Jump to content

Returnless fuel system pressure at different BAPs


Recommended Posts

Hi all,

I would like to understand if the fuel pressure behavior in my NB1 Mazda MX-5 is normal. This generation MX-5 has a returnless fuel system with the regulator in the tank and a target pressure of 400kPa/4bar. I have a fuel pressure sensor in the rail and the ECU uses it for compensations. It has a new fuel filter, and retrofitted OEM pulsation dampers to the fuel rail. It is the same as stock except for 550cc injectors and a different fuel rail.

To my curiosity, the FP sensor has never registered 400kPa at idle, cruise or WOT. Even on the previous ECU it was always lower than 400kPa, but it seems to follow somewhat follow BAP changes. At the racetrack BAP is ~83kPa, and FP was around 377kPa at WOT. Through the mountains around 73kPa BAP, FP was 371kPa at WOT. The data makes it seem like the FP is related to barometric pressures in this setup. Is this normal since the regulator is in the tank and referenced to BAP, not MAP?

I'm using the following sensor, and the linear calibration provided by the company: https://lowdoller-motorsports.com/collections/pressure-sensors/products/0-100-psi-5v-pressure-sensor

Another related question. When using Fuel System Type = FP Sensor, does the ECU need a 3D table for injector deadtime referencing Differential Fuel Pressure, or a 2D table with voltage only? Currently I have it setup as a 3D table with differential pressure and battery voltage because that's how I interpreted the help file, but I'm not clear if it is performing double compensations as a result.

Thank you in advance,

Ricardo

There are two logs in this directory to show what I mean: 

"ECU Log 2023-04-29 7;47;26 pm_TA2" is from trackday and "ECU Log 2023-04-28 7;42;36 pm_MountainScoot" is a drive through the mountains.

https://1drv.ms/f/s!Amu_1gWOWFlvmzvylI-WYSu8K6wN?e=rlfFfh

Link to comment
Share on other sites

I can't see it in the advertised specs, but you'll probably find that's an Absolute pressure sensor.

So in your example at 83kPa BAP if you're reading 377kPa "Absolute", it's actually 294kPa "Gauge" pressure.
And 371kPa Absolute at 73kPa BAP is 298kPa Gauge.

Which means you're regulating at around 300kPa Gauge pressure, which is 43.5psi - a very common fuel pressure.
My guess is if the fuel pressure regulator is stock and the spec is "400kPa", then they are assuming you'd be at sea level (101kPa) when checking the pressure. So 299+101 = 400.

Link to comment
Share on other sites

In a fuel press regulator you have the spring and atmosphere on one side of the diaphragm balancing fuel pressure on the other side.  Therefore the fuel pressure should always be the same "gauge pressure" above the reference (BAP) - it shouldn't vary with BAP, it should always be say 400Kpa above BAP.  It works the same way with a manifold pressure reference also - FP should always be the same gauge pressure above the reference pressure (MGP). 

For your deadtime question - yes you should use differential press on one axis if you have that info.  

 

 

Link to comment
Share on other sites

15 hours ago, DerekAE86 said:

I can't see it in the advertised specs, but you'll probably find that's an Absolute pressure sensor.

So in your example at 83kPa BAP if you're reading 377kPa "Absolute", it's actually 294kPa "Gauge" pressure.
And 371kPa Absolute at 73kPa BAP is 298kPa Gauge.

Which means you're regulating at around 300kPa Gauge pressure, which is 43.5psi - a very common fuel pressure.
My guess is if the fuel pressure regulator is stock and the spec is "400kPa", then they are assuming you'd be at sea level (101kPa) when checking the pressure. So 299+101 = 400.

This is incorrect according to the instructions for the sensor he has.  It's listed as a gauge type sensor not absolute.  It should always be reading differential vs BAP, though changes in pressure inside the tank can also affect the pressure regulators control pressure.

Link to comment
Share on other sites

Thank you all for the replies.

@koracing is correct, this is a gauge pressure sensor. I verified with Lowdoller when I purchased it last year because I was confused as well. @koracing, which instructions did you find for it? Even the specsheet they sent me with it didn't include this tidbit, I asked over the phone. It's good to have double validation here though.

If I'm understanding correctly, something is slightly amiss with the fuel system then. The only other modification I can think of that may affect the pressure in the tank is that I capped off the purge line out of the tank. I'll remove the cap and retest later in the week. I suppose other than that, maybe a weak pump or failing regulator? I'm not excited to do either.

15 hours ago, Adamw said:

For your deadtime question - yes you should use differential press on one axis if you have that info.  

Awesome. One less thing to stress over.

Link to comment
Share on other sites

I wouldn't worry about 370kpa too much as it's well over what you should need and you do have FP compensation anyway.  Maybe set up an alert of some kind if fuel pressure drops below a particular threshold.

