[P-51]

Well, I pretty much spent all weekend playing with my car to increase the fuel pressure and I got mixed results.

First, I installed the fuel pressure gauge. This went fine, and it was a good thing that Jim gave me a 5th injector setup cause the plate for the fuel rail came in really handy to hook up the pressure gauge.

It took until about noon Saturday to get that up and running. I installed the gauge where the coin tray used to be. It will be a nice setup once I finish it.

I tested this, to see how it runs stock. It runs about 30psi at idle, 42psi at atmospheric (lugging the engine at 1500rpm before boost hits), and about 45 psi under boost. The pressure sensor is referenced to manifold pressure, that is how this happens.

I also hooked up the pressure switch to the manifold.

Next, I started trying to increase the fuel pressure.

I was splicing into the wires of the computer by finding the right wire (Pin 63) for the signal from the fuel pressure sensor, and cut it then put connectors on the ends. The circuit connects to the connectors, and I can easily return to stock by taking out the circuit, and reconnecting the wires by plugging the connectors together. (Do I get an award for most use of the word "connect" in a paragraph?)

I get this all going, but I broke a pin off my electronic relay, about 6pm (after all the stores close!). So I hook it up with the fuel pressure working 100% of the time, and the switch hooked up to the LED just to see how the switch works.

Turns out the switch works well, the LED comes on under boost.

BUT, the resistor mod DOES NOT WORK by interrupting the signal from the fuel pressure sensor. I started at 0 Ohms, and as I start cranking in resistance, I watch the pressure GO DOWN!!!

This is not the right wire to use. I don't understand the circuit inside the pressure sensor. I thought it was a variable resistor, but it is not. I think it is a Wheatstone Bridge or something like that. I don't understand those. It's a damn good thing I installed the fuel pressure gauge first! Otherwise I'd be in trouble.

Sunday morning, I go and splice into the yellow with green stripe wire from the fuel pressure sensor, and work the resistor, and VOILA that increases the fuel pressure.

I get a new relay from Radio Shack, and wire that up.

Now, I have it working as planned. The boost pressure closes the pressure switch, which trips the relay which puts the resistor into the circuit and turns on the LED. The higher the resistance the higher the pressure. I had to go up to around 450 ohms to get about 70psi under boost. It is stock pressure at cruise and idle.

Sweet.

So I'm trying it out, and it works great. The pressure switch activates great, the only problem is it "flutters" when it is just kicking in. Because their are pressure pulses in the intake manifold. I think I can damp this out with a capacitor.

Anyway, I can run at a full 4-5psi at 70psi of fuel pressure with NO DETONATION. Well, there is the odd <tick> but nothing to worry about.

BUT...

This morning, the whole thing started acting wonky.

The idle fuel pressure was 40psi. And when I get in the boost, the circuit is working but the pressure sometimes doesn't go over 50!

It's kinda sometimes it works, sometimes it doesn't. I don't get it.

For a while I was getting only 50psi at full boost. Then, I turned the resistance to zero, and floored it. All the fuel pressures go back to normal. I turn the resistance back up, and the fuel pressure goes up and no detonation again...

I think the computer somehow "figures out" what is going on, and is compensating by turning the pressure. But how could it...?

I'm going to check my circuit tonight. Make sure nothing is wrong. Then, I'll try the fifth injector setup.

I hope that works, cause it's pretty frustrating.

 

[P-51]

What is the best way to reset the computer?

Apparently, you can pull some fuse, and step on the brakes or turn on the headlights?

 

[P-51]

I'm at work, so I wanna know if there is a way to reset it without needing any tools... some people have mentioned a fuse or something.

I was thinking about that, wether or not it could have happened on start-up. It's possible, but not likely. It really needs to be in boost to work. And the pressure was fine when I left home. It was after a few pulls that it started going wonky.

I was thinking about the second point of reference too. There is nothing that I can think of, other than the fuel tank pressure transducer. But I don't see how that would have a relationship to fuel rail pressure...

Maybe there is a table with fuel injector duty cycle vs. fuel pump duty cycle and it sees that it is "out of bounds" or something?

 

[P-51]

I think I found the problem.

I looked under the hood, and saw that one of the wires to the splice from the fuel pressure sensor came out of the connector.

So, essentially it didn't have any fuel pressure reading....

Yeah, that could cause a problem!

I don't know how long that condition existed, so I'm not sure how much of the problem is from that.

But in hindsight, the computer was doing a bang-up job of regulating fuel pressure with no signal!

I'll keep everyone posted on how well it works over the next couple of days. But if that was the only problem, this will be a great way to give a little more fuel to the car.

After I made the repair, it seems to be working properly again.

PowerFocus:

I don't think making the fuel pump go 100% is a good idea. It will send fuel pressures through the roof!

The only other thing wrong with this circuit is that as it is crossing the pressure threshold where the pressure switch turns on the relay, there is that flutter. Is there any way to dampen that out electronically? Or do I have to do with an orifice on the pressure line?

Could I connect a capacitor across the terminals of the relay? When the switch comes on at first, it charges the cap. Then, ~1/20th of a second later when it the pressure switch disconnects, the cap discharges and can supply enough current to keep the relay open until the pressure switch reconnects again?

I think the switch flutters with a frequency of about 30-60Hz, for about 1/4 second. I want it so when it's on, it's on.

 

[P-51]

Seems to be working pretty well. There is some funny stuff still, but I don't think it's a problem. I have to cross the border everyday, and while waiting in line idling for a hour, the fuel pressure goes up to 40psi at idle. Returns to normal once I start driving. I'll find out later if that is normal operation.

