EFI Fuel Pump Failure Risk

[airguy]

Let's stick to facts. You're reporting your system goes rich with two pumps running. So, detail it. What size are your return lines, and are you using the SDS pump module or individual wing root pumps?

It does go rich(er), slightly, with two pumps running, not nearly enough change to worry about and you wouldn't notice it without an oxygen sensor. I'll put a $100 bill right now that says System32 does too, if properly instrumented to see it and you're honest about the results. You're going to be hard pressed to find/install a backpressure regulator that is so well controlled for the fuel flow range from 1 to 30 gph that you don't see any rise (or drop) for that entire range.

EFI (either one), is obviously the absolute worst way possible to run your engine - except for all the other ways.

I am running the SDS module with Andair full duplex valve, #6 lines in and out, 100 micron prefilter and 40 micron postfilter.

You've made it obvious you are not a fan of the EFI camp - why do you feel you need to **** on people using it? Is 1950's carb/mag tech so wonderful that we should stop innovating? What carb/mag setup do you have in your current road vehicle, so we can compare?

 

[Toobuilder]

Close to 100 posts about one pump vs two, and finally we hear the fuel rail won't maintain set pressure with two pumps running?

Depends. If one is really leaning on the bypass orifice because the engine does not consume much fuel (think, 160 hp), then its reasonable to view that as a fixed restriction. My 540 does not "bypass" much fuel in comparison because a lot more of it is going through the injectors. I get no perceptable pressure rise on TO between 1 pump or 2. And on approach (or even idle) I have not noticed any perceptable change in the engine behavior.

...Any notion about the perfect accuracy of EFI just got shot in the ***.

Was this a real notion in your head? You do understand the significance of "open loop", right?

 

[dmattmul]

Data if accurate and understood worth it:

Greg's data is always welcome and yes, our EFI systems are not perfect air/fuel delivery systems. We need to understand the design, data, risk/rewards and decide our own path. Unfortunately, the EFI run down the opponent doesn't seem to stop at the top. (Wish it wouldn't start at all) I think both systems are better than the alternatives out there and tell people that. What I hate to rehear is they ask me what about this or that incident and do you feel safe flying a EFI system? (Even most panel builders refuse to build EFI systems now) I try to explain that pretty much every single issue I've read about comes about a builder shooting themselves in the foot. Unfortunately, that probably will never stop.

I debated about posting the data but after talking with another builder who has performed similar testing and similar results thought I would. Please if you doubt the data perform your own testing. Manometers are very cheap on Amazon and although not calibrated to ASTM standards provide a decent reference point.

One pump or two:

My flow data for both SDS and EFI running one pump at a time where ~8% lower than the Walbro Performance Charts at 40 psi and 13.5 volts. I placed the pumps 3 inches above the height of my fuel reservoir which I estimated the pumps are located in my 14 tunnel when down to 5 gals. I'm guessing Walbro provided a flooded suction when they ran their tests. The pressure drop located on the suction side of the pump was right at 1 psi. I did not add a fuel selector valve just one container and return line.

My flow data for 2 pumps was pretty much exactly twice what one pump delivered both SDS and EFII. My fuel pressure regulator went up about 1.5 psi using 2 pumps both systems. Because of the added flow rate pressure drop was a little over 2 psi suction side again both systems. This will cause a richer mixture (because of the slightly higher FP) but on take-off probably a good thing for a minute or 2.

I'm not at my shop that has a bunch of my previously purchased billet filters but the one that I had (Holley 100-micron stainless screen) produced 0.12 pressure drop 1 pump and about twice that with 2 pumps.

I went out on a limb and connected 4 hard AN6 90's together and only with 4 could I even get a discernable reading and it was very small. (less than 0.04 psi for 4 fittings)

I wish I could have come up with a method to measure cavitation but judging by the "noise" running 2 pumps seemed to be a lot more. When I connected the filter plus the 4 90's running 2 sounded quite a bit noisier. (More "rattlier") This does not make it wrong as there are some distinct advantages running 2 pumps, especially on take-off. The positive outweighs the negative in my mind if you don't have a system that provides automatic reliable switch-over. One side effect for the pressure regulator I use is the metal-to-metal seat. Cavitation could quickly damage the surfaces as the diaphragm will want to bounce possibly bring the two surfaces together causing erosion. This is not just theoretically possible but actual performance in the field. (I have pictures to document it, a fitting that was not perfectly tight on the suction side allowed air to enter the piping and after about 100 hrs. damaged the seat) When this occurs the MP fuel boost does not perform 1:1 as designed. Since the fitting is above the level in the tanks no fuel leaked so zero smell and no blue coloring or it least if the level was higher not enough pressure to induce a leak)

I am ok discussing the downside of our system. Every system has them just for me the positives outweigh the negatives and hopefully my grandkids will be comfortable flying behind either EFI system. VAF is a great place to share best practices, hopefully we can continue.

