Really strange idea of the day!!!

[1project2many]

I wonder if MTBF for tbi fuel pumps is in on/off cycles, or total run time?

On/off cycles are minimal for most pumps compared to the total number of hours of operation. And start/stop cycles do little to cause wear to the pump. Consistently, pumps fail when the copper contacts on the armature wear through. IME the MTBF is directly related to the amount of fuel pumped vs the amount of run time. IOW running the pump dry or with low fuel level will take it out of service much faster than ensuring fuel is available to cool and lubricate the armature and the brush/armature interface. I have also noticed that almost all manufacturers except GM tend to have large intervals between failures. I have Ford and Dodge vehicles with 300k miles which have never received pumps, while most of the GM's have used as many as four pumps in the same time.

MSD?? or maybe Holley? has a pump control module which steps up voltage to the fuel pump to create higher than stock pressure and volume. I've worked with vehicles using this system although ultimately I always try to remove it and use proper tuning to make the engine run correctly. There are quite a few stories around about these modules causing premature pump failure but I haven't left one in place long enough to know.

Brushed DC motor controllers aren't new, of course. There are quite a few guys using PWM controllers on brushed motors in Power Wheels type vehicles (yes, the kids toy). Soft starts as well as variable speed control are usually the goals here. These motors usually get abused once Dad is done "fixing up" the toy and if they burned out quickly after modifications I'd be willing to bet there would be a lot of discussion on the forums.

You know, there is another option that might cause less worry and effort. The vacuum regulator is a very nice analog pressure control device. Why not PWM the signal to the regulator? This allows you to use the stock fuel system without worry about purging the vapors which will inevetably form, or purging air after a filter change or extended storage, or worry about deadheading the pump. A small vacuum pump and reservoir could ensure a constant supply of vacuum. An EGR control solenoid already designed to PWM a vacuum signal would be a perfect match for the control system, and the existing EGR control code and ecm output could be repurposed for the FPR. And really, the system could even be connected to pressure rather than vacuum, which would allow increases in pressure above base regulator settings vs reductions below max. Just a thought.

[JeepsAndGuns]

You know, there is another option that might cause less worry and effort. The vacuum regulator is a very nice analog pressure control device. Why not PWM the signal to the regulator? This allows you to use the stock fuel system without worry about purging the vapors which will inevetably form, or purging air after a filter change or extended storage, or worry about deadheading the pump. A small vacuum pump and reservoir could ensure a constant supply of vacuum. An EGR control solenoid already designed to PWM a vacuum signal would be a perfect match for the control system, and the existing EGR control code and ecm output could be repurposed for the FPR. And really, the system could even be connected to pressure rather than vacuum, which would allow increases in pressure above base regulator settings vs reductions below max. Just a thought.

I've been reading this thread with a little intrest. I think you might be onto something there. Find a vac refrenced pressure regulator and re use/re purpose a part of the bin thats already there. Use the egr solenoid to controll the vac to the regulator. Very interesting.

[EagleMark]

No, I just answered it awkwardly. What I meant was I don't think a tbi/tpi pump will work PW switched or voltage varied. Then you come along with a controller that is specific to fuel pumps in automotive applications. I think I'm officially confused right now, probably from lack of sleep. I'm off to bed for the night, Later.

Surely was not trying to confuse you!

Just trying to gain some insight to a worthy project. If I knew the answer I would tell!

The fuel pump is a DC motor, it will take varied voltage just like an RC car motor will. But switching it on off repeatedly to get the effect is what will kill it. I think that is where the controller comes in. Seems it's part of the cars and diesel trucks I have found with same system. The one above is for home made. There is a link above that to jbperf that has one cheap, they make other things for megasquirt like the JimStim.

[jim_in_dorris]

Mark, did you read the PWM controller pdf you posted? It seems to indicate that PWM controlled DC motors will last longer than voltage controlled DC motors. That circuit was specifically designed to control a fuel pump in a boosted application. The article even claims that most DC motor manufacturers recommend PWM control rather than voltage control.

I'm still trying to get my head around this problem. I also need to keep in mind that I can run 2 80# 454 injectors off a simple vacuum referenced pressure regulator and be done with it. I just worry about big injectors even at low pressures having trouble at the bottom end where I don't need as much fuel.

[gregs78cam]

I'm still trying to get my head around this problem. I also need to keep in mind that I can run 2 80# 454 injectors off a simple vacuum referenced pressure regulator and be done with it. I just worry about big injectors even at low pressures having trouble at the bottom end where I don't need as much fuel.

I do like this thread's whole idea, but......
What engine is this going on again? I have 4 61lb injectors feeding my 383 and @13psi, I can idle it at around 15:1. I could even swap one pair to 45lb injectors and still have plenty at the top end.

[EagleMark]

Mark, did you read the PWM controller pdf you posted? It seems to indicate that PWM controlled DC motors will last longer than voltage controlled DC motors. That circuit was specifically designed to control a fuel pump in a boosted application. The article even claims that most DC motor manufacturers recommend PWM control rather than voltage control.

I'm still trying to get my head around this problem. I also need to keep in mind that I can run 2 80# 454 injectors off a simple vacuum referenced pressure regulator and be done with it. I just worry about big injectors even at low pressures having trouble at the bottom end where I don't need as much fuel.