Also, when the relay is flickering, it makes the speedometer jump a bit, but nothing major. It will go away once I stop the flickering.

Other than that it works pretty good.

Tom:

Yes, you could do that, but this setup costs $50, and is much easier to perform. These are just a few development bugs to work out.

 

[P-51]

So I would just put it right across the two terminals?

The guy at the electronics store suggested I attach one end to the 12V feed, and the other to ground.

 

[P-51]

Great page, I'll review that at home.

You electrical dudes are a great help!

 

[P-51]

Ok, the pressure in the intake manifold closes a pressure activated switch.

That switch, turns on the current to a relay which activates my circuit.

The problem is the pressure waves in the intake manifold make the switch flutter, and as the pressure is passing over the threshold where the switch will turn on, the pressure pulse make the switch flutter instead of coming on clean. So I need to get rid of those pulses.

As I said, the pulse frequency seems to be about 20-50 Hz. I think the relay has a resistance of about 200 Ohms, but I need to verify.

I need a capacitor or something that can either A)absorb or stop the current from the switch from activating the relay until the current from the switch is on full. Or B) keep the relay switched on while the pulses take place, yet won't hold the circuit open after pressure is gone (so a flip-flop won't work).

I forget the math behind how a capacitor works... If the first spike is only a few hundreds of a second long, will it charge enough to be able to supply the power to keep the relay open for another few hundreds of second? I don't know how fast they charge... I imagine if there is no resistance in the line going to the capacitor, it will charge fast, and then when the switch opens (current stops) the charge will last longer because it's only way out of the cap is to go through the relay which has resistance?

This stuff is fun, but I always have to get a physical world analogy so I can understand.

 

[P-51]

I looked on the radioshack website to find out exactly which relay I got. I only know it by sight, and I couldn't find the exact one. It looked an aweful lot like this guy:
http://www.radioshack.com/product.asp?catalog%5Fname=CTLG&category%5Fname=CTLG%5F010%5F008%5F005%5F000&product%5Fid=900%2D5864

Except it's blue, and is a 12V relay. I think it was some sort of "PC relay", I guess that is for a computer power supply? Definitely has the same pin arrangement, it's a DPDT.

The pressure switch powers the coil on the relay. I have a 12V diode hooked up on one of the normally open switches. And the resistor is short circuited by the other pole, it's soldered to the normally closed relay.

I'm pretty sure it's the switch that is causing the problem, because the problem changes amplitude as I adjust the pressure that the switch activates at. Problem is if I keep reducing the activation pressure, it activates before I'm in boost.

Maybe I'll have to put the switch on the other side of the throttle body where it isn't affected by pulses from the pistons as much.

The speedometer does a little dance when the relay is buzzing, so I'm sure the computer isn't too happy about the situation either.

 

[P-51]

If one terminal of the coil on the relay is connected to the switched 12V, and the other terminal of the coil is connected to ground, then wouldn't connecting the Capacitor to:

A)one end connected to the switched 12V line and the other to ground (like Sandman's diagram)

B)one end connected to one terminal of the relay coil, and the other end connected to the other end of the relay coil.

Wouldn't that do the same thing?

If terminal A of the relay coil is connected to the switched 12V power, and terminal B is connected to ground, it wouldn't matter if the cap was connected to terminals A and B, or connected to switched 12V power and ground, the circuit would be the same.

 

[P-51]

Is the polarity clearly marked, or is it another one of these "inside jokes" you electrical dudes play on us mechs?

"Duh, the short lead on the LED is the positive one! You mechs are so dumb..."

 

[P-51]

Yeah, I'm aware of the possibility of fixing this with an orifice. But so far this can all be done electronically, and I'd like to keep it that way if possible. If a cap will work, then I'd rather do that.

So is the general consensus that I should go to the nearest electronics store, and buy a fistfull of caps and see what happens?

 

[P-51]

The electronics store near me has caps the size of pop cans... But I'll try to get a 1000microfarad one then. See what happens.

I'll just solder the negative side to the grounded pin on the relay, and the other side to the positive pin?

How bad is the explosion if I get it wrong? Like... would I lose a finger?

Is there ANY other risk here? The whole circuit is powered by a splice into the wire powering the ECU, and grounded the same. I think that is why it makes the speedo go wonky when it's buzzing. That nasty collapsing electromagnetic field stuff...I'm aware of that from my mech training.

But what are you saying about the cap making this worse?

[This message has been edited by P-51 (edited 10-03-2001).]

 

[P-51]

Yeah, it's probably a good idea to wire this circuit away from the CPU... I'll do that. Since it won't be making boost with the engine off, I don't need a switches source!

 

[P-51]

Is anybody able to figure out with some kind of simulation what size capacitor is optimum? What information would you need from me to do so?

 

[P-51]

Ok... I'll try to get you better values.

What happened to the initial thing you said saying I'd need .3F?

 

[P-51]

Ok, my electrical dudes:

I just picked up a 1000microF capacitor, rated at 25V. Which end is negative and positive? The guy said the end with metal showing is negative (chassis) and the end that is insulated is positive. Does that sound right?

They have a 22,000microF 16V cap for $10 there. If this doesn't work, should I make the jump to that? It's about the size of my thumb. Any risks?

I'm going to go ahead and move circuit to a better powre supply to. Right off the battery.

Lastly, wouldn't an inductor help me?

 

[P-51]

YOU GUYS ROCK!

The 1000microF cap works PERFECT. The flutter is COMPLETELY gone. It just comes on, and that's it.

There is the added bonus that the relay stays closed for about 1/4 second, so if you shift quick, the circuit stays and the pressure never drops.