 

[DanH]

I am running the SDS module with Andair full duplex valve, #6 lines in and out, 100 micron prefilter and 40 micron postfilter.

Thank you. I think we can call that setup typical.

You've made it obvious you are not a fan of the EFI camp - why do you feel you need to **** on people using it?

Greg, take a deep breath, because you're way off base.

I was running a Honda store in 1985 when the first EFI Accords came off the truck. I managed used car recon throughout the entire US fleet transition from carbs, so I'm well aware of EFI superiority, in hard dollar terms. I bought an SDS system from Ross more than 20 years ago (to fly on a Subaru, BTW). And I'm installing an EM-6 right now.

Here's the deal...I don't play favorites. I don't live in your camp, or their camp, or anybody's camp. I'm just interesting in the facts. Here we went through 100 posts to learn there is a rail pressure difference when running one pump or two. Thank you for casting light on the subject.

Next question (noting Mike comments): The pressure rise with the addition of pump #2 was 6 psi at takeoff power. When you reduce fuel consumption for cruise, thus increasing return flow, how much pressure rise do you see? How much at idle?

 

[DanH]

My flow data for 2 pumps was pretty much exactly twice what one pump delivered both SDS and EFII. My fuel pressure regulator went up about 1.5 psi using 2 pumps both systems. Because of the added flow rate pressure drop was a little over 2 psi suction side again both systems.

I'm not at my shop that has a bunch of my previously purchased billet filters but the one that I had (Holley 100-micron stainless screen) produced 0.12 pressure drop 1 pump and about twice that with 2 pumps.

I went out on a limb and connected 4 hard AN6 90's together and only with 4 could I even get a discernable reading and it was very small. (less than 0.04 psi for 4 fittings)

Nice work Matt.

One question...which regulator?

 

[dmattmul]

Bus Manager and 2 for take-off

To the person who uses a Bus Manager did you not use the auto-switch over feature, or did you energize the second pump in another way?

 

[Toobuilder]

Just did an impromptu hangar test with my airplane. Engine off (fully bypassing), PSI reading

1 pump = 49.2 PSI
2 pumps = 51.4 PSI

So thats a 2.2 PSI rise with all the flow returning through the bypass orifice. Also noted that I have an alarm set at this higher pressure threshold that I have never tripped in flight to my knowledge. Next time I fly I will record the pressure numbers at various phases of flight and report back. Also will look hard at any EGT variation with 1 pump vs. 2.

For the record: SDS fuel pump module, Borla regulator, Andair 2020, #6AN lines supply and return.

 

[dmattmul]

Nice work Matt.

One question...which regulator?

Aeromotive 13129 was the model, I think.

 

[DanH]

You do understand the significance of "open loop", right?

Oh ye of little faith...

Just did an impromptu hangar test with my airplane. Engine off (fully bypassing), PSI reading

1 pump = 49.2 PSI
2 pumps = 51.4 PSI

So thats a 2.2 PSI rise with all the flow returning through the bypass orifice.

Thanks Mike. Good info, and interesting too. Your delta 2.2 psi, Matt's 1.5, and Greg's was 6:

I usually take off with my AFR set in high 12's region for a single pump with fuel pressure at 44, and when both pumps are on the fuel pressure comes up to 50 and my AFR will go low-to-mid 12 regions..

You and Greg running the same regulator?

For the record: SDS fuel pump module, Borla regulator, Andair 2020, #6AN lines supply and return.

 

[airguy]

For the record: SDS fuel pump module, Borla regulator, Andair 2020, #6AN lines supply and return.

You and Greg running the same regulator?

Yes, I'm running the Borla regulator. I cannot explain why one install sees 1-2 psi increase and mine sees 6 with the second pump, but mine sees 6 and I'm using #6 hoses all the way.

Greg, take a deep breath, because you're way off base.

