Hello everyone. Sorry if this is the wrong topic to post in, but hopefully some automotive people could help me on some information.

I have done an extensive amount of research on the topic of automotive fuel pumps. I have an application that is limited to very low current ~1500mAh in which a fuel pump type system should be used. We are exploring various options at the moment, but I need some clarification.

I have read on several sites that when a standard fuel pump turns on by a person putting an ignition key into the car, the pump runs for say 5 seconds then "shuts off". This is essentially the pump building pressure in the system. However in a return type pump system, I would assume the pump runs continuously, thus drawing a massive current.

In a return less type system, the fuel will not be recirculated, instead the pump should accumulate the fuel and pressure in a fuel line via a check valve. So my understanding is that once the pressure is built up, the pump could then be shut off entirely thus drawing little or no current. Is this the case? We essentially want to build a system that draws as little current as possible by only running a massive fuel pump when we need the pressure to the system to be maintained. A standard fuel pump can draw say 10 amps. I assume this means when the pump is on and not in the 'regulating' state of maintaining the pressure in a return less system. Is this assumption correct?

We can buy or build a pressure regulator with a controller. This is to be integrated into a fuel injection system. However, in order to save current, since the pressure should only decrease when a fuel injector is opened, the pump should run during these times to 'fill the pressure back up'. Do you think this is possible by pwming or turning the pump on and off dynamically?

Any help is appreciated. Thanks.

 

The fuel rail pressurized to 20 PSI. Sounds like 120 watts is a lot to do that. What does your extensive research suggest is the actual draw of a pump?

 

20 PSI? We are looking for around 40PSI for this application. The 10 amp draw was just a figure (as I literally said the word "say", which should indicate that its not a real figure). We have found standard fuel pumps anywhere from 1.4 amps (not sold) to probably about 8 amps. Either way, a continuous draw of say 5 amps from a 1500mAh battery would last what.. 20 minutes? It is hard to find a low current draw fuel pump made for small engines and not giant automotive ones, which have the PSI rating we would like.

The idea is to run the fuel pump only when pressure is below a certain point. Doing the calculations, for 4000RPM and a 4 stroke engine, you get one cycle (4 strokes) lasting around 30 milliseconds. This means that the only time that pressure in the system would drop would be when the injector is open, which is once per cycle, so in essence 1/4 of 30 milliseconds or 7.5 milliseconds. Now factoring in all that off time of 3/4 of 30 milliseconds, we could get a massive savings if the pump did not run during this time. We have other electronics that must fit within the current consumption budget as well. We are trying to optimize sections of the design so that we can get the lowest consumption possible. Of course, we have to factor in inrush currents and the like. we plan on factoring in inrush current times and building simple inrush current limiters.

We can essentially use a high powered pump if its not on all the time, but my original question is with regards to the viability of switching a pump this way. Since the device is electromechanical, and I do not have a ton of experience with automotive fuel pumps, I decided to ask if anyone else had any info.

Thanks

 

Some of the reason for a return-type system is to keep the fuel rail from getting to hot, causing the fuel to boil, resulting in a vapor lock type situation.

Don't recall offhand at what temperature gasoline boils and develops pressure, but keeping the fuel rail pressurized to a couple of atmospheres plus helps to increase that boiling point - just like water boils at a lower temp on a mountaintop than it does at sea level.

If you wind up with the fuel boiling in the rail(s), you'll get less fuel out of the injectors which will lead to detonation. Fuel that's in steam form would meter differently through the injectors.

Turning the pump on for several seconds when you first turn the key will help to reduce the possibility of vapor lock, but not much point in keeping it running unless the engine is running too.

You might get some kind of energy savings using a synchronous buck arrangement, having its' output controlled by the fuel demand. However, you're going to have some rotational delay (mechanical lag) and electrical delay that will be much slower than merely changing the PWM signal to a fuel pressure regulator.

I suppose it depends upon your application; if it's strictly an economy-type vehicle where you can control the engine performance (flat-out acceleration isn't important) you might be able to get away with something like that. I'm skeptical if it would work well in a racing application, though.

