fuel pump primes 17 times then stops, engine won't run
Discussion
Penelope Stopit said:
Toilet - Are you a member and able to reach this topic?
ECU /relay box wiring diagram - Ariel.Club Forum
https://ariel.club/forum/viewtopic.php?t=15749
Yes I'm a member on there. I posted a duplicate thread on that forum the same time as on here but didn't get any luck solving. To be honest that community is pretty small, especially compared to the number of experts on PH ECU /relay box wiring diagram - Ariel.Club Forum
https://ariel.club/forum/viewtopic.php?t=15749
To answer your other questions, the Atom comes prebuilt (you can't buy it as a "kit") and I've had it circa 7 years
Edited by Toilet Duck on Tuesday 30th October 19:54
Edited by Toilet Duck on Tuesday 30th October 19:55
Mignon said:
Toilet Duck said:
I think (assume)
Don't assume - measure.Mignon said:
Toilet Duck said:
it runs rough because the fuel pump isn't constantly running, it primes for 2-3 seconds then stops, then repeats a total of 16 times.
So does the engine run rough exactly in sync with the pump cycling or is it rough all the time?Mignon said:
Toilet Duck said:
So I'm guessing the injectors are getting fuel but the delivery pressure is fluctuating with the cycling pump. Even if the pressure wasn't "right" I don't understand why that would make the fuel pump (and throttle relay!) cycle 17 times.
What makes it even stranger is that with the fuel pump disconnected the Freelander relay module thing and throttle relay STILL cycle 17 times
I'm not really fussed about the 17 times thing. Presumably the ecu has to have some number programmed in to it to try and operate something before it gives up. 17 times sounds as good a number as any except maybe 42 which is obviously the universe's best number. Stop trying to figure out how every bit of the ecu system is designed and measure the things that are important. Fuel pressure for starters.What makes it even stranger is that with the fuel pump disconnected the Freelander relay module thing and throttle relay STILL cycle 17 times
Try sticking a 12v feed direct to the fuel pump. Does the engine start and run ok?
stevieturbo said:
Cant be arsed reading every reply...but has the OP confirmed that the ecu is staying live throughout these 17 cycles, or more likely for whatever reason the ecu is losing power and going live again...hence priming every time it powers up ?
No, but I will call the place that currently has my car tomorrow and ask, good question!Many thanks for the continued help etc.
I've asked the place my car is currently at some questions based on what's been suggested here, please see below. For some reason the formatting isn't working (I tried to put the responses in bold)
I was looking on various forums online and one thing was suggested that
I thought maybe worth passing on to you:
the ECU could be using a 4 bit "retry" counter that starts at 0000
(binary) and increments by one after each cycle.
By the time it's got to 1111 (binary) it will have tried 17 times. It
doesn't help identify the problem though, which presumably is some kind
of failure with the ECU deciding it can't (or shouldn't) start, so tries
the cycle again.
[b]We can't access the ecu at binary level, however swapping the ecu should
have eliminated the possibility of the ecu being at fault.[/b]
Is the ECU constantly powered up during the "17 cycle"? If it
was losing power for some reason would that cause it to cycle the fuel pump
etc. each time?
[b]I did not check this, I only checked the fuel pump output from the ecu
which was cycling with the relay clicking. I would have thought that if
it was the main power supply that was switching on and off that the dash
would also switch on and off too and I didn't notice that happening.[/b]
One other thing just so I fully understand, if my ECU is plugged into
your diagnostic equipment is it able to run some sort of all
encompassing "system check" that would show any and all
faults/corruption etc (if present) within the ECU? Or is it just that my
ECU has been swapped out with a known working one and as the symptoms
remained it is believed my original ECU is "good"??