I was running a Honda store in 1985 when the first EFI Accords came off the truck. I managed used car recon throughout the entire US fleet transition from carbs, so I'm well aware of EFI superiority, in hard dollar terms. I bought an SDS system from Ross more than 20 years ago (to fly on a Subaru, BTW). And I'm installing an EM-6 right now.

This doesn't surprise me, knowing your background and the discussions you've had on these forums. That also means that the pressure rise with two pumps comes as no surprise to you, which raises the question of why you would want to hold it up as an "Ah HA!" moment acting like we are hiding something by not talking about it. It's simple hydraulics, you are not surprised by it, why act like that? Maybe I'm being overly cranky here but that is not your usual style.

Next question (noting Mike comments): The pressure rise with the addition of pump #2 was 6 psi at takeoff power. When you reduce fuel consumption for cruise, thus increasing return flow, how much pressure rise do you see? How much at idle?

In answer to that question, I really don't. After immediate takeoff I climb on one pump, all the way to altitude, rich of peak at 2500 rpm. When I level off I pull the mixture back to LOP (usually 16-16.5 AFR) and advance ignition to 29 degrees (above 7000') using the LOP button on the SDS. I have noted no pressure increase when reducing fuel flow from ROP to LOP. I theorize that this is within the reactive zone of the spring and seat in the Borla regulator and thus we are not limited to the orifice effect. The data seems to indicate that with both pumps on the spring in the Borla is saturated (or very close to it), the seat is well lifted, and we are seeing orifice drop.

 

[DanH]

In answer to that question, I really don't.

I was asking about operation on two pumps, the subject of interest. When at cruise power and leaned, what is the rail pressure delta between one pump and two pumps? Likewise, at very low power?

 

[airguy]

I was asking about operation on two pumps, the subject of interest. When at cruise power and leaned, what is the rail pressure delta between one pump and two pumps? Likewise, at very low power?

I had to go back into some past recorded data, but I'm seeing a delta of 5 to 6 psi when I switch pumps at cruise. Single to dual to single. When I switch tanks I also switch pumps, so that both get exercised regularly and I can detect a weakness before a failure.

Low power? The lowest I can give you is a runup scenario, with a constant speed prop set for 2700 rpm you don't pull much power at 1600-1700, and I'm still seeing the same 6 psi there. It's repeatable across my recorded data.

If the EM-6 uses AFR for closed loop, then this becomes a non-issue.

 

[rocketman1988]

I had to go back into some past recorded data, but I'm seeing a delta of 5 to 6 psi when I switch pumps at cruise. Single to dual to single. When I switch tanks I also switch pumps, so that both get exercised regularly and I can detect a weakness before a failure.

Low power? The lowest I can give you is a runup scenario, with a constant speed prop set for 2700 rpm you don't pull much power at 1600-1700, and I'm still seeing the same 6 psi there. It's repeatable across my recorded data.

If the EM-6 uses AFR for closed loop, then this becomes a non-issue.

I will be interested to see how the closed loop system is implemented as the O2 sensors do not last very long with 100LL...

 

[rv6ejguy]

I will be interested to see how the closed loop system is implemented as the O2 sensors do not last very long with 100LL...

Not really designed for 100LL for this reason. We developed this now for Mogas and Swift unleaded users with an eye towards the future when UL avgas is more readily available. GAMI has approval now but I expect several more years at least before that's widely available.

 

[rocketman1988]

Not really designed for 100LL for this reason. We developed this now for Mogas and Swift unleaded users with an eye towards the future when UL avgas is more readily available. GAMI has approval now but I expect several more years at least before that's widely available.

I am not expecting any of the new fuels to be widely available any time soon…

 

[rv6ejguy]

Close to 100 posts about one pump vs two, and finally we hear the fuel rail won't maintain set pressure with two pumps running?



Any notion about the perfect accuracy of EFI just got shot in the ***.

It's never been a secret. Almost all users reported this. With the typical 5 psi rise in FP with both pumps on and consequent roughly 5% change in AFR, there is no significant impact. It you were targeting 12 AFR, both pumps would change that to around 11.4. Two pumps might be used for TO and initial climb, that slight richness wouldn't impact anything except fuel flow and a slight lowering of CHTs and EGTs.

In cruise, we'd normally run one pump and lean to a specific AFR or EGT target either with the knob or LOP button. No impact in this case. There is no point in running both pumps above 1000 AGL in my view.

With Bendix FI or a carb, you might guess at where you want the mixture and lean it a tad for a high elevation takeoff. I doubt if most folks can guess between 12 and 11.4 AFR.

