Erik Gross method revisited

[RealMcCoy]

We were asked to start migrating content here... This is my first contribution. Originally from a couple years ago, but I'm still using it and very happy with the results.
--------------------------------------------------------------------------

I made a post about this on 3Si that sparked some interesting discussion, but it was never intended to be a case study/searchable resource... More of a "woot, look what I found" thread..

Cliff notes of other post:

1: I install Supra pump in my car.
2: I think I'm really smart wiring in the stock resistor to keep from overrunning my stock FPR..
3: I'm not smart enough to search, and discover the pitfalls of this plan until AFTER I experience them myself..
4: I'm bent because it "should" have worked.. and go on a mission to find out why it didn't..

Now there are several methods of wiring a fuel pump that are attributed to Erik's work... For the purposes of this discussion, it is solely the practice of wiring in the stock fuel pump resistor to slow down a hotwired pump at slow engine speeds.. My entire fuel system is wired as standalone from the stock wiring. I used two relays signaled from the MPFI and the ECU.. Even the resistor is fed from the hotwire. Here's a diagram of my system:

http://i90.photobucket.com/albums/k252/lnmccoy/fuelSys.jpg

From reading the old threads, Erik gave up on the idea because there was an obvious fuel pressure variation between the low and high speeds that screwed with the fuel trims and would cause drivability issues like lean spots right before the transfer... These are the same symptoms I experienced.

Many of you will say "why bother.. Put in and aftermarket FPR" And you would be justified in that position... But I'm far too stubborn to let a solution lay undiscovered. And I had a pretty solid theory about what was going on. (and wasn't about to take a couple hundred bucks away from the white car build if I didn't have to)

So here is the Data I gathered: (bench tests w/stock regulator)

The stock pump hotwired at approx 13.6Volts, draws 10 amps making 45 psi (w/no vacuum)
The Supra pump under the same conditions draws 14 amps and makes 50 psi(overrunning the FPR)

Deadheaded,(up against the internal bypass) the stock pump draws 12 amps making 90 psi.
Under the same conditions the Supra pump draws 20 amps making 100 psi.

Conclusion: the supra pump is a huge power hog...

On to my car... Why are we getting the pressure fluctuations?

KOEO, battery charger maintaining 13.6v, jumper relay to hotwire mode: The installed Supra pump screaming along at 7500rpm, drawing 14-15 amps, and overrunning the FPR...

http://i90.photobucket.com/albums/k252/lnmccoy/hiSpd2.jpg

Jumper the resistor on line: The supra pump gets the wind knocked out of it.. Slows down to 3500rpm, drawing 7.5 amps... At this speed it will make less than 50 psi deadheaded... Starting the car, I discovered that on a snap throttle the pressure would drop, rather than go up like it should. The Supra pump simply could not supply enough fuel to run the car while turning that slow..

http://i90.photobucket.com/albums/k252/lnmccoy/slw1.jpg

Conclusion: Too much resistor for that pump...

Now what to do about it..? I already had a plan... I wired a second stock resistor in parallel to the original, cutting the resistance in half.

http://i90.photobucket.com/albums/k252/lnmccoy/rstr1.jpg
http://i90.photobucket.com/albums/k252/lnmccoy/rstr2.jpg

Results:

5000 rpms, 10amps, normal fuel pressure(no overrun), and able to make 80 psi deadheaded... WOOT! I win!

http://i90.photobucket.com/albums/k252/lnmccoy/med.jpg

I've been driving it around for a week looking for any signs of fuel control or pressure issues. I have to call this one a fix..

Thanks for reading..

Larry

[Austin@STM]

Very interesting work.

[HydeDW]

Well this is a timely Post! I have a Supra Pump dropped in with all Stock wiring (no hot wire). I'll keep eye out for this symptom.
Thx
Dan
97VR4

[RealMcCoy]

Crap... I just noticed the typo butchering Erik's name in the title. Need to find a Mod...

[Greg E]

Great work! Can you build me one? :)

[Roybatty]

this needs to go in a knowledge database section

[IPD]

so the moral here is, AFPR?

[CoopKill]

The moral is, no need for the AFPR.