[b]The nearest I could get to checking sensors etc would be to lock the
Flashpro unit to your ecu and carry out some datalogs. As explained on
the telephone, if I do this the ecu calibration will be lost and the car
will have to be tuned on the dyno at a cost of £400.00 I can do this if
you wish.[/b]
I know you explained that the red thing by the fuel pump is a
pressure regulator that diverts fuel back into the tank if excess
pressure is delivered. If this regulator was faulty (either allowing too
much or too little pressure), could that cause these symptoms
[b]No, as the fuel pressure regulator (the red thing) is purely mechanical
and not linked to the electronics.[/b]
If I can't get this resolved locally, which is looking highly likely, then I'll take it back to the factory
Cheers
I've asked the place my car is currently at some questions based on what's been suggested here, please see below. For some reason the formatting isn't working (I tried to put the responses in bold)
I was looking on various forums online and one thing was suggested that
I thought maybe worth passing on to you:
the ECU could be using a 4 bit "retry" counter that starts at 0000
(binary) and increments by one after each cycle.
By the time it's got to 1111 (binary) it will have tried 17 times. It
doesn't help identify the problem though, which presumably is some kind
of failure with the ECU deciding it can't (or shouldn't) start, so tries
the cycle again.
[b]We can't access the ecu at binary level, however swapping the ecu should
have eliminated the possibility of the ecu being at fault.[/b]
Is the ECU constantly powered up during the "17 cycle"? If it
was losing power for some reason would that cause it to cycle the fuel pump
etc. each time?
[b]I did not check this, I only checked the fuel pump output from the ecu
which was cycling with the relay clicking. I would have thought that if
it was the main power supply that was switching on and off that the dash
would also switch on and off too and I didn't notice that happening.[/b]
One other thing just so I fully understand, if my ECU is plugged into
your diagnostic equipment is it able to run some sort of all
encompassing "system check" that would show any and all
faults/corruption etc (if present) within the ECU? Or is it just that my
ECU has been swapped out with a known working one and as the symptoms
remained it is believed my original ECU is "good"??
[b]The nearest I could get to checking sensors etc would be to lock the
Flashpro unit to your ecu and carry out some datalogs. As explained on
the telephone, if I do this the ecu calibration will be lost and the car
will have to be tuned on the dyno at a cost of £400.00 I can do this if
you wish.[/b]
I know you explained that the red thing by the fuel pump is a
pressure regulator that diverts fuel back into the tank if excess
pressure is delivered. If this regulator was faulty (either allowing too
much or too little pressure), could that cause these symptoms
[b]No, as the fuel pressure regulator (the red thing) is purely mechanical
and not linked to the electronics.[/b]
Kccv23highliftcam said:
On, or rather in, your mfrp do you have an internal or external line to pin 30 of the fpr.
(May as well start at the bottom)
I don't have my car at the moment (picking it up next week). Apologies but is the "MFRP" the same as the "MFRU" Freelander module? What is the "FPR"?(May as well start at the bottom)
Penelope Stopit said:
I posted the OP a link https://www.lotus7.club/forum/techtalk/which-fault... to better diagrams and information that mentions the same as you have above
The thing is the OP's car was a runner and developed a fault that caused the fuel pump to keep running rather than for a 2 second or however long prime
Me bearing in mind that the OP didn't alter any wiring but did start (excuse the pun) switching the ignition on and off to check the priming function explained in another post what could cause the fuel pump to keep running, I also posted my thoughts on what may have been happening due to switching the ignition on and off up to 17 times (flooding came to mind)
The OP hasn't been back here to discuss all posters findings, at the bottom of this post are the good diagrams from the above link and my thoughts about the problem and causes
Sorry if I've misunderstood, but my fuel pump is NOT continually running. The place my car is currently at have pulled the spark plugs after repeat primings etc and apparently there isn't any signs of excess fuelThe thing is the OP's car was a runner and developed a fault that caused the fuel pump to keep running rather than for a 2 second or however long prime
Me bearing in mind that the OP didn't alter any wiring but did start (excuse the pun) switching the ignition on and off to check the priming function explained in another post what could cause the fuel pump to keep running, I also posted my thoughts on what may have been happening due to switching the ignition on and off up to 17 times (flooding came to mind)
The OP hasn't been back here to discuss all posters findings, at the bottom of this post are the good diagrams from the above link and my thoughts about the problem and causes
Penelope Stopit said:
"Poster Mignon" mentioned that the "17 times and out" could have nothing to do with the problem and is correct, Mignon having mentioned about "17 times and out" and myself having thought about it am now 100 % certain the "17 times and out" is nothing to do with the pump continuously running or the engine running rough
The fuel pump is NOT continually runningPenelope Stopit said:
The "17 times and out" mystery has very likely been there for much time and the OP has only noticed it recently due to another problem encouraging him/her to test it further by cycling the ignition on/off..............