I think the instructions were on their site and they listed some part numbers as "MAP sensors" and some as "Standard Pressure sensors" - but the calibration scales told me they are gauge vs. absolute when comparing them.  Their documentation could be better.

Link to comment
Share on other sites

  • 1 month later...

I think I need to sweat this a little bit. In a recent trip through various Colorado mountain passes I ran into fuel pressure problems. The fuel got very hot and it was coming out of the stock EVAP line in the engine bay that I had then uncapped. As a result my fuel pressure started fluctuating and the ECU overcompensated and the car ran rich, which is good. Opening the fuel cap released the pressure and I got to a hardware store, got a cork to cap the EVAP line, and continued my trip and racing with no further issues. After doing major elevation gains I would stop and depressurize the tank, so not ideal. So I guess I'm dealing with a fuel vaporization problem but I have not clue where to start diagnosing. Could someone give me some pointers on this one? I still have the stock canister in the rear of the car, under the tank, but no purge valve. I see Link can do basic control of one so I could find a way to retrofit it but I'm not sure this is the issue.

Link to comment
Share on other sites

@koracing yes it’s definitely getting pressurized. When I open the cap after going higher elevation there is at least a couple of PSI in there.

 

I’ve ordered another vacuum manifold to use for emissions, PCV and EVAP for now. Should be here this week.

@curly I saw your response about AN Volt 8 but it’s on a separate thread. I configured the input and it has a reading! I have no idea what the correct values would be so I scaled the 5 volt value until it read BAP (with the cap off and poking the breather). Apparently 155kPa at 5volts? I’ll drive the car tonight and see what the readings say.

Link to comment
Share on other sites

It seems the fuel tank pressure increasing correlates to how hard I drive the car and not so much with altitude, validating what curly said about my exhaust possibly heating the tank and no proper venting in place? It's possible there is an exhaust leak back there, and the muffler is slightly higher than stock and could be rubbing a bit.

The purge control logic through a GP output seems to work as expected, but I don't have it plumbed into the manifold yet. I hope this helps the situation once that extra manifold arrives and a vacuum gets pulled on the EVAP line. Does this seem sound to you @koracing @curly?

Log: https://1drv.ms/u/s!Amu_1gWOWFlvoAzlo9PnN718rn64?e=h0dN8D

LqdopAL.png

lsfCG9A.png

Link to comment
Share on other sites

If there are any two bolt flanges back there, put your hand around them while the cars running, you'll feel a puff of exhaust pretty easily. My experience is with enduro cars, and since a lot of miata exhausts seem to have a two bolt flange back there, I've seen it a couple times when the gasket blows out only on the top, heating up the fuel. I saw a 10 foot geyser of fuel during a pit stop once, that was fun.

Otherwise, check the top of the tank for a check valve like the NAs have, forget if the NBs do too.

 

 

Link to comment
Share on other sites

I'll get under the car on Friday and look for leaks.

I've been going through the service manual and many forum posts to understand how this EVAP is supposed to work together, but I'm seeing conflicting information and could use a sanity check from more knowledgeable folk. Turns out the manual also has expected values to voltages for the Fuel Tank Pressure sensor and Fuel Level, which is nice, and my previous scale was very wrong. The NB1s have a more complicated system than NAs with 2 valves in the rear (canister drain control valve and tank pressure control valve), a rollover valve, and the purge valve in the bay. The service manual also indicates that the purge valve is supposed to work on PWM, not just on/off. Aux 12 is not selectable for PWM, will it be ok if the logic is on/off?

My main doubt is whether the CDCV should be open or closed when the purge valve is active, and is it ok for the ECU not to control the pressure control valve?

8aMvysE.png

UbSCBRA.png

BzD3ZyE.png

qCEOj8U.png

Link to comment
Share on other sites

@curlyis going to be the miata expert and works at a shop that primary does track vehicles (and tracks his own miata as well as does track side support for many others including a fair amout of bmws) so I would yield to his knowledge on that subject.  Before I moved to Washington in 2021 I worked with his shop pretty regularly as they used my dyno almost weekly it seemed.  There is probably a way to do some really low frequency PWM via on/off delay on the aux 12 output with a math block - as I've used the 11-16 outputs for check engine lights and I'm sure there's a threshold of frequency that you can turn them on/off.  Otherwise repinning to a PWM capable output (maybe swapping Aux12 pin 21/3L purge pin with ignition 5 pin 20/3N that's for the fuel pump) or using expansion (ign 3/4 or Aux 9/10) for it may be an option.  

Link to comment
Share on other sites

Oof, expert in race miatas, yes, expert in NB evap systems, probably not. If it were at my shop, it'd have a 16 gallon cell and all the related vents :D. Make sure you can hear both the front purge valve and TPVC clicking in PWM test mode, and then my first suggestion would be to leave both off, wait for fuel tank pressure to increase, then use on/off test mode on each valve individually to see which one results in a rapid depressurization of the tank. In my mind, one or both should. Both may need to be open at the same time if I'm reading the vacuum diagram correctly.

