Fuel Delivery Redesign: "Project Number Two"

[rbattelle]

Well, I decided to test with vegetable oil because I was anxious to get some testing in and the weather here has recently turned to complete crapola. Veg. oil lets me test in the garage with no fears of fumes/fires.



Anyway, I ran the first phase of tests last night. These were done without the filter in the loop; just the pump and regulator. I found the pump self-primes very easily (as expected).



The pump current varies linearly with changes in RPM, with maximum RPM corresponding to maximum current draw. I plugged the regulator output line and ran the pump at maximum pressure and the maximum current draw I could achieve was ~4. 6 amp. Very mild.



So I ran an endurance test for 1 full hour, collecting temperature data at the MOSFET chip the whole time (sampling at 1 minute intervals). Pump at maximum RPM, current draw ~4. 5 amp average. Maximum temperature achieved was 86. 3*F. Horizontal asymptote is probably around 100*F, but that's just a guess (not enough data).



The pump was pretty noisy, although that's probably because it's bolted to a piece of plywood. When I fabricate the mounting bracket I'll need a lot of rubber isolation. Lmills, who was running the same pump, said he was able to hear it running with the truck at idle.



Tonight I'll plumb the filter into the system and take some pressure data. But it's pretty clear that the PWM will work quite well in this application.



-Ryan

 

[rbattelle]

A bad day for science

Ran the pump today with the filter installed and ran into 2 problems. The pump ran fine, and the filter worked fine, but I discovered a faulty sending unit on my fuel pressure gauge. What's worse, I think it's my fault because I caught an error in my wiring diagram (had the sending unit grounded to the wrong spot). So I had to order a new sending unit.



Then I was setting up for another maximum-current test and managed to blow the fuse on my brand new fancy-pants multimeter. Gotta wait a few days for a replacement. Note to self: do not short battery terminals across your ammeter.



After that I decided it was a bad day for science so I quit. I think I'm going to remove the regulator assembly (I was keeping it in order to simulate the presence of the CP-3) as I think I can test the system alright without it since the pump maximum current doesn't seem to be much of a function of output pressure. When I test the variation in pressure with RPM I'll just run the output line plugged (so the gauge reads total pressure). This will cause the pump to cavitate (I think), but it'll be for short periods anyway.



I love experimentation.



-Ryan

 

[nickleinonen]

something that might help save some fuses, make up your own set of leads for the DMM with an external fuse on the leads... i am guessing your meter has a 10a fuse in the meter, so some external fusing at say 8a would save your internal fuse... . my pos digital meter is pretty good, but the one thing i don't like on it is the higher amp side isn't fused, but it is rated for 20a for 10sec [but the 300ma is fused]... so i have a cable set made for it with a 10a fuse inline to save my meter

 

[rbattelle]

Yesterday I tested the system's ability to make pressure. I removed the regulator from the pump and tested the pump with and without the filter. The pump builds pressure very quickly, and I should be able to achieve the 8-10 psi I'm looking for at the CP-3 with only 2. 5-3. 5 amp current draw. Maximum current draw is 4. 5 amp with the pump at 97% duty cycle, 20 psi total pressure (pump output plugged).



I did not measure flow rate.



Next step is to make a mounting bracket for the pump. Unfortunately, the weather here has recently turned to some of the absolute worst kind you can get (high winds, mix of rain/sleet/snow/hail/thunderstorms, frigid temperatures), so I'm not motivated to go crawling around under the truck measuring things at the moment.



-Ryan

 

[rbattelle]

Problem?

Tonight I decided to test the flow rate on the pump. My results were interesting. Working with vegetable oil (soybean), I was able to achieve a flow rate of only 41 GPH at maximum speed across the PWM. It turns out that the PWM isn't able to run the pump at 100% speed (not sure why, yet). But when I ran the pump directly (bypassed the PWM), I was able to achieve 5. 8 amp current draw but still only 49 GPH flow rate (max PWM performance is 4. 7 amp and 41 GPH). Note that the factory minimum spec is 38 GPH.



There are 2 possible explanations: the thicker consistency of vegetable oil reduces the flow rate or the -6AN lines I've used are not large enough.



It's interesting to note that Aeromotive provides their own data indicating the maximum flow rate on the pump is around 175 GPH.



Aeromotive specifically recommends -6AN fittings and line for this pump, so that can't be it.



Must be the difference in viscosity between petroleum fuel and vegetable oil? The difference is around 9 times (vegetable oil is 9 times more viscous than diesel).



