Diagnosing loss of fuel pressure

[magpie]

First off many thanks to Dasing (forum member) and Cameron (Racepak) for showing that the Datalink II software is actually useful!

Apologies for the 'big' images, however needed to display the data...

History.
At a recent track day at Lakeside (Queensland, Australia) the car ran in four sessions. However by session 3 the car would not accelerate as it previously was and by session 4 it actually felt like the car was having a soft cut. The dash (IQ3) was giving no warnings (oil/water temp, oil pressure, low fuel, high/low EGT). Fuel surge has never been an issue in the car as it has been run in the past down to below 1ltr, during all the 4 sessions fuel level was never below 10.

FP by Eipeip, on Flickr

It took me a while to look at the fuel pressure trace as the source of the problem, so I took the opportunity to add to my 'Car Health' tab. This showed that he fuel pressure issue was not related to G forces nor fuel level or below 5,750 RPM. EGT 1 above 5,750 showed that when compared to the average EGT it was up to 30°c hotter than the other 4, however in general No1 has always been a little warmer than the other 3. Interestingly the shape of each EGT is almost identical.

Car Health by Eipeip, on Flickr

Actions:

1. Swap out fuel pump for new one - done. The current one is a bout 3 years old
   DSCN4023 by Eipeip, on Flickr[/*:m:1bl6wbdz]
2. Change fuel filter - done. Current one was about 18 months old[/*:m:1bl6wbdz]
3. Clean injectors - having done this week
   DSCN4052 by Eipeip, on Flickr
   
   DSCN4051 by Eipeip, on Flickr[/*:m:1bl6wbdz]
4. Add low fuel pressure warning to dash & ECU - not done, will do after injectors cleaned[/*:m:1bl6wbdz]

Whilst the Datalink II could make customising graphs/reports less time consuming, it does allow data analysis. It would be great to have 3 way data/scatter plots but I'm getting used to it (with the help of Dasing).

[dasing]

Re: Diagnosing loss of fuel pressure

Very nice. Congrats on using some of your new capabilities in the software for understanding an actual problem. That's great!

[magpie]

Re: Diagnosing loss of fuel pressure

THANKS Dasing! Gets better, looking at 2016 data it would appear that the 'issue' happened between prior to 10/16 (WTAC). Since then the problem has got worse, however it starts off 'normal' then gets worse. This would indicate a 'time' based issue like heat soak. It could be the pump becoming less efficient or maybe even a loss of voltage due to incorrect wire sizing to the pump.

FP 2016 by Eipeip, on Flickr

[magpie]

Re: Diagnosing loss of fuel pressure

Bad news, still had fuel pressure dropping :(

Time to change the fuel pressure regulator and/or look at a surge tank. Looking at something like this http://www.radiumauto.com/FST-R-Fuel...-FPR-P348.aspx

Again the issue seems to be time related...
20170218 QR FP by Eipeip, on Flickr

[Blue]

Re: Diagnosing loss of fuel pressure

While I don't have the Racepak dash any more, mine did provide some similar insight. Mine was not on a scatter plot but rather the normal trace, it was immediately obvious that my oil pressure was collapsing under heavy braking. A bit of extra oil in the system prevented it and the data showed that too.

Your plot does just look like a fuel pump that is not up to the task, seems to pretty significantly start to drop pressure around 5000 RPM and by 6000 or so it would be enough to lean mixtures off. Do you have a wideband O2 sensor in the car? I alarmed based on AFR over a certain manifold pressure which sent me off to look on another page at my fuel pressure and manifold pressure together to work out real fuel pressure. It was harder to look at as my old ECU sent raw gauge pressure for fuel rather than the more useful differential pressure calculation. If you have the diff fuel pressure sent you can alarm on it easily too. It is probably easier on an atmo engine too as you don't have such large fuel pressure swings to deal with. As gauge PSI on the rail mine runs anywhere between about 34 and 62 PSI depending on throttle opening and boost pressure! Differential pressure should only vary by a couple of PSI through the whole operating range.

[magpie]

Re: Diagnosing loss of fuel pressure

Thanks for the feedback Blue.

Whilst my car does have an AFR sensor it is not used for the tune and regularly stops working :) The car does have 4 X exhaust gas sensors, these can be used to monitor what is happening if the AFR sensor fails. I'll be replacing the AFR sensor as part of the problem solving.

It appears to be time related and after a certain amount of time the pressure drops off as RPM increases.

Since the fuel pump/filter have been replaced and the injectors cleaned there are 2 possibilities. 1 fuel pump voltage drop or fuel pressure regulator fault.

I have checked the voltage drop at the pump and it is fine, however difficult to test under track conditions. I may wire something in to measure/log current draw for future reference.

Fuel pressure regulator - the current one is a Tomei which is not serviceable so no way to 'test' it. Since the fuel system is a return type and the FPR has 1/4" in/out whereas the rest of the system is AN6 size the fuel (E85) could be heated on its return to the tank and hence heat up the rest of the fuel. The heat may be enough to cause a fault in the FPR... Long shot....

