I was talking to a friend today about running the return line back into the feed line on the low pressure side between tank and pump when doing EFI LS swaps as a way to avoid drilling into the stock tank to send the return. He asked why I'm not just doing PWM on the fuel pump.

Good question.

Seeing as how PWM controllers can be had for $300, why are more people not using them? You eliminate the return line and regulator all together. The controller pulses the fuel pump to maintain pressure. You get a cooler running pump, less noise, less amp draw, and more pump life. The cost is roughly a break-even with less plumbing hassle.

I have everything I need to make a closed loop PWM controller to run my dual AEM 380s full time so I can do away with hobbs switches, regulator, return line, and check valves. They'll run quieter, cooler, and with more pump life. I can even do rising rate pressure with a MAP signal.

So again, why don't more people go this route???

 

It might not be as common a knowledge as it should be.

 

LOTS of OEMs are doing it now. You'll even see it on the vehicles with 100,000 mile warranties, etc. I'm going to set one up just to see how well it works, I just might have to borrow someone's car lol.

 

It DOES look like a good idea, should simplify any fuel injection system.

 

Links to these controllers ?

The only aftermarket place I've seen offer a standalone setup ( based on GM parts ) is Vaporworx ?

I know some people who have built their own setups...but controlled by the ecu.

IMO it is definitely something worth looking into, but at the same time I wouldnt be running it without failsafes to ensure no harm can come to the engine if things do not work as planned, which will require a good aftermarket ecu anyway.

And the OEM controllers....what sort of current can they handle ? Will they handle the 20....30...40 or more AMPS a powerful multipump setup will require for anything that actually makes some power ?

 

The aeromotive setup is around $300 and varies fuel pump duty cycle with RPM.

The Vaporworx setup is better since it takes fuel pressure and MAP into consideration.

I want to do something like the Vaporworx setup, and set it up to completely eliminate the return line, regulator, hobbs for 2nd pump, and check valves for when I run 1 pump at a time. Maybe at first I'll set it up to do closed loop static pressure, and then I'll put a MAP sensor 5v signal on it so I can adjust pressure based on vacuum/boost. The good thing about that, I can do whatever formula I want. So if I want 30 PSI at idle, 60 PSI at 100 KPA, and then 2 PSI per pound of boost, no problem.

 

Ford has been doing PWM fuel pumps for a long time. Lots of foreign car companies have been doing PWM for a long time. I always though they had a tank mounted regulator like most do, but there is zero regulator in the system at all.

 

I dont think it's that common really, especially as some modern stuff is now moving to direct injection.

The Gen5 stuff use a PWM setup ( which is where Vaporworx base there's from ? ), but other than that, I cant think of any other cars that run a dead end, PWM pump control and no form of mechanical regulation.

 

No, they have a 'pressure relief valve'.
Simple valve with a very high opening pressure that bleeds off in case of some sort of failure. For example, Audi/VW pressure limiting valve that goes in between the fuel line and rail:



Examples of electronic returnless systems:


Look up Ford ERFS. There is no mechanical pressure regulation.
http://www.searchautoparts.com/motor...ssure?page=0,0

 

Actually a VERY simple system, mechanically!

 

Getting it to hold ~50 PSI at idle with a pair of AEM 380s will be the real task. I'm going to give it a shot.

 

Hi Joe, your "tech" is fine though ONLY SOME Motors will operate correctly using PWM of current flow.
There ARE MANY Motors that will NOT AND run very hot.
The test method is easy, just connect a scope and view the wave form when operated.

Thus your answer.

The NEXT "tech" is cost, the cost of a Solid State Relay (35 AMP) is under $50.00.
The signal can be a GPOI from the EMS/PCM.
I supply a PWM controller for $200.00 with software GUI. (2.5, 4, 7 Bar)

I always design with a "vapor return" back to the tank(s)

Lance, BTW = Sticky for Fuel Pumps that can use PWM ?

 

I'm sure almost every pump can accept a PWM signal.....however, how many of those pumps will be happy at very low duty cycles is another matter

I suspect not too many.

 

i wonder how fast the response is when taking into account the damping in the lines, even if hard lines were used.

 

The controller and SSR I'll be using will have 1ms response. As far as actual pressure response, I'm not sure. I can set one up and try it but I don't want to do it on one of my turbo engines. I've tested it on my 4.8 and the fuel pressure stayed at a steady 60 PSI at idle but I never hit the throttle.

 

There's no reason why response shouldnt be very good, and even if there was a little lag, the entire system would be tuned around it anyway.

Again....all a fuel system has to be is consistent and repeatable. As long as behaviour when tuned remains the same....ultimately any slight oddities here and there arent really a problem.

But an SSR is not an ideal candidate, most pumps/motors will want in the 10-20KHz range for this sort of work..whereas most SSR's cannot even do 1KHz
Although many are using cheap SSR's at low frequencies for a basic PWM'd pump and seem to be getting away with it. No idea how they'll do long term ( or the pumps too )

 

Someone on the Holley support forum did some scope work on the C6 PWM fan controller. It is not suitable for fuel pump use, because it has a slow response time, which is great for the cooling fan. The reason I am talking about it here is that in this analysis, the person determined that it was actually some sort of frequency amplifier and operated in the frequency range that your mentioning above.

Andrew

 

so take this (or any of the ohers out there):

https://www.ebay.com/itm/PWM-30A-Mot...cAAOSwKUhZsgco

and solder up something to make it work with the fp sensor?

 

For a fan, it wouldnt surprise me if it has "slowness" built in, after all, that would be fairly standard for any big heavy current motor....simply so it doesnt suddenly pull a ton of current. And a cooling fan has no need for high speed changes. So it's probably limited so the controller of say 30A....will never actually pull more than 30A even during transients, even though the motor itself could happily pull more if allowed to

A fuel pump motor would have much much less inertia and be far more able to perform faster changes in speed with less current loadings to do so than a fan motor.

It'd be interesting to know how much current is passing through the Tesla motors when they perform hard launches...because they launch bloody hard !

Again when the topic comes up, always worth referring back to this thread from a few years back where "Max_Torque" built his own controller. Although his technical abilities would be a little higher than most, towards the end though he makes reference to a suitable PWM controller fitted to some Range Rovers or something to save building an entire unit from scratch.

https://www.pistonheads.com/gassing/...&f=66&t=916550

 

Here is a scope I did years ago, first hit is simply key on, monitoring ignition voltage ( pump voltage ) and also pump current for my 2x044's upon startup and idle. It doesnt let me alter sidescales for different traces, but I know both pumps draw around 22A at idle.

Immediately on key on, there is a large current spike, no doubt this is what OEM PWM Fan controllers will be avoiding at all times so they can run more powerful fans without such side effects.
Same with any industrial motor these days which are all inverter controlled. So current is always well monitored and pump speed etc controlled.