 

We can get a pressure regulator. The question is, will the regulator maintain a constant pressure in the system by dynamically pulsing the fuel pump? What I mean is say the pump draws 5 amps fully on. If we get a regulator, when the pump pressurizes the system to 40PSI, will the pump then draw only say 0.5amps? The reasoning is that there is an initial burst of current to pressurize the system. then the regulator essentially would top off the pressure as the injector opens and closes with a much lower current draw. From many websites it seems like the regulator maintains the proper pressure, but they neglect to comment about current draw when in the 'topping state', if this state exists.

The vehicle is purely economical.

Thanks

 

I have done an extensive amount of research on the topic of automotive fuel pumps. I have an application that is limited to very low current ~1500mAh in which a fuel pump type system should be used. We are exploring various options at the moment, but I need some clarification.

fistly, you'll need to consider what volume of fuel your consuming and at what pressure. Do this at full load. From there, you'll need to size the pump to deliver that requirement, and formulate a prime mover to provide the required pump drive. Less than full load, you could modulate the pump.

because your fuel volume is a function of designed engine output, you may have the ability to work with pressure. Injectors require a certain pressure to perform thier task, however, for years auto makers utilized a low pressure 'injector' that replaced the venturi of carburation. The old float and needle of the carburator required a low pressure pump, most of which drew no current.

Regarding your last question, most automotive fuel pumps are fixed displacement. As such they pump the same quantity of fuel at all pressures, at all times, so thier draws are relatively constant. If you could research and find a variable displacement pump, then you'll have an opportunity to save power, as such pumps maintain preset pressures, but vary the flow according to demand, and subsequently have variable power requirements.

Irregardless, you need to service demand, so get that figured out first.

 

Check how motorcycles do it. Small pump, small injectors.

 

Hi .

you will find most automotive fuel pumps run for a short period when you first turn on the ignition but once the engine is running the pump runs continuously at full current and they aint light things just look at the size or the relays that normally control them .

You could always use a pump from a modern HDi style diesel engine for greater pressure but you still have the current consumption , it boils down to you aint going to get pressure and maintain it for peanuts it is going to require current

 

i would think turning the pump on and off would use a lot more energy then having it on

I would look at trying to speed control the pump so there's no sudden surge when turned on and when you 40 psi is reached turns the pump right down but not off

When the engines running the pump will be more or less on all the time even with you original pressure switch design

On my car when i turn the ign on yes you get the first 5 secs of fuel pump running them it shuts off and this is a return system, Then when you start the car up it runs all the time

Normal petrol injection cars the fuel flows from the pump through the filter then to the pressure regulator. On the pressure regulator there is 3 connections one is flow in the other flow return to the tank and the third is the outlet to the fuel rail where the pressure is maintained for the injectors


You cant use the hdi diesel pumps for more pressure, The only pump on them that are electric are the low pressure fuel pump that are sat in the tank which have less pressure then the petrol pumps, Where the hdi engines build their pressure is off the high pressure pump driven of the cambelt

The low pressure side of the hdi is around 2.5 bar around 36psi std petrol regulators are around 3 bar

 

Sg Wookie is right the fuel is constantly circulated to cool the fuel rail & diesels do the same. Once the engine is running who cares if the pump run all the time they do have a fairly long life. The only thing i can see as a gain is vary the pump RPM by PWM to match how mutch fuel is actualy being used. Could be done by using Injector pulse width as a reference. Ime suprised manufacturers haven gone this way. Ive cut a few pumps apart after they got noisy Or int start. Always the comutator has had sements worn through to the insulation. Comutator seem to wear out long before brushes. Most cars in Australia run on about 30 to 35PSI. It would still need to run fastenough to give some fuel return.

 

This is a standard Bosch fuel pump for a 6Cyl car & pulls 5.5A @ 35PSI.

 

Oh wow. Thanks for all the comments guys and the pictures.

The engine I am working on is a single cylinder. We found a small engine fuel pump that does 40 PSI at 1.4amps, but unfortunately, we can't find a place to buy it. Its the Delphi T11 series.