Link to comment
Share on other sites

Thank you both very much. I got to drive the car on a 170 mile trip with a mocked up EVAP system plumbed in. Stock solenoid and overflow canister connected to the vacuum manifold I use for the brake booster, MAP, and IACV.

In this trip I found the EVAP made the car pig rich anytime it opened, especially after a big elevation climb where I leave the car at 4k-6kRPM for a while. Specific example is the hill up to town where I go from 5600ft to 7200ft. After this climb, Purge active made the car 15-20% richer, then CLL made the Lambda oscillate a lot which affected drivability. I also found I had to close the purge valve on overrun in addition to the other conditions (at idle, near BAP, engine cold). That said, the tank was nowhere near as pressurized as other times. More testing needed, especially on a hot trackday.

It seems PWMing the purge valve is necessary to control how rich EVAP makes the engine. It really caught me off-guard how much of a difference this made at part throttle. @koracing I've tried to find a way to use a math channel as you suggested, but I can't think of it. Could you give me an example? Otherwise I'll plan to swap the wire to a PWM capable output.

@curly That counts as an expert to me! So yes, I can hear the ECU clicking the purge and the CDCV. Link doesn't have an output for the TPCV, but from what I've gathered that one is a last resort vent valve when pressure goes too high. The stock ECU closes the CDCV when it checks for leaks in the EVAP system. The CDCV is closed, a small vacuum is pulled with the purge valve, and the ECU checks the output of the tank pressure sensor. If it's out of spec, a CEL is thrown (small EVAP leak, or large EVAP leak, depends). So the CDCV is normally open until the ECU actuates it as you pointed out, so I left it open in my testing above. Tank pressure stayed at 0kPa using the scaling I found in the service manual, so it most be working?

Here is the logic I'm using for the Purge Control at the moment.

KrFnTh3.png

Link to comment
Share on other sites

Have you looked into the exhaust to make sure there aren't any leaks back there? How long is the valve open for after tank pressure returns to zero? I'm not immediately seeing a solution, but maybe there's a way to only open the valve for a few seconds once your above conditions are met.

Link to comment
Share on other sites

I've been busy with other projects and haven't checked the exhaust for leaks yet, but I will tomorrow. Hadn't even looked at the car until today. I have to get it ready for Saturday trackday and exhaust leaks are in the To-Do list.

7 hours ago, curly said:

How long is the valve open for after tank pressure returns to zero?

Do you mean in the stock ECU? If so I have no idea. I haven't found much about the stock PWM logic other than it does PWM during cruise states. As for the logic I was using above, the valve was open anytime I was cruising, regardless of tank pressure. I'm not seeing the tank pressure move at all, so I'm starting to second guess the scaling I found in the manual.

Just today I figured out a way to PWM the valve using a math channel. After driving the car around a bit the richness oscillation is gone. The car still runs ~9% richer than it used to when the purge is active but CLL took care of it. The formula I'm using is "sin(a*10.3)-0.5", where "a" is cnt(a), and cnt(a) is counting when the "Cruise" GP Output turns on (RPM>2400, TPS>2%, MGP< -10kPa).

Link to comment
Share on other sites

On 7/3/2023 at 2:46 PM, curly said:

Have you looked into the exhaust to make sure there aren't any leaks back there? How long is the valve open for after tank pressure returns to zero? I'm not immediately seeing a solution, but maybe there's a way to only open the valve for a few seconds once your above conditions are met.

I did have an exhaust leak exactly as you described. I couldn't source a new gasket before the trackday so I grabbed a generic gasket that fit and made it work. No leak for the weekend.

So the weekend's results. The tank pressure showed 1kPa after my hotlaps in 100+F ambient temps, so good result by fixing the exhaust leak. Gaining elevation does increase the tank pressure to about 3-4kPa with my current purge control. I used a 4D fuel correction to compensate for the richness caused by the purge. After ~300 miles of mountainous driving I'm satisfied with the results. I can smell fuel as I gain big elevation but goes away after some cruising, so maybe I just need to make the purge more aggressive and should check that the carbon canister is still in good condition. I will also move the purge wire to a PWMable pin later, I just wanted to test this out quickly. I also played with the CDCV after hard driving/tank hot. Closing the valve increased pressure, opening made it stop increasing. I haven't had a chance to play with closing it as the purge is active, but I imagine that's not a wanted state?

To recap: I had an exhaust leak making the tank hot. Elevation gain does increase my tank pressure. Used Mazda's Tank Pressure sensor calibration and seems to be accurate (was using the wrong calibration earlier). Purge control is working, makes the car rich as it activates.

If anyone has any further feedback I would love to hear it and learn more, but I think I have resolved my issue for now.

Here are some pictures of my current settings for anyone that may want/need to setup a similar control:

j2A1k83.png

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...