[KeithMac]

Good post by Larry, I recently re-wired my Twin Walbros in parrallel with 2 resistors and the car dove terribly with the resistors getting red hot ( odd as I had no problem with a single Walbro in 2 speed mode).

Gone 2 pumps full speed at the moment but the noise is annoying!.

[RealMcCoy]

Great work! Can you build me one? :)

I'd be much more willing to help you do it yourself, but if you really wanted me to build it, I suppose it could be arranged...


so the moral here is, AFPR?

That's what I usually recommend, but there are downsides... If you drive your car on long trips and lots of highway use, slowing the pump down will significantly prolong it's life. There is also the all too common complaint of the aftermarket regulators leaking, causing long crank times.


Good post by Larry, I recently re-wired my Twin Walbros in parallel with 2 resistors and the car dove terribly with the resistors getting red hot ( odd as I had no problem with a single Walbro in 2 speed mode).

Gone 2 pumps full speed at the moment but the noise is annoying!.

You have any way to test the current draw on those things?

The nature of an electric motor is that it will increase it's current draw as it's slowed down. If you bring it to a full stop, it's a dead short. This is why you will see starters and battery cables turned to bacon from a depleted battery, it goes against all logic until you examine the nature of a segmented armature electric motor.

In my above data, those resistors were only allowing 7.5 amps each through them with the pump slowed down to under half normal speed. It sounds like two resistors is still too much for dual pumps.

The Walbro has significantly less current draw under normal operation than a Denso, but I would theorize that it's much less tolerant to being slowed down. The Denso has 12 armature segments, while the Walbro has only 8. You get much less leeway for speed changes before you approach dead short status.

In your situation, I'd be tempted to experiment with an injector resistor pack... You'd have 6 separate segments you could run in parallel one at a time until you found the sweet spot.

[Greg E]

I'd be much more willing to help you do it yourself, but if you really wanted me to build it, I suppose it could be arranged...

Haha, I can make one. If you could make up a wire schematic, I'd greatly appreciate it! Where can I get the relays from?

[RealMcCoy]

I already posted a diagram of my system:

http://i90.photobucket.com/albums/k252/lnmccoy/fuelSys.jpg

Let me try and clarify what it is we're doing and what the diagram represents...

The relays I used were standard Bosch style relays. You can pick them up just about anywhere. They are all plainly labeled with the pin #'s 30, 87, 87a, 85, and 86.

Pins 85 and 86 are the signal circuit, when current flows through them, the relay is engaged. Pin 30 is the common high current post, when the relay is not engaged, it is connected to pin 87a, when engaged, it's connected to pin 87.

The circuit labeled "MPFI" is the fuel pump power source from the factory MPFI relay. We're using that to turn on the hotwire relays.
The circuit labeled ECU is the factory signal from the ECU to engage the resisted mode. The ECU grounds this circuit to engage the relay. We're pretty much duplicating the factory setup here.

So what we do is feed power from the MPFI relay to pin 85 on BOTH our relays, the hotwire relay pin 86 is grounded, so it automatically engages whenever the MPFI is powered up. the resistor relay pin 86 is grounded by the ECU, so it will engage when the ECU switches to non-resisted mode. The power out of the hotwire relay is split into two parallel circuits, one resisted, one not... The resisted leg is hooked to pin 87a of the resistor relay so that any time the relay is NOT engaged, you have resisted power. The non-resisted side is hooked to pin 87 so that when the ECU engages the relay, you have non-resisted power out pin 30.


This is the factory system we're replacing:


http://i90.photobucket.com/albums/k252/lnmccoy/ForumPics/92TTFPR.gif


Let me know if I can clarify anything further...

[RealMcCoy]

Thank you to whoever fixed the thread title...

After several edits and second guessing my own thoughts, I think I finally clarified my clarification.... LOL

[green-lantern]

Good read, I admire the stubborn never let it beat you attitude that drives you to do the research. When I want to know something or why I got the results I did I pretty much won’t quit until I get the answer.

Good Job!