If we assume Aeromotive's data is accurate, then the pump flows 3. 5 times more at maximum power with gasoline than it does with vegetable oil (175 GPH/49 GPH = 3. 5). Then, assuming that factor is roughly constant, it should flow 97 GPH minimum and 146 GPH at maximum speed with the PWM.



All this could be cleared up by simply running the pump using diesel instead of vegetable oil, but I'm just trying to get an idea whether the system is good before moving on to diesel. I'd hate to get it all installed on the truck only to find out it doesn't flow enough.



-Ryan

 

[roverhybrids]

yes the thicker veggie oil reduces flow rates.

Does the PWM limit the maximum amp draw?

 

[csutton7]

I haven't visited this thread since I posted on it in the beginning, I will say that you are taking your time and doing as much research as possible--very intelligent --



I will add--I went thru 3 pumps using the stock pickup and that's why I went full gravity when I did this 4 yrs ago--one was a Holley, the next was the Barry Grant and the last the Mallory 110 FI---all three bit the dust within months of installation--all 3 were wired using the relay, so it wasn't lack of power on that end--the only thing I could attribute it to was the extra work needed to suck the fuel out of the tank and all of these pumps were mounted on the frame rail with regulators and return lines--I think Gary k7gld had made an early comment about flow coming out of the stock pickup--now one thing to consider is that when the tank is full there may be sufficient head to provide proper head pressure, but when the tank has less and less fuel in it I think the pump has to work much harder than it's designed for and thus the failures--my return line I tapped into the rubber hose part of the fuel fill line vent using a tee--only draw back is you can't fill up your truck with the engine running--if some of you do this---I did change my pump after like 31 months of use, but it was just a loose connection and could've stayed on, but I was going on a trip and said screw it--it is now my spare---it will be interesting to see how long the FASS, Air Dog and all the other systems work with sucking out of the stock pickup--they all have some time to go before they catch me---downfall of my system is the ground clearance lost as stated earlier, but beyond that I think it's proven----one other think is my truck is highly modified and I drive it everydayand I use at sled pull and dyno events and the fuel system has performed flawlessly---chris

 

[rbattelle]

I will add--I went thru 3 pumps using the stock pickup and that's why I went full gravity when I did this 4 yrs ago--one was a Holley, the next was the Barry Grant and the last the Mallory 110 FI---all three bit the dust within months of installation--all 3 were wired using the relay, so it wasn't lack of power on that end--the only thing I could attribute it to was the extra work needed to suck the fuel out of the tank and all of these pumps were mounted on the frame rail with regulators and return lines--I think Gary k7gld had made an early comment about flow coming out of the stock pickup--now one thing to consider is that when the tank is full there may be sufficient head to provide proper head pressure, but when the tank has less and less fuel in it I think the pump has to work much harder than it's designed for and thus the failures--my return line I tapped into the rubber hose part of the fuel fill line vent using a tee--only draw back is you can't fill up your truck with the engine running--if some of you do this---I did change my pump after like 31 months of use, but it was just a loose connection and could've stayed on, but I was going on a trip and said screw it--it is now my spare---it will be interesting to see how long the FASS, Air Dog and all the other systems work with sucking out of the stock pickup--they all have some time to go before they catch me---downfall of my system is the ground clearance lost as stated earlier, but beyond that I think it's proven----one other think is my truck is highly modified and I drive it everydayand I use at sled pull and dyno events and the fuel system has performed flawlessly---chris

Chris, your contributions to this thread are greatly appreciated, as you are one of the "pioneers" in this game. Indeed we will see if the stock pickup kills this pump. If it does, I will certainly need to reconfigure it down the road. I wonder if the 3rd gen stock pickup has a larger capacity than the 2nd gen pickup did?



roverhybrids - the PWM does not place any limits on the current flow to the pump (I once accidentally flowed 120 amp through the PWM ). I need to do some more research to see why the PWM isn't giving me 100% rpm... I think it may need a resistance change at one of the timer circuits.



-Ryan

 

[csutton7]

Ryan--good point about the pick-ups--of that I have no info--would be interesting to hear--keep up the work and good luck--chris

 

[rbattelle]

Made a minor change to the PWM circuit that bought me another 1/2 amp. So now I can achieve 5. 2 amp at the maximum setting. That should be plenty, assuming the pump will flow diesel fuel at least twice as fast as vegetable oil (a reasonable assumption, given the earlier data). I'm attaching a plot of my flow rate vs. current data (RPM is linear with current, so this plot is effectively flow vs. RPM).