One idea I had is similar to what someone had in another reply, which was to not turn the pump completely off, but rather turn it down low to save some power when pressure was regulated. I am doing more research on the matter, but as it stands now, a return less type system seems to be the best option for us in terms of current. Since a check valve operates as such to only let fuel flow in one direction, a check valve seems like the best option to build up the pressure in the fuel line. Once the pressure builds up initially, I can turn the pump way down to conserve current. When the engine starts running, the pump can dynamically be sped up or down depending on various operating conditions with the hope that the pump wont be in the continuous on state all the time, drawing maximal current.

 

Most fuel pumps Ive dealt with already have a check valve built in them. They also have an over pressure valve in them, which is a spring loaded valve which you dont want to activate if at all posible as they dont always reseat, this will render the pump as useless. The pumps are relatively expensive. Here in Aust $250 or more. Its mutch safer to use a bypass regulator as in pic, easy to source from wreckers. For your pump why not see what EFI motorbikes use, these must be reasonably compact to fit in a motor bike fuel tank.

 

There are two small engine EFI kits we have been looking at. One is from MBE motorsports and the other is from delphi. The delphi one.. is not commercially available. We have been having trouble getting our order through MBE motorsports. We are in the process of getting the pump in that that kit uses.

The problem is that.. we believe that they are just small fuel pumps designed to be used with a lighter load, such as lower psi and current consumption rating. This is not really what we are looking for to solve the problem. The lower current consumption does lessen the problem, but it doesn't 'go away' so to speak. A 2 amp draw is better than a 5 amp draw, but the inherent problem is still there. We have a battery that can only do so much, and in addition to the fuel pump, we have other electronics that need to go on the same battery. We just think there should be a better way to do a constant pressure system with some sort of valves.

I mean.. the problem to me is kinda like a reservoir. You have to fill up the tank initially with a giant load of water; however, after the tank is full, you only have to top it off with a much lighter water load. The water in this case would be analogous to current draw.

 

Build or buy a small pressure tank. Tee it to the injector feed, add fuel to the lower portion of the tank, then charge the top of the tank with 40# of air. Use a pressure transducer on the top of the tank to control pump operation. The tank will maintain pressure to the very low flow injector and by tweaking the pressures and hysteresis, the pump would hardly run at all as all it would be required to do is maintain tank pressure. This is how irrigation pumps work down on the farm.

 

What are you using this engine for? If you use an alternator you wont have to worry about power use. If that's a problem get a mechanical fuel pump and use a pressure regulator to send any fuel back to the tank.

 

Well, as far as I know - the pump uses the fuel for it's own cooling. If the pump is running at all, it needs to have a flow through it, which is another reason why they have that fuel-return system where most of the fuel returns to the tank.

One fairly common failure mode for fuel pumps is if the pickup in the fuel tank or a filter prior to the pump becomes mostly clogged, and the flow through the fuel pump becomes severely restricted. It'll still supply enough fuel to keep the engine running, but not enough to keep the pump cool, and it fails.

I'm not sure how effective putting some sort of heat sink on the outside of the fuel pump would be, but that dissipated power needs to go somewhere.

 

We just think there should be a better way to do a constant pressure system with some sort of valves.

Pressure stems from flow, more precisely, the restriction of flow. Flow, flow, flow. How much flow do you need to get the desired output from your engine???

 

I believe what JaguarJoe said was a very good suggestion. After talking to one of my comrades, we think we will investigate this option at it seems to have a lot of promise.

GetDeviceInfo:
The flow for our engine is extremely low. Our engine is a 150cc engine that is designed to operate at a max of 3.5hp. Again, any flow rates are extremely low. I don't know the exact flowrate, but anything that is akin to something an EFI kit on a scooter would use. At this point, we are not exactly interested in discrete numbers, but rather, we are interested in a lot of theory in order to get current consumption down to its lowest possible levels. What we know are the basic numbers that an off the shelf fuel pump gives us: PSI, current consumption, etc. We can adjust a lot of other variables later.

 

Frankly, the post should be pulled. You will basically be building a bomb.



The best current consumption is going to be obtained from a high speed pump that is rated marginally over your flow requirements.