This is incorrect. The 17 times cycling of the pump etc was not present before it started mid way through a trackday. I always let the fuel pump prime before starting the car so would have noticed if it kept priming. You hear all sorts of noises on the Atom due to it being "open" so could not have missed the fuel pump repeatedly priming. Also, it cuts out after 17 times and then won't runPenelope Stopit said:
You could replace the MFRU in an attempt to eliminate the fuel pump running on rather than priming, there is more of a chance that the fuel pump relay stuck closed rather than the ECU having a fault
The MFRU was replaced with a known working one whilst stuck at the trackday as per my opening post - it didn't make any differenceIf I can't get this resolved locally, which is looking highly likely, then I'll take it back to the factory
Cheers
Penelope Stopit said:
Ok I think I've got it now
Yes I have been wasting my time on this due to misunderstanding your opening post
Where you comment - "The fuel pump normally primes and then cuts out after 3 seconds or so, but for some reason it kept priming" I misunderstood this as meaning the pump kept running
Am I right in thinking you switched the ignition on and left it on and the fuel pump was auto switched on and off 17 times before stopping?
Do you understand how your opening post that mentions "kept priming" could be read as kept running?
You surely do mean that when switching the ignition on the priming stage is repeated automatically 17 times
No, I think I was quite specific. "Priming" is not the same as "running". Priming is normally a one time affair to pressurise the fuel lines etc. Once a car is started the fuel pump runs continuously. You've obviously misread my post but no harm done. Yes I have been wasting my time on this due to misunderstanding your opening post
Where you comment - "The fuel pump normally primes and then cuts out after 3 seconds or so, but for some reason it kept priming" I misunderstood this as meaning the pump kept running
Am I right in thinking you switched the ignition on and left it on and the fuel pump was auto switched on and off 17 times before stopping?
Do you understand how your opening post that mentions "kept priming" could be read as kept running?
You surely do mean that when switching the ignition on the priming stage is repeated automatically 17 times
Did you watch/listen to the video clip I linked to? It shows exactly what is happening - fuel pump priming 17 times and then cutting out
Toilet Duck said:
put some fuel in and then switched ignition on just to get a reading off the dash fuel guage. The fuel pump normally primes and then cuts out after 3 seconds or so, but for some reason it kept priming. Having investigated a bit further today, it primes 17 times and then stops. It does it 17 times every time you put the ignition on and repeat. If you try and start the car whilst it is doing this fuel pump priming/cycling it will start but idle very rough. After the 17 cycles of priming the engine dies. The engine will not start at all after this (the fuel pump is not running)
Here is a video I took today whilst investigating:
https://www.youtube.com/watch?v=5vwAjCGWtNc
Here is a video I took today whilst investigating:
https://www.youtube.com/watch?v=5vwAjCGWtNc
GreenV8S said:
Those replies from your mechanic all seem sensible.
One thing I'm uncertain about and don't remember whether you already told us this: will the engine actually start and run while the fuel pump is going through its cyclic priming? If there is any life from the engine I'd consider jerry-rigging a permanent supply to the pump so it runs continuously and see whether the engine then behaves as normal. If so, it suggests the ECU problem is specific to the fuel pump supply logic and not a more general failure.