On these air cooled engines, we usually throw fuel at them at TO and climb power anyway to keep the CHTs in range.

 

[rv6ejguy]

My flow data for 2 pumps was pretty much exactly twice what one pump delivered both SDS and EFII. My fuel pressure regulator went up about 1.5 psi using 2 pumps both systems. Because of the added flow rate pressure drop was a little over 2 psi suction side again both systems. This will cause a richer mixture (because of the slightly higher FP) but on take-off probably a good thing for a minute or 2.

I'm not at my shop that has a bunch of my previously purchased billet filters but the one that I had (Holley 100-micron stainless screen) produced 0.12 pressure drop 1 pump and about twice that with 2 pumps.

I went out on a limb and connected 4 hard AN6 90's together and only with 4 could I even get a discernable reading and it was very small. (less than 0.04 psi for 4 fittings)

I wish I could have come up with a method to measure cavitation but judging by the "noise" running 2 pumps seemed to be a lot more. When I connected the filter plus the 4 90's running 2 sounded quite a bit noisier. (More "rattlier") This does not make it wrong as there are some distinct advantages running 2 pumps, especially on take-off.

VAF is a great place to share best practices, hopefully we can continue.

I appreciate the data you present here from your testing. Good stuff.

 

[rv6ejguy]

I am not expecting any of the new fuels to be widely available any time soon…

Same here. We have quite a few users flying on mogas and some occasionally on Swift. We thought we'd be ready and gain experience for GAMI or other UL avgas in future.

 

[rv6ejguy]

Further to my post #116. From Liston, here is what we are talking about here with a 5% change in AFR. Red circle denotes 11.4 to 12 AFR:

 

[Mark33]

AFR displayed on G3x

Ross,

I know that once all of the parameters are dialed in, the controller can actually be disconnected and the mixture be controlled via a small rheostat knob. I don’t mean to hijack the thread, but while it has your attention, I’d like to ask a question: Is it possible to connect the O2 sensor to, and the AFR be displayed on, the G3x screen? If your answer requires any discussion beyond a simple reply, we can always start a separate thread.

Thanks!

 

[Toobuilder]

Flew today and here's what I saw:

Idle: FP1=40.9, FP2= 41.8, EGT change = undetermined/maybe 5 degrees rich
Climb: FP1=48.6, FP2= 49.2, EGT change = undetermined
Cruise LOP: FP1=49.3, FP2= 50.1, EGT change = 10 degrees rich
Downwind: FP1=43.2 FP2= 50.1, EGT change = undetermined

 

[sbalmos]

Mark, it absolutely can be. It's just set as another one of the voltage-based General Purpose inputs. Then set the voltage boundaries to the appropriate lower and upper AFR value limits as specified by your O2 sensor's lambda controller.

Ross,

I know that once all of the parameters are dialed in, the controller can actually be disconnected and the mixture be controlled via a small rheostat knob. I don’t mean to hijack the thread, but while it has your attention, I’d like to ask a question: Is it possible to connect the O2 sensor to, and the AFR be displayed on, the G3x screen? If your answer requires any discussion beyond a simple reply, we can always start a separate thread.

Thanks!

 

[Mark33]

Mark, it absolutely can be. It's just set as another one of the voltage-based General Purpose inputs. Then set the voltage boundaries to the appropriate lower and upper AFR value limits as specified by your O2 sensor's lambda controller.

Outstanding Scott….thanks for the reply!

 

[dmattmul]

FP vs MAP

As a way to monitor my fuel pressure regulator's health I like to look at this and insure it's consistent with previous flights. I ruined a regulator early (My fault) on and this helps insure me it's still functioning properly.

 

[Mark33]

As a way to monitor my fuel pressure regulator's health I like to look at this and insure it's consistent with previous flights. I ruined a regulator early (My fault) on and this helps insure me it's still functioning properly.

Can you elaborate on how you ruined your FPR?

 

[DanH]

Flew today and here's what I saw:

Idle: FP1=40.9, FP2= 41.8, EGT change = undetermined/maybe 5 degrees rich
Climb: FP1=48.6, FP2= 49.2, EGT change = undetermined
Cruise LOP: FP1=49.3, FP2= 50.1, EGT change = 10 degrees rich
Downwind: FP1=43.2 FP2= 50.1, EGT change = undetermined

Deltas, psi:

Idle 0.9
Climb 0.6
Cruise LOP 0.8
Downwind 6.9

That's strange. Mike, do you have thoughts about 6.9 on downwind?