The highest-current point (flow rate 49 GPH) is with the pump connected directly to the battery (no PWM). The other points are with the PWM circuit connected.



I'm really excited to get this thing on the truck and see what she'll do!



-Ryan

 

[nickleinonen]

the PWM does not place any limits on the current flow to the pump (I once accidentally flowed 120 amp through the PWM ). I need to do some more research to see why the PWM isn't giving me 100% rpm... I think it may need a resistance change at one of the timer circuits.

i was reading that web page you linked to earlier that had the PWM controller. . there is a limit on there... 7 amps using just the printed circuit board as the conductors, and 16a if you add extra copper wires to externally trace the power path... . and i don't think you will ever get 100% voltage out of that model PWM, as they say on their site, the max PWM is 99% on cycle... all that is in the pdf link on the page with the pwm controller

 

[rbattelle]

Yeah, all that is true which is why I'm content to be at 5. 2 amp. I did thicken up the traces on the PCB as they recommend using some bare 12 ga. solid copper wire.



-Ryan

 

[rbattelle]

Update

I've got the system completely wired into the vehicle electrical system. Cab controller is mounted on the dash, underhood junction box is installed atop the fuse box. All that remains is the plumbing (which is the hard part).



I've attached a photo of the filter mounted in place but not plumbed. I was working on cutting the hoses to length. The elbow with the brass plug in it is where the fuel pressure will be taken. Originally, I was going to just screw the pressure transducer directly into that elbow, but it dawned on me that the vibration might not be so good for it, so I think I'm going to put in a braided stainless isolator. That mounting location looks really nice (almost factory!), but to be honest it's not very easy to unlock the filter from that position (hard to get a good grip on the locking ring). I may have to unmount the filter to change it out. Oh well.



One thing I recently became concerned with is heat transfer from the engine to the filter mount. The mounting bracket is aluminum, so it's an excellent heat sink. I noticed after shutting the truck off the mount got "very hot". I began to fear the possibility of vaporizing fuel at the filter if it gets too hot (boiling point of diesel is at least 392F). So I decided to take some temperature readings. I've attached a plot of the mount temperature over a 75 minute period. The first reading is with the engine running at idle (it was shut off right after the first reading, readings are taken once per minute). I think the plot is a very interesting commentary on the heat soaking of the engine once there's no airflow across it. Obviously, "very hot" to my hand must be 137F, since that's where the temperature ends up over time.



Bottom line: heat transfer from the engine to the filter is of no concern.



I thought others might be interested in seeing the data.



-Ryan

 

[nickleinonen]

very good so far... i wish i had so much free time

 

[rbattelle]

Good news bad news

I have good news and bad news.



The good news is I got the entire system plumbed and up and running! As advertised, I'm able to vary pressure at the CP-3 by turning my little dial on the cab controller box. I tested the flow rate of the pump yesterday, and with the pump running around 8psi I'm able to achieve around 60 gph. Minimum factory spec is 40 gph, so I think I'm in good shape.



Now for the bad news. I started the truck up before all the hoses and fuel lines were tied up in their final positions (can you see where this is going?). Engine started and ran good, and I had good pressure, so I decided to take it for a quick spin down the block. After rolling back about 10 ft, I hear this sudden horrible grinding/ripping/tearing/snapping noise from under the truck and the fuel pressure drops to 0. Instantly I get that horrible sick feeling in my stomach as I begin to visualize my poor sweet innocent truck being rendered utterly destroyed by my new fuel system. Shut the engine off real quick and run up to check under the hood. All hoses and electrical connections are fine. Go back to the cab, turn on the manual pump switch, no pump.



So I crawled under the truck to discover that the wiring from the pump had been in contact with the front driveshaft. It had wrapped itself up around the driveshaft, snapped the wires apart up toward the engine compartment, and pulled the positive lead right out from inside the fuel pump. Took the pump apart and sure enough, the positive wire had been ripped off where it's crimped to the positive bushing wire.



Feeling like an utter failure, I began to despair. My neighbor stopped over and suggested that the motor might be repaired, since other than the wiring there was no apparent mechanical damage. So I carefully soldered a new positive wire and managed to get the pump reassembled. I checked it across one of my batteries and to my amazement it actually ran (I've had bad luck repairing electric motors in the past)!!