[i3igpete]

The moral is, no need for the AFPR.

the moral of the story is, twin stock FPR's... :suspect:

[green-lantern]

And two throttle bodies :suspect:

[CoreyB]

Dont listen to them, you will break a ringland.

[aaronatstate]

the moral of the story is, twin stock FPR's... :suspect:

GTFO now :p Really we just need ram air, that way we'll make boost with an N/A!

[mb7050]

I was thinking of doing something similar because I hated the noise that walbro makes when hotwired but then I bought a denso fuel pump and an afpr ..

[KeithMac]

I'd be much more willing to help you do it yourself, but if you really wanted me to build it, I suppose it could be arranged...



That's what I usually recommend, but there are downsides... If you drive your car on long trips and lots of highway use, slowing the pump down will significantly prolong it's life. There is also the all too common complaint of the aftermarket regulators leaking, causing long crank times.



You have any way to test the current draw on those things?

The nature of an electric motor is that it will increase it's current draw as it's slowed down. If you bring it to a full stop, it's a dead short. This is why you will see starters and battery cables turned to bacon from a depleted battery, it goes against all logic until you examine the nature of a segmented armature electric motor.

In my above data, those resistors were only allowing 7.5 amps each through them with the pump slowed down to under half normal speed. It sounds like two resistors is still too much for dual pumps.

The Walbro has significantly less current draw under normal operation than a Denso, but I would theorize that it's much less tolerant to being slowed down. The Denso has 12 armature segments, while the Walbro has only 8. You get much less leeway for speed changes before you approach dead short status.

In your situation, I'd be tempted to experiment with an injector resistor pack... You'd have 6 separate segments you could run in parallel one at a time until you found the sweet spot.

I had 2 Walbros in parallel powered by 2 Resistor packs in parrallel, I have another Resistor so I could try 3 in parrallel I suppose.

Looked at the injector resistor pack but it doesn`t look substantial enough compared to the pump ones?.

It never dawned on me that they would inrease current draw as they slow down but it makes sense, have split many motors down before for repair so I`ve got a good understanding of what`s inside.

[TwIzTeD_3kGt]

This is cool stuff. Kudos OP.

[mb7050]

:bump:


Have you noticed any issues yet ? Im thinking of trying a single walbro 450 with dual stock resistors. I'm really trying to avoid going the afpr route this time.

[RealMcCoy]

:bump:


Have you noticed any issues yet ? Im thinking of trying a single walbro 450 with dual stock resistors. I'm really trying to avoid going the afpr route this time.

I've had it in my car for over three years, and probably somewhere around 20K miles... I'm running 550's now, still works great.

The only concern I ever had was the switch point climbing up with piggyback airflow adjustments... That hasn't been an issue with up to 550's at this point. If you had a clone ECU and Greg's software you could adjust it anyway....

I've got the same system in a Walbro 255 equipped car with 450's works great in that one as well.

[DG]

I'm bent because it "should" have worked.. and go on a mission to find out why it didn't..

This is the way I found 90% of the stuff I learned working on race cars. "Huh, that's funny" coupled to hard, objective data obtained during an investigation ultimately produces results.

I wish more people worked like that. Good job.

Everybody else: could we maybe stop with the jokes and inside references? This isn't Reddit; the noise is drowning out the signal.

DG

[CoopKill]

This is the way I found 90% of the stuff I learned working on race cars. "Huh, that's funny" coupled to hard, objective data obtained during an investigation ultimately produces results.

I wish more people worked like that. Good job.

Everybody else: could we maybe stop with the jokes and inside references? This isn't Reddit; the noise is drowning out the signal.

DG
LOL, DG you are yelling about comments made well over a year ago...

[DG]

Still applicable.

And how does your comment add value?

DG

[CoopKill]

About as well as yours I guess?

[pat3000gt]

I confused why not go to the AFPR. I have a supra pump hot wired and a AFPR, did not see any pressure issues?

[Greg E]

I confused why not go to the AFPR. I have a supra pump hot wired and a AFPR, did not see any pressure issues?

Because its cheaper and easier to rewire the fuel pump the way Mitsubishi intended.

[mb7050]

because the afprs have a tendency to leak down very fast and causing longer cranking times.