BTW, what regulator are you using?

 

[FinnFlyer]

I know there are strong opinions about this, but if you use Tyco W31 series switchable circuit breakers, you get something that functions similarly to the Honeywell switches you're looking at while also providing circuit protection without needing a separate fuse panel, which is handy if space is an issue.

- mark

Do not use these for critical items. They have a design that guarantees failure (flexing of a copper braid near spot weld). Don't know how may cycles it takes for them to break. Perhaps you'll never reach that with new ones, but with used ones (unknown number of on/off cycles) it's a gamble. Had two (used ones) fail -- not by over-current, simply on/off cycles.

Finn

 

[dmattmul]

How to ruin a FPR

Can you elaborate on how you ruined your FPR?

I had a small leak (Flare was scratched) on one of the fittings on the suction side of the pump that allowed air to be sucked into the system and caused the FPR ball to diaphragm to bounce together. (They are both metal) and wore away part of the seat. The leak was small enough to not allow fuel to escape (negative pressure was greater than the positive pressure) or possibly fuel molecules are larger than air molecules (atomic physicist please add input) so no smell. Damage to ball and seat pics below.

 

[DanH]

It's never been a secret. All users see this.

Good to hear more data is available. So far we have pump #2 ON/OFF deltas from 0.6 to 6.9 across a tiny sample. Who else can report?

There is no point in running both pumps above 1000 AGL in my view.

Or mine. Traditionally we add a boost pump for flight regimes nearer to the surface. Here it means both pumps ON for takeoff and landing, or an auto pump switching device.

It seems like the regulator can't fully bypass high fuel volumes at low power, which drives up the rail pressure. Apparently no one knows for sure. Greg is reporting the same delta at high and low power, and Mike's numbers are puzzling. What do you think?

 

[airguy]

I am not expecting any of the new fuels to be widely available any time soon…

It's already beginning - Centennial in Colorado announced today they are making 94UL available - https://www.cbsnews.com/colorado/news/centennial-airport-unleaded-fuel-airplanes/

It will take a while, sure - but the tide is turning.

 

[Toobuilder]

Deltas, psi:

Idle 0.9
Climb 0.6
Cruise LOP 0.8
Downwind 6.9

That's strange. Mike, do you have thoughts about 6.9 on downwind?

BTW, what regulator are you using?

Not sure if the downwind data is anomalous or not. Will pay more attention on the next few flights and try duplicate.

This is a Borla unit (which has a MAP input).

 

[rv6ejguy]

Ross,

I know that once all of the parameters are dialed in, the controller can actually be disconnected and the mixture be controlled via a small rheostat knob. I don’t mean to hijack the thread, but while it has your attention, I’d like to ask a question: Is it possible to connect the O2 sensor to, and the AFR be displayed on, the G3x screen? If your answer requires any discussion beyond a simple reply, we can always start a separate thread.

Thanks!

We don't recommend flying without a programmer but it doesn't need to be plugged in to run the engine. You'd lose the ability to use the LOP function through the keypad and error code display. If something went wrong, you'd have no idea what and therefore no idea on a course of action to possibly mitigate. We always recommend running with the mixture knob in place too. Can be valuable under certain circumstances.

The G3X could be configured to read the 0-5V linear output from the wideband and convert to AFR.

 

[rv6ejguy]

Good to hear more data is available. So far we have pump #2 ON/OFF deltas from 0.6 to 6.9 across a tiny sample. Who else can report?

It seems like the regulator can't fully bypass high fuel volumes at low power, which drives up the rail pressure. Apparently no one knows for sure. Greg is reporting the same delta at high and low power, and Mike's numbers are puzzling. What do you think?

My 6A uses a factory Subaru FPR and fuel pressure climbs 10 psi at idle when the second pumps goes on. I only normally use only one pump for TO.

I hope to be flying in Les' -10 later this week, will report on FP increase we see there.

I think we know, it's logical and explainable. With enough flow volume, the metering valve is incapable of passing enough fuel to control the pressure. Can't explain the data Mike is seeing at low throttle though. Have to ponder that one.

 

[TASEsq]

One small query:

The increase in fuel pressure after the pumps has been demonstrated - this just means a slightly richer mixture to the engine. Not a big deal.

But would this increase on the output side mean a comparable decrease on the suction side?