Yes I have and a bunch more, like you I am still trying to get the entire concept in order.

Back to your engine, 400 HP can be done and tuned with BB injectors at around 15PSI without vacuum referenced regulater...

But I don't want to kill a sweet idea! If/when this works a VE table would be a fairly straight line?

[jim_in_dorris]

Mark
Using the formula's from the sticky on getting enough fuel from the "Performance" forum on fullsizechevy, I need 103#/hr to feed 350hp. with an 80# injector (rated I think at 13psi) i need 22psi to feed the beast. If I wasn't in California (home of the Smog Nazi's) I would have an easy time feeding this engine using an mpfi system or 4 bbl TBI.

I still like this approach, and can get a take-out vortech fuel pump (not delivering enough pressure) to use as a test mule to see what happens, then build the PWM circuit and actually test the pump to see how it handles it.

[EagleMark]

Mark
Using the formula's from the sticky on getting enough fuel from the "Performance" forum on fullsizechevy, I need 103#/hr to feed 350hp. with an 80# injector (rated I think at 13psi) i need 22psi to feed the beast. If I wasn't in California (home of the Smog Nazi's) I would have an easy time feeding this engine using an mpfi system or 4 bbl TBI.

The injector sizing spreadsheet I have gives 80# 15PSI BSFC .5 90% injector duty cycle 309 HP. But that is at 108 BPW. Stock BPW is 135 I beleive, I ended up at 130. The Injector sizing calulater I have is not set to adjust BPW... it's the numbers I used for 398 HP motor on Dyno so it worked with WideBand showing 12.7 to 1 AFR from 4000 to 6200 RPM.

[jim_in_dorris]

Well, I think I know the "fly in the ointment". If I am thinking clearly, the problem is going to be at the injector end of things. I think the injectors would not work properly. Let me try to explain. At idle, the pump would cycle on once delivering 34 psi to the injector (remember that the regulator caps it at 34 psi) followed by 2 off pulses. Depending on the pulse width of the pump, the injectors could see 34 psi dribbling off to nothing, all the way to opening with no fuel and spiking at 34 psi. PWM of pumps may work for modern systems designed to deal with a dead-head system, but I think that this is the deal breaker. Guess I need to learn more about vacuum referenced regulators. Mark, How did that 398 hp motor idle with the 80# injectors set at 15 psi with a bpw of 130?

[EagleMark]

Fine of course! I tuned it!

It takes a little work which is easy with Emulater, things like raising idle RPM, idle timing, making min air adjustments so there's still a little IAC control, while engine gets to temp while watching a vacuum gauge, then rinse and repeat when warmed up. IIRC this one ended at 850 RPM and 14-15 inches of vacuum and under 10 IAC counts. Started with not being able to idle and vacuum of around 10-11 with a screw driver tip in stop screw to raise idle so I could work on tune.

From there it's tune as usual.

[EagleMark]

Well, I think I know the "fly in the ointment". If I am thinking clearly, the problem is going to be at the injector end of things. I think the injectors would not work properly. Let me try to explain. At idle, the pump would cycle on once delivering 34 psi to the injector (remember that the regulator caps it at 34 psi) followed by 2 off pulses. Depending on the pulse width of the pump, the injectors could see 34 psi dribbling off to nothing, all the way to opening with no fuel and spiking at 34 psi. PWM of pumps may work for modern systems designed to deal with a dead-head system, but I think that this is the deal breaker. Guess I need to learn more about vacuum referenced regulators.

I think that is the purpose of the module between the PWM regulater and fuel pump, to keep a steady voltage? It's be nice to test one...

[gregs78cam]

Well, I think I know the "fly in the ointment". If I am thinking clearly, the problem is going to be at the injector end of things. I think the injectors would not work properly. Let me try to explain. At idle, the pump would cycle on once delivering 34 psi to the injector (remember that the regulator caps it at 34 psi) followed by 2 off pulses. Depending on the pulse width of the pump, the injectors could see 34 psi dribbling off to nothing, all the way to opening with no fuel and spiking at 34 psi. PWM of pumps may work for modern systems designed to deal with a dead-head system, but I think that this is the deal breaker. Guess I need to learn more about vacuum referenced regulators. Mark, How did that 398 hp motor idle with the 80# injectors set at 15 psi with a bpw of 130?

Actually the on-off action is so fast that the only thing you would notice would be the pump would run slower, there won't be any pulsing of the fuel pressure at the injectors, your fuel lines and hoses will act somewhat like an accumaulator, and buffer any high frequency pulsing in the pressure.

[jim_in_dorris]

It may still be possible to figure out a variable voltage driver to do the trick, just trick the ECM into not seeing the voltage changes. Just have to figure out what pump voltage = what pressure.

[EagleMark]

There's acually 2 kinds of pumps as well. The vane drive the Carter makes does not pulsate like whatever kind the AC/Delco part is...

[RobertISaar]

It may still be possible to figure out a variable voltage driver to do the trick, just trick the ECM into not seeing the voltage changes. Just have to figure out what pump voltage = what pressure.

resulting pressure will be considerably modified by the flow of the injectors.

and as long as the cycle is fast enough (64Hz or higher should be plenty, i would think), then the injectors would never know the difference.