Plan

1. install surge tank (Radium with built in FPR)[/*:m:2e1ytwh9]
2. change fuel system to dead end[/*:m:2e1ytwh9]
3. wire in surge tank (via relay) to battery with trigger from tank pump. This will be a short wire run as the battery is in the boot[/*:m:2e1ytwh9]

If this does not work, then it could be tune related.

[Blue]

Re: Diagnosing loss of fuel pressure

it is not an older Innovate unit with an LSU4.2 sensor is it? I have one of them and firstly the LSU4.2 requires regular free air calibration to stay accurate, and secondly in my car (A turbo engine) it was prone to overheating the sensor on track and shutting itself down.

The Haltech CAN wideband controllers also currently use LSU4.2 sensors and I could not get any answer if they are updating to the better LSU4.9 in the future. I am going a different way with my ECU as a result (And because of other issues I am seeing and do not like) I am probably going to put an Emtron ECU in my car and that has LSU4.9 wideband control on board (Two if them in fact but one would be unused in my car)

Any that high fuel temps cause it to fall over? My car used to get the fuel so hot at events in the (Victorian) summer that I had to fit a fuel cooler to it. That went on the car before I had a data logger but it could get so hot that the pump would become noisy.

[magpie]

Re: Diagnosing loss of fuel pressure

Again thanks Bruce.

I will need to check the Innovate sensor, however I suspect it is a LSU4.2. Strange I was looking at the Haltech CAN wideband today and did not notice the sensor used.

I had my first ever run of Sandown in January this year in a Victorian 'summer'.

[Blue]

Re: Diagnosing loss of fuel pressure

Was that a Tampered motorsport day? If I had realised you were down I might have been able to come for a look. I live about an hour and a half (On a good day) away from Sandown. A little off the Hume Freeway, you probably drove within 10KM of my place.

Depending on the age of your innovate unit, later ones than mine can use a 4.2 or 4.9 sensor, but mine is the best part of ten years old.

[magpie]

Re: Diagnosing loss of fuel pressure

I was down for the Mazda MX5 Fan Fest on 21/01/17. Manged to get almost 700 MX5s on the Sandown track.

Towed the MX5 down for the event, 500ltrs of diesel for 50ltrs of E85...

The Innovate unit is about 5 years old. So yes more than likely due for replacement. Will calibrate the sensor on the weekend and see how it goes.

[Blue]

Re: Diagnosing loss of fuel pressure

It would be very handy of Haltech would update to the LSU4.9 sensor type. The advantage of the CAN based Haltech one (Apart from having Yet Another Box to find a space for) is that when you use the Haltech CAN one the ECU can see the status of the sensor in the CAN messages and ignore it if it is not ready or indicating a fault.

[magpie]

Re: Diagnosing loss of fuel pressure

Installed the Radium Surge tank before I had to fly to New Zealand for work. Pump not wired in yet but the tank pump works and there are no leaks...

Bracket for the surge tank. Where the jack used to go. I may add a brace as whilst it feels solid who knows at track speeds.
DSCN4096 by Eipeip, on Flickr

Bracket in place
DSCN4097 by Eipeip, on Flickr

Surge tank mounted
DSCN4098 by Eipeip, on Flickr

DSCN4099 by Eipeip, on Flickr

in line 100 micron filter to replace the OEM. This goes to the surge tank
DSCN4100 by Eipeip, on Flickr

Bulkhead connectors installed
DSCN4101 by Eipeip, on Flickr

Hoses nipped up. Need to purchase some hose seperators and clamps to keep the hoses from rubbing on anything.
DSCN4107 by Eipeip, on Flickr

DSCN4108 by Eipeip, on Flickr[/quote]

[magpie]

Re: Diagnosing loss of fuel pressure

Finally the issue has been 'fixed'. Turns out fixed one problem to find another that needed to be fixed...

Due to heat soak issues the surge tank was moved:
DSCN4143 by Eipeip, on Flickr

The surge tank fixed the fuel delivery issue, however the car then was reporting erratic fuel pressure. With the car's ECU sending the car into limp mode if fuel pressure dropped outside pre-set limits it was frustrating to say the least...

As part of the surge tank install the fuel sensor was moved to the end of the fuel rail, however this was corrected and moved off the engine.
DSCN4285 by Eipeip, on Flickr

When this job was done, the reported fuel pressure was reporting low again... Jiggled the plug and suprise, suprise, suprise normal pressure. Turns out that the ground pin on the plug was not crimped properly and all the playing around with the plug caused a dodgy connection.


Fuel pressure before sensor wiring fix. The 'noise' on the channel should have been obvious, but for some unknown reason I did not think to check the sensor wiring:
20170429 Fuel_Presure by Eipeip, on Flickr

After sensor fix
20170601 Fuel_Presure by Eipeip, on Flickr

When you look at when the problem was first found, the fuel pressure trace is not 'noisy' and there was an obvious power loss. The 'dodgy' wiring was different as there was no loss of power and was intermittent...
Car Health by Eipeip, on Flickr