If it did the numbers were say 1-7 psi. -7 psi is the equivalent of raising the fuel to approx 15,000ft. If you then took off at a high airport after the fuel was sitting on a hot day… you could have fuel pressure near the 20,000ft mark. Is this an issue? I know people regularly fly at these heights with avgas but not at full power!

Combined with a few elbows at the fuel selector and another sharp 90 going into the EFII pump (banjo fitting to backup pump) perhaps there is some value in Robert’s concerns with running both pumps on takeoff (and using the EFII pump module)?

 

[TASEsq]

There have been 4 respondents to the poll who use EFII with no Bus Manager (one which runs 2 pumps for takeoff).

If it’s not too much trouble, can I ask them to PM me with some details of their electrical system design to support EFII? (Or post here if you prefer)

Does the pilot who uses 2 pumps for takeoff on EFII use the EFII boost pump module and have they ever seen any issues?

For the 1 pump at takeoff respondents - whether they have a home made fuel pump switching device?

Thanks in advance!

 

[M McGraw]

Over thinking

While I find the discussion educational, it feels like a lot of over thinking is happening. I have taken off from Leadville, Density Altitude well in excess of 11,500’, many times. I used both pumps until 1000AGL and my standard target EGT. She performed well. IO-390 SDSEFI EM-5F.

From my past career:

“No one believes the theoretical data except the man that wrote the theory. Everyone believes the test data except the man who ran the test.”

 

[dmattmul]

A/F on EFIS

Ross,

I know that once all of the parameters are dialed in, the controller can actually be disconnected and the mixture be controlled via a small rheostat knob. I don’t mean to hijack the thread, but while it has your attention, I’d like to ask a question: Is it possible to connect the O2 sensor to, and the AFR be displayed on, the G3x screen? If your answer requires any discussion beyond a simple reply, we can always start a separate thread.

Thanks!

See wiring diagram but as others have said no issues. I also like to monitor when my pump 2 is on as it should only be used when pump 1 goes out OR when the system starts up as it senses a low-pressure indication (pumps where initially off during start sequence and switched to pump 2). Then you need to manually toggle to pump 2 then back to auto which engages pump 1. When this occurs, I get a CAS warning.

 

[DanH]

I have taken off from Leadville, Density Altitude well in excess of 11,500’, many times. I used both pumps until 1000AGL and my standard target EGT. She performed well. IO-390 SDSEFI EM-5F.

Marvin, "standard target EGT" means you adjust mixture with the rotary knob to reach a target EGT?

 

[airguy]

In general, I'm not a fan of "automatic switchover" type devices, because they can be set up to switch and not alert the pilot of what is happening. It depends on a light or buzzer to indicate the status, both of which can fail and you don't realize you have a problem. I like having the ability to rapidly bring on a backup system manually, and rely on myself to do that so there is no possibility of missing the fact that the primary has failed. Worst case failure would be the fuel pump - and how long exactly does it take to reach the backup switch, engage, and have fuel pressure again? Two seconds, maybe three? The modern glass will alert you to the nature of problem in your headset "Fuel Pressure!" so there's not a lot of guesswork about what to do next - engage backup.

I'm not going to pour cold water on those that like those systems, just some food for thought. I have dual ECU boards, dual alternators, dual fuel pumps - all on manual backup.

 

[rocketman1988]

In general, I'm not a fan of "automatic switchover" type devices, because they can be set up to switch and not alert the pilot of what is happening. It depends on a light or buzzer to indicate the status, both of which can fail and you don't realize you have a problem. I like having the ability to rapidly bring on a backup system manually, and rely on myself to do that so there is no possibility of missing the fact that the primary has failed. Worst case failure would be the fuel pump - and how long exactly does it take to reach the backup switch, engage, and have fuel pressure again? Two seconds, maybe three? The modern glass will alert you to the nature of problem in your headset "Fuel Pressure!" so there's not a lot of guesswork about what to do next - engage backup.

I'm not going to pour cold water on those that like those systems, just some food for thought. I have dual ECU boards, dual alternators, dual fuel pumps - all on manual backup.

…and one injector for each cylinder, one throttle body, one crank, one cam, one prop…

You get the idea; dig deep enough and you will ALWAYS find a SPOF…

 

[airguy]

…and one injector for each cylinder, one throttle body, one crank, one cam, one prop…

You get the idea; dig deep enough and you will ALWAYS find a SPOF…

Yep, sure enough. And my body has only one heart, one brain, one spine, etc etc ad nauseum. Perhaps we should just stay in bed and have no fun?

