate the fuel flow.  In practice... ? I'm sanity-checking with the ME guys to see if they can drive this fuel pump module, but I assume the recommendation will be to simplify and add lightness, by removing it and splicing or repinning a couple connectors.

[MV8]

Fuel pressure must be stable enough for open loop tune to prevent washing down the cylinder walls with fuel from being too rich or running problems from being too lean. Pump output varies with voltage and wear. The Bosch design return systems have excess capacity pumps for a long service life and the regulator maintains the pressure so the tune can be effective and the injectors sized for idle and low speed control with a vacuum reference to reduce the pressure.

 

It would be odd for Cat to use the ford ecu and not have the necessary sensors on the fuel line on the chassis (not on the engine) but I don't have one to inspect, so I will take your word for it.

 

The simple solution is to add a return fitting to the tank (they sell them as a bolt-on or the pump assy can be modified without modding the tank) and nicopp line with addel clamps parallel to the existing line, then fit a vac ref, ethanol compatible FPR like Aeromotive sells. The existing pump can be reused.

Edited September 2 by MV8

[Austin David]

I know several of those terms

 

To my knowledge there's a regulator in the tank.  If I blow up my fuel pump in the next 8 years I'll look at a return line and regulator up by the fuel rail.  For now, based on what I've seen and touched, there is no regulator outside the original tank and no sensors.  It's a single soft line coming off the fuel tank, clipped straight into the fuel rail, and the rail doesn't have a regulator on it.

 

My setup has a wideband sensor on it, and I will be watching AFR pretty closely.  If I burn a fuel pump early, or if I feel like I'm getting starved for fuel, I'll look at adding more $$ magic in line.  From what I can see so far (famous last words...) I should be OK with the simpler solution.

[JohnCh]

I took a closer look at the in-tank fuel pump assembly, and it does appear to have a fuel pressure sensor.  So perhaps this is less Heath Robinson an implementation than I suspected.  But given they don't even use a post pump filter, my expectations were understandably low  

 

This does mean though that if you bypass the controller, you are running the pump flat out and may have issues.  Is the shop providing the ECU and ITBs familiar with the Caterham or the stock Ford system?  

 

[JohnCh]

Never mind, I just saw your other thread with a link to the kit.  Based on how they pitch their kit, I suspect they should have familiarity with the Ford set up and may have a map that runs the pump.  I would certainly reach out and ask before bypassing anything.

[Austin David]

there are only 4 wires headed down into the tank; I assume (based on other cars) two are for the sender, and two are for the pump.  Is there a feedback mechanism between the fuel pump, and that controller?

 

if that's a new photo, can you stand a little further back and get shots of the wiring up to the socket at top?

[JohnCh]

One blue wire goes to a terminal on the output hose nipple, where I'm assuming it attaches to another part of the sensor.  The other blue wire goes to the pump and terminates on the same pin as a black wire that goes to pin 1 of the lid connector (see last photo).  The other black wire from the pump attaches to pin 4. Pins 2 & 3 are for the fuel level sensor.  The pressure sensor shows part number 2112518 068 and a rating of 625 KPa.  

 

 

 

 

[MV8]

There have been many different methods to control pressure. I don't have any info on that valve, but it looks like a safety relief to prevent pressurizing the plastic assembly above 90 psi/625kpa and the blue wires just ground in parallel to the pump ground. Something designed to have some flow under normal conditions would minimize the splash that causes foaming/aeration.

 

This system looks like pressure would be regulated with voltage level across the pump, not basic hydraulics.

[Austin David]

My plan is to run this one "bang bang" for now; 100% on (bypassing the fuel pump control module), driven by the ECU.  This ECU has a way to run a general-purpose map and PWM, so maybe it's possible to drive that FCM somehow.  Docs are pretty scarce so I'd be guessing... plan "b" would be a $100-150 self-regulated in-tank unit, or a regulator at the rail + return line.  I don't think I'd want to bother trying to fake out that FCM, rather than just rebuilding the fueling with something better understood.

 

If this blows up or causes me any trouble I'll be sure to speak up

 

[Austin David]

Saving some info here, for posterity: 

 

https://www.scannerdanner.com/media/kunena/attachments/417/fpdmtheory.pdf

 

Electronic Returnless Fuel Systems (ERFS) utilize a Fuel Pump Driver Module (FPDM) to control fuel pressure. The PCM uses a Fuel Rail Pressure Sensor (FRP) for feedback. The PCM outputs a duty cycle to the FPDM to maintain the desired fuel rail pressure. During normal operation, the PCM will output a FP duty cycle from 5% to 51%. The FPDM will run the fuel pump at twice this duty cycle, e.g. if the PCM outputs a 42% duty cycle, the FPDM will run the fuel pump at 84%. If the PCM outputs a 75% duty cycle, the FPDM will turn off the fuel pump.

[sf4018]

I encountered the same stuff during a fuel cell installation project a few years ago. It's crazy that the pump control module is basically guessing the fuel demand. With your new ITB the "guess" is no doubt going to be wrong, especially if you have the green injectors. The Pump On command wire GY from the ECU has to be a pulse to work, you cant just leave it on to run the pump, so I ended up removing the pump control module, though It's not as safe. Also the fuel pump really drains the battery quickly when it's on full time, so watch out for that. I made a thread on at the time, it has some diagrams and videos with and without the pump control module, you've figured most everything out already but maybe there will be something in there that is useful...

 

[Austin David]

Thanks, I'll take a look!  Last night I figured out how to drive the FPDM from my ECU, I've got a sensor + gauge on order, and I think I can set up a feedback loop to maintain constant fuel pressure... I should know more this weekend.   I'll post final configs when I've confirmed consistent pressure, but it looks like 10 000 Hz cycle, "negative" sense on my ECU.  75% is indeed off, and the full sweep is indeed 5-50% duty cycle, which definitely ranges from "FP barely running" to "sounds like full blast".  I've also got the ECU taking over the fuel relay.  Pending confirmation over the weekend (and hoping I don't jynx anything) I should be able to have the ECU enable/disable the FP entirely, and drive it from 10-100% to maintain a target pressure.