Hiya One thing I'm uncertain about and don't remember whether you already told us this: will the engine actually start and run while the fuel pump is going through its cyclic priming? If there is any life from the engine I'd consider jerry-rigging a permanent supply to the pump so it runs continuously and see whether the engine then behaves as normal. If so, it suggests the ECU problem is specific to the fuel pump supply logic and not a more general failure.
Yes, when at the trackday and this fault first materialised, I tried starting the car during the fuel pump cycling and it did start. However, it idled very rough and wouldn't accelerate cleanly (gearbox was in neutral, I just reached in and blipped throttle with my hand). After 17 cycles the fuel pump stops.
During this time the throttle relay also cycles 17 times in sync. If pressing the throttle pedal during this cycling (and without engine running) I can hear the DBW butterfly valve opening, but it shuts every time the relay cycles.
jeremyc said:
It is possible that there is some fuel pressure sensor somewhere (in the fuel pump itself, in the fuel system) that feeds back a go/nogo to the ECU (either only when priming or also when running). If there is a loss of fuel pressure, which could be caused by a faulty FPR or a faulty pressure sensor, then the sequence could be explained:
1. Ignition on.
2. ECU instructs pump to start priming process (applies power to relays).
3. ECU doesn't see adequate fuel pressure (faulty FPR, sensor, pump, whatever) so turns off power to relays, restarts sequence for step 2.
4. When ECU has retried steps 2 to 3 seventeen times it gives up.
Try swapping out that red FPR and/or find out if there is some kind of fuel pressure sensing going on?
I rang the Ariel factory yesterday to ask whether there is a fuel pressure sensor in the pump, I was told that there definitely isn't. There are 4 wires on the connector going to the fuel pump, apparently they are just for the running of the pump (power?) and the others go to a fuel level sender that gives a reading on the dash. 1. Ignition on.
2. ECU instructs pump to start priming process (applies power to relays).
3. ECU doesn't see adequate fuel pressure (faulty FPR, sensor, pump, whatever) so turns off power to relays, restarts sequence for step 2.
4. When ECU has retried steps 2 to 3 seventeen times it gives up.
Try swapping out that red FPR and/or find out if there is some kind of fuel pressure sensing going on?
From what the current place has told me, as the fuel pressure regulator is purely mechanical it has no link to the electronics/ECU side of things. It doesn't appear that there is any way for the car to "measure" fuel pressure via a sensor. This seems very odd to me but I'm far from being qualified to say different. I can try swapping the red FPR when I get the car back, but so far have been advised it won't be causing this issue.
Cheers
jeremyc said:
It's not low on fuel is it, or a faulty fuel level sender?
Maybe the ECU won't allow it to start if there's no petrol in the tank.
If only! Maybe the ECU won't allow it to start if there's no petrol in the tank.
The level sender purely talks to the dash and not the ECU. It's notorious amongst Atom owners as being crap, it fluctuates so much its unbelievable, especially on track. What I can say is that if the fuel level does drop too low on track, on extended left handed bends you can get fuel starvation as the fuel sloshes to the right. However, there is plenty of fuel in my tank
GreenV8S said:
Another question you may already have answered - what is the throttle relay for? Is it something that would normally be powered up from the engine run signal, presumably together with the fuel pump and lambda sensor heaters?
I'm afraid I do not know the answer to this, google let me downeliot said:
What if you disconnect the wire (or fuse if it's after the relay) to the actual pump?
Does the relay(s) act the same or do nothing?
I don't know. However, with the connector block going to the fuel pump disconnected (which I believe has the same effect as what you have suggested?), the relay(s) within the MFRU and throttle sensor relay both continue to click/cycle 17 timesDoes the relay(s) act the same or do nothing?
eliot said:
No experience of this ECU, but with megasquirt the ECU will fire the injectors during the pump priming phase as well. You could have a faulty low impedance injector that could also be tripping the ECU. Prove by removing all the injector connectors. (may also explain bad running when it does run)
Before taking my car to the garage it's currently at, I tried unplugging all the fuel injectors from the loom and it didn't make any difference to the fuel pump cycling etceliot said:
Also be wary of repeatedly cycling the ignition - you could be slowly filling the cylinders and the block with fuel, watering down your oil - when you finally fix it you can expect trashed big ends. Check oil level and smell it too.