 

[rv6ejguy]

If it did the numbers were say 1-7 psi. -7 psi is the equivalent of raising the fuel to approx 15,000ft. If you then took off at a high airport after the fuel was sitting on a hot day… you could have fuel pressure near the 20,000ft mark. Is this an issue? I know people regularly fly at these heights with avgas but not at full power!

Dave Anders has flown out of Leadville several times under hot conditions. Elevation 9934 feet. Hasn't said anything to me about having problems. Exclusively 100LL though so I wouldn't expect any.

 

[Mark33]

We don't recommend flying without a programmer but it doesn't need to be plugged in to run the engine. You'd lose the ability to use the LOP function through the keypad and error code display. If something went wrong, you'd have no idea what and therefore no idea on a course of action to possibly mitigate. We always recommend running with the mixture knob in place too. Can be valuable under certain circumstances.

The G3X could be configured to read the 0-5V linear output from the wideband and convert to AFR.

Roger that. I agree, having the controller in place is best, but I was thinking about situations where there may be limited panel space.

 

[Mark33]

See wiring diagram but as others have said no issues. I also like to monitor when my pump 2 is on as it should only be used when pump 1 goes out OR when the system starts up as it senses a low-pressure indication (pumps where initially off during start sequence and switched to pump 2). Then you need to manually toggle to pump 2 then back to auto which engages pump 1. When this occurs, I get a CAS warning.

I’m not a huge fan of the automatic switching of the pump’s, but in general, it works well. My main concern that I have with this setup is that it relies on a single automotive relay to control both pumps. If that one relay were to burn out or malfunction, both fuel pumps could lose all power. I’m setting mine up with two separate relay’s….one for each pump.

 

[rv6ejguy]

Yup one switch, one breaker and one ground for each pump is best from a redundancy and reliability aspect in our experience.

 

[rocketman1988]

Yep, sure enough. And my body has only one heart, one brain, one spine, etc etc ad nauseum. Perhaps we should just stay in bed and have no fun?

Nope, you are missing the point, which is risk mitigation…and that mitigation is decided by each individual.

I also have dual systems to mitigate risk but still, there are many SPOFs in any system.

 

[M McGraw]

Hello Dan!

Marvin, "standard target EGT" means you adjust mixture with the rotary knob to reach a target EGT?

Yes, I use target EGT (manually rotate the knob) for climb on all flights and for takeoff when I get more than 4,000’ from my base airfield. This is a choice when setting up the SDS system. Auto lean can be programmed, I chose not to do so because I use the system for WAY more than cross country and the secondary curve is important for my varied types of flights. I like having a base curve set as a standard ROP situation at all altitudes (personal choice). The secondary curve is really a linear shift (ignition) and/or a percent fuel reduction, as programmed by the owner.

I set my system up as follows (I suspect you could help me find a better compromise):

Base curve is set every 100RPM as a function of AFR and MAP. 2500RPM and above are set for Sea Level full takeoff power, 175 degrees ROP. Once below 2500RPM I am set for 100ROP for the remainder of the flight. There is an aneroid barometer that adjust for air density and I have tested it to 17,500’MSL many years ago, the results of that test were satisfactory, but as you know I am an engineer at heart. I still use the knob to maintain the target EGT. I probably save less than 1/2 a gallon of fuel for the total flight, but it gives me something to do. Once in cruise, I have the option of manual control or curve 2.

The SDS controller will display fuel knob percentage off zero. From Sea Level to 4,000’ DA I leave the knob at zero. From 4,000’ through 12,000’ DA I vary the knob linearly up to -11%. This procedure is no different than the yellow line I painted near the mixture knob on my last aircraft with a standard aviation carb, except I can see a consistent number on the controller. So less guess work. During the initial climb I adjust for target EGT the same way I did with my carbed aircraft; however, the adjustments tend to be less necessary and the EGT far more reliable. Once in cruise, I typically set the fuel at -30% sometimes -33% as a function of altitude, up to 50% is available. I’m actually choosing a value relative to PEAK EGT. I removed the O2 sensor more than 800 hours ago.

The secondary curve can adjust timing and fuel percentage individually or together. Because the ignition is a linear offset from the base curve and the fuel reduction is shown in percentage, I have a system for setting Curve 2 based on my anticipated flight and I can change my mind in flight with no adverse effects. I use Curve 2 several different ways.