[anyonebutme]

because the afprs have a tendency to leak down very fast and causing longer cranking times.

:|

thought I just covered that

[RealMcCoy]

:|

thought I just covered that

Is this an inside conversation the rest of us aren't privy to? You haven't covered anything in this thread... This was your first post here.

[mb7050]

:|

thought I just covered that

I´m still very very sure its the afpr that is causing the longer cranking not the fuel pump etc because when I switched back to the oem one no more long cranking.

[Greg E]

I´m still very very sure its the afpr that is causing the longer cranking not the fuel pump etc because when I switched back to the oem one no more long cranking.

This was the main reason for the "fuel pump on clutch" chrome mod. The ECU will activate the pump when you push the clutch instead of waiting for the engine to crank.

[anyonebutme]

Is this an inside conversation the rest of us aren't privy to? You haven't covered anything in this thread... This was your first post here.

Yes, it was an inside conversation.

[Chris@Rvengeperformance]

I´m still very very sure its the afpr that is causing the longer cranking not the fuel pump etc because when I switched back to the oem one no more long cranking.

I had the same experience

[RealMcCoy]

I´m still very very sure its the afpr that is causing the longer cranking not the fuel pump etc because when I switched back to the oem one no more long cranking.

That's a fairly common complaint... Easily diagnosed with a pressure gauge and a pair of pliers...


This was the main reason for the "fuel pump on clutch" chrome mod. The ECU will activate the pump when you push the clutch instead of waiting for the engine to crank.

I like... Very cool.

[J. Fast]

This was the main reason for the "fuel pump on clutch" chrome mod. The ECU will activate the pump when you push the clutch instead of waiting for the engine to crank.

Please point me to discussion. I like this idea!

[Greg E]

Please point me to discussion. I like this idea!

No discussion. I had the idea, wrote the code and made it happen. :)

[vr4tune]

Can that be done on aem? (Fuel pump on with clutch)

[sergechronos]

In my above data, those resistors were only allowing 7.5 amps each through them with the pump slowed down to under half normal speed. It sounds like two resistors is still too much for dual pumps.

The Walbro has significantly less current draw under normal operation than a Denso, but I would theorize that it's much less tolerant to being slowed down. The Denso has 12 armature segments, while the Walbro has only 8. You get much less leeway for speed changes before you approach dead short status.

Not to renecro a mostly dead thread, (although mostly dead is still partially alive), but given that you said the Walbro has less current draw and is likely less tolerant to slow downs (in regards to dual Walbro), would this mean that the dual resistor idea may not work near as well for a single Walbro as it did for the Denso? Or should it still work fine?

I've been running a Walbro single but had yet to hotwire it as I was still running stock FPR, Injectors and Turbos. Now I'm going to be going to 560cc, 13ts and replacing the FPR (most likely with another factory FPR if this works), and would rather do dual resistors than hotwire and AFPR.

[RealMcCoy]

Not to renecro a mostly dead thread, (although mostly dead is still partially alive), but given that you said the Walbro has less current draw and is likely less tolerant to slow downs (in regards to dual Walbro), would this mean that the dual resistor idea may not work near as well for a single Walbro as it did for the Denso? Or should it still work fine?

I've been running a Walbro single but had yet to hotwire it as I was still running stock FPR, Injectors and Turbos. Now I'm going to be going to 560cc, 13ts and replacing the FPR (most likely with another factory FPR if this works), and would rather do dual resistors than hotwire and AFPR.

Several people have done this with a single walbro, with no issues that I've been made aware of... I just did it on a wally equipped car last month, and it seems to be working just fine.

[importrunner]

Would this only be done with a aftermarket pump? I'm strictly bpu and was going to hotwire my pump just wanted clarify.

Sent from a wannabe ipad

[RealMcCoy]

Would this only be done with a aftermarket pump? I'm strictly bpu and was going to hotwire my pump just wanted clarify.

Sent from a wannabe ipad

I've never seen a hotwired stock pump overrun the regulator, so I would consider it mostly unnecessary. If you did want to mimic the stock system, The factory resistor would be fine for the stock pump. It just can't flow enough current to run the bigger pumps.