So I repaired the damaged wiring harness and reinstalled the pump. Turned on the pump, started the truck, adjusted to 12. 4 psi, and everything was good. After idling for about 20 seconds or so, suddenly fuel pressure drops to 0 again. Shut it down, pump wouldn't run. Checked all the voltages and connections, then pulled the pump again. Connected the pump right to a battery and it ran for about 1/2 second then stopped.



Pulled the pump apart again, and one of the bushing springs had popped out and mangled itself in the armature. Pump is now beyond repair. Called Summit, and I'll have a new one here Tuesday. That means no transportation for me until Tuesday night. My wife and I will be a 1-car family for a few days.



To be honest, I'm feeling rather discouraged. I could use some encouraging words.



-Ryan

 

[Matt400]

To be honest, I'm feeling rather discouraged. I could use some encouraging words.

We thank you for making those mistakes so that we can learn what not to do :-laf



Seems you had it and excitement won over patience, reminds me of some of the stuff I do!

 

[rbattelle]

Seems you had it and excitement won over patience

Amen to that! I've learned to double check things for binding/catching in the past, but I guess that lesson was "expired".



Actually, I'm thankful that it wasn't much MUCH worse. I shudder to think of the carnage that could have happened if the shaft had caught the braided stainless fuel line instead. Yikes.



-Ryan

 

[ThomasLawrence]

To be honest, I'm feeling rather discouraged. I could use some encouraging words.

Well... like Forrest Gump said, "it happens". And it happens to all of us. I had the same gut-wrenching feeling when I watched my transfer case try and commit suicide by jumping off my transmission jack onto the concrete floor - because I was too impatient to secure it with the safety chains.



In the end, you'll get the new pump, throw it in, triple-check everything, and it'll run just fine. A week from now, you'll laugh at yourself (or at least you won't be beating yourself up over it). Like you said, it could have been worse.



Of course... had Dodge not gone with the STUPID non-disconnecting-front-axle design, and left well enough alone with the CAD, this never would have happened (that is, until you were out somewhere and needed 4WD, only to be left stranded in mud up to your knees, with the sun going down on a Friday evening... ). Yeah... definately could have turned out worse



-Tom

 

[nickleinonen]

I could use some encouraging words.

i guess shiz happens doesn't really fit here does it... ?



i've messed up like that before... at home and work... at work i don't care so much. . i've assisted in very much mangling a 30T overhead crane, making a traction motor/wheel combo [about 13,000lbs] drop nearly 2' [leaving it dangling 10' in the air] and having parts rain down from the drum, brake and motor... nearly had to change my shorts on that one...



this is just a small f-up... no worries... get the new pump, chrome the old one and keep it as a trophy/battle scar memory... it'll remind you to not do this again :-laf

 

[rbattelle]

Final Cost: $1000 - $1200 (Science is Expensive)

I compiled a list of the final components I used to build this system. You may notice the list does not include wire. There's actually around 240 ft of wiring in the system, but the wire cost is not that high and it could actually be done with much less wiring (I included a lot of extra for future expansion). The total cost is a little misleading, as it assumes only the absolute highest quality components (in most cases), and it could be built much cheaper. For example, the total cost for the MIL-Spec circular connectors (made by the same company that makes them for the space shuttle) is $170. That cost could be eliminated altogether if the system were simply soldered, for example. Also, I used only Aeroquip AN fittings; they can be had cheaper elsewhere. Swagelok makes some of the world's best pipe fittings, but the same stuff can be had much cheaper from elsewhere. I made all fittings stainless steel to avoid any possible interaction between brass and fuel. If the system were built from brass, the fittings cost would be about 2/3 to 1/2 that of stainless.



Attached is my spreadsheet that covers exactly what was actually used on the system, except for wire and other misc. components like self-fusing tape, RTV, and a few resistors for the LEDs. There was no point in including those items, since their cost is pretty low and most people who would do a project like this have them laying around anyway. I also didn't bother to include the braided sleeving I used to cover all the wiring.



With the exception of the fuel pump wire, the entire system uses 18 ga. stranded copper wire with SXL cross-linked insulation. That wire is resistant to every chemical on earth (almost) and is very high temperature and very flexible. The fuel pump wire is plain old 14 ga. stranded copper wire with PVC insulation.



Including all my development work and all the fittings, hose, wire, and components that I bought but weren't actually used, the price probably goes up to around $1500. Ouch. I had hoped the final cost would be less than FASS or Airdog, but it doesn't look like that's the case. I can only hope the system is as reliable as it's price suggests. I've had it on for 1 week and have no problems (no leaks, no electrical trouble)... yet.



#ad




-Ryan