- Standard local flights.
- Long LOP flights
- formation flights
- Aerobatics at altitude
- occasionally long LOP climbs

The above is why I chose my base curve, every flight is relative to a known starting point. Were I to use mogas, an O2 sensor, and closed loop my choices would be significantly different.

As far as one fuel pump versus two, my preference is two pumps below 1000AGL. It just makes me happy. My fuel pressures vary similar to what was posted by TooBuilder. If there is an over rich condition with two pumps I have never noticed it, possibly because I have a system for takeoff based on DA?

You and I watched a carbed O-540 make a poor attempt at an over-rich takeoff in Leadville last year. Your aircraft (mechanical) and mine (EFI) did quite well. Perhaps Ross will read this and improve upon my understanding of the SDS system. I’m a gear-head wannabe, with limited experience on only one EI/EFI airplane. I am very open to improvements on my use of the system which is why I love this forum. Perhaps Ross will read this and offer an improvement on my understanding of SDSEFI.

 

[rv6ejguy]

What you describe here makes perfect sense Marvin. You understand the system well.

What we've learned as that people have different preferences and missions so we need to make the system adaptable to those differences.

The EM-6 will bring some more possibilities into the mix and at this time, we are not sure what the most popular setup will be at this time and that may change as unleaded fuel becomes more prevalent.

BTW, Les is seeing around a 4 psi rise with the 2nd pump going on. 393 pumps, AN6 plumbing, Borla reg.

 

[DanH]

Yes, I use target EGT (manually rotate the knob) for climb on all flights and for takeoff when I get more than 4,000’ from my base airfield.

Got it. Me too.

As far as one fuel pump versus two, my preference is two pumps below 1000AGL. It just makes me happy.

Same again, notably for takeoff and landing.

My fuel pressures vary similar to what was posted by ToolBuilder. If there is an over rich condition with two pumps I have never noticed it...

As Ross noted, moderate additional pressure should not richen delivery to an unacceptable degree. Under full power, the mixture vs power curve is relatively flat, so lacking manual or automatic compensation (i.e. a knob tweak as above, or a closed loop system), the power loss is not huge.

The SDS ECU derives fuel flow from injector pulse width, not actual flow, so a rail pressure change also degrades indicated FF accuracy.

Is it a Big Deal? Not really, but...

I have always assumed the regulator held rail pressure to the set value, and the caution for builders was to ensure adequate return flow capacity for all conditions. Note a recent thread regarding return line size and fitting size for the Andair valve. Turned out there doesn't seem to be any hard data relating return capacity to rail pressure...and folks are using both -4 and -6.

Now it seems no one really knows why rail pressures vary with two pumps, or why the deltas vary when comparing installations, or why the deltas don't seem to mirror return flow.

I'm not the guy who sees a gremlin on the wing, but I do like to know why.

Break.

FWIW, a new player in the EFI market is incorporating a pressure sensor, the output of which is used to modify injector open time with variation of rail pressure...closed loop compensation.

You and I watched a carbed O-540 make a poor attempt at an over-rich takeoff in Leadville last year. Your aircraft (mechanical) and mine (EFI) did quite well.

We're both using manually adjusted target EGT on our 390's. The practical difference is the EFI system begins at a point closer to the desired mixture, as it follows air density more closely than my constant flow mechanical cousin. Still, the trusty FM200 doesn't lose much even with the knob full rich. For example, I once did a test climb at WOT and full rich from takeoff to 15,000 feet. The resulting EGT shift (200F) suggests less than 5% power loss, again due to the mixture vs power curve being relatively flat in the rich of peak regime.

I don't have any values to quantify power loss with that 540's carb. Sure looked scary from overhead!

 

[rv6ejguy]

Now it seems no one really knows why rail pressures vary with two pumps, or why the deltas vary when comparing installations, or why the deltas don't seem to mirror return flow.

As I have mentioned a couple times in this thread, yes we do know why- The FPR valve/ seat can't flow enough volume to hold pressure down. Analogous to putting your finger over the end of a garden hose- pressure rises as you restrict the flow further.

I think Marvin used to run 50psi FP as that was our recommendation on 390/580 engines due to a limitation of 32ms on injector pulse width back then. This was addressed in 2020 with V31 software, mainly for Reno racers. The higher the static FP, the less return flow is required so this may be one reason why Marvin sees less FP change with both pumps on.

Finally, there is huge variation in plumbing, fittings and bends between installations which will create either more or less restriction in the system.