Unless an injector is leaking/faulty, would it actually squirt any fuel when the fuel pump is just priming and the engine not actually running? I will be changing all fluids anyway once it's fixed. As mentioned before, the spark plugs (when removed by the garage after fuel pump priming etc) apparently did not show signs of being "wet" with fuel.Cheers
Penelope Stopit said:
Ok my speakers are back on and I've viewed your video
Some of what I'm posting here may have already been advised but I don't think so, the topic is getting too big for me to read all the way through right now
I'm posting the following assuming that the pin numbers in the above image are for your cars MFRU and ECU's pin-outs
The first thing to do is connect a multimeter between the battery positive (not any positive....it must be the battery positive) and the cable going to or the terminal in the MFRU 6 way plug (6/3) White/Pink or Pink/White, the MFRU must be fully connected
Is that ECU Switched Negative switching on and off 17 times?
Many thanks, I will check this when I get my car back next weekSome of what I'm posting here may have already been advised but I don't think so, the topic is getting too big for me to read all the way through right now
I'm posting the following assuming that the pin numbers in the above image are for your cars MFRU and ECU's pin-outs
The first thing to do is connect a multimeter between the battery positive (not any positive....it must be the battery positive) and the cable going to or the terminal in the MFRU 6 way plug (6/3) White/Pink or Pink/White, the MFRU must be fully connected
Is that ECU Switched Negative switching on and off 17 times?
Penelope Stopit said:
I could do with a diagram. What Honda model/year car is the engine fitted to, is it K20 or K20a or??? Is the ECU Standard to look at for correct pin-outs???
It's a K20z. I understand the ECU is standard but with the Hondata add on to allow reflashing/changing parameters etc, and therefore the "Atom specific map."It would seem that the ECU is supposed to just prime the fuel pump for a set period as opposed to continuously/repeatedly or until another signal is received:
Just for clarification, that screenshot is courtesy of another Atom owner with a Hondata Flashpro adapter. That plugs into the ECU and talks to it via the Hondata add on installed on the ECU board. You need to purchase the Flashpro in order to map the ECU or to make any adjustments etc. I don't currently have one although I will be purchasing one when my car is ready to be remapped. The Flashpro is "locked" to the ECU it's plugged into so won't work with others. I also believe that once it's locked/plugged in then the vehicle has to be dynoed in order to give it the required settings
Thank you very much for the continued help, really appreciated
jeremyc said:
here are some interesting sensors in that list on the left that might be worth learning more about: Fuel Status, PTANK (which is a pressure reading).
It might also be instructive to understand what data is available on the 'Analog Inputs', 'Digital Inputs' and 'Multiplexer/Digital Output' tabs, as well as the 'Onboard Datalogging' of course.
I suspect a lot of the fields are "generic" and do not necessarily mean that a particular sensor/input etc is therefore present/available, but don't know for sure. The other thing is I don't own a Flashpro to connect to the Hondata ECU so I can't check myself. I plan to buy one, but don't want to connect/lock it to my ECU as that will cause it to loose its existing map etc and require dynoing. My plan was to do some other changes before getting it dynoed/mapped, but this current fault has thrown a spanner in the works.It might also be instructive to understand what data is available on the 'Analog Inputs', 'Digital Inputs' and 'Multiplexer/Digital Output' tabs, as well as the 'Onboard Datalogging' of course.
Kccv23highliftcam said:
Ok. As I intimated earlier the fault finding needs to back to stage 1 and thoroughly recorded step by step.
There are differences in relay panel and harness that need to be physically discounted for a start. Once this is done then the relay panel has to be checked for both continuity and insulation resistance.
This problem will only be solved by taking a logical approach and carefully eliminating items piece by piece.
It is disappointing that you appear to have been left holding the baby by the ecu vendor and vehicle manufacturer.
I agree, i think this probably needs to be looked at from the beginning againThere are differences in relay panel and harness that need to be physically discounted for a start. Once this is done then the relay panel has to be checked for both continuity and insulation resistance.
This problem will only be solved by taking a logical approach and carefully eliminating items piece by piece.
It is disappointing that you appear to have been left holding the baby by the ecu vendor and vehicle manufacturer.
Kccv23highliftcam said:
Btw has anyone checked from the ignition switch down to the relay box (power supplies) ?
Not that I'm aware from (I haven't)eliot said:
And yes, taking the live for the o2 heater from the ignition seems an odd choice instead of the main fused 12v supply. Again easy to eliminate; pull the o2 relay out.
Will try disconnecting the O2 sensor once I have the car back. I can't pull the relay as its soldered inside the MFRUPenelope Stopit said:
FAO Toilet Duck
Many thanks for all the info/pics etc, I will have to wait until I get my car back to check this out Edited by Toilet Duck on Friday 2nd November 12:33
Holy s
t, issue has been found! 
Another Atom owner got in touch with me to say they had experienced a similar issue, ironically at a track day. They told me to check the wiring to the lambda sensor in the exhaust as it passes extremely close to the exhaust manifold, theirs had melted and the wires shorted. As discovered by posts within this thread one of the relays in the MFRU is for the lambda/O2 sensor so therefore linked to the fuel pump etc via the MFRU circuitry. Contacted the garage my car is at with this suggestion and after checking they confirmed that the wires are melted together. Cut back and removed and car now apparently starts and runs so just needs new section of loom making up and this time rerouting away from the exhaust unlike the OE wiring route.
Absolutely over the moon that the issue has been found. Many thanks to everyone who posted and for all the help offered, much appreciated
t, issue has been found! Another Atom owner got in touch with me to say they had experienced a similar issue, ironically at a track day. They told me to check the wiring to the lambda sensor in the exhaust as it passes extremely close to the exhaust manifold, theirs had melted and the wires shorted. As discovered by posts within this thread one of the relays in the MFRU is for the lambda/O2 sensor so therefore linked to the fuel pump etc via the MFRU circuitry. Contacted the garage my car is at with this suggestion and after checking they confirmed that the wires are melted together. Cut back and removed and car now apparently starts and runs so just needs new section of loom making up and this time rerouting away from the exhaust unlike the OE wiring route.
Absolutely over the moon that the issue has been found. Many thanks to everyone who posted and for all the help offered, much appreciated
Penelope Stopit said:
eliot said:
Yeh - I’ve got good electronic test equipment including a digital scope - i wanted to look at the feed into the ecu for sagging/spikes. It was your drawings that led me in that direction - so props for looking into it and posting it up.
I hope Toilet Duck presents you a Blue Peter Badge for your kind thoughts of making a journey to his part of the world to fix the faultI don't own a scope. You, your knowledge and scope would have no doubt soon found the fault
Some very generous people on here, really grateful for all the invaluable help Penelope Stopit said:
All of what I post above is assuming that the MFRU hasn't been modified and we both know what assuming can cause don't we. It would have been good if Toilet Duck had noticed if the MFRU had been modified in any way but I doubt we will ever find out as he/she is very likely screaming the Atom around for the foreseeable future
I haven't got the Atom back yet, I'm letting them redo the burnt bit of loom as I just want a fully working car when I pick it up. When I get it back, I will double check the MRFU internals but I am 99.99% sure it was unmodified. I went over all the pads on the back of the PCB with a soldering iron and re-blobbed them in case of a dry joint (obviously a pointless exercise knowing what we do now). and it all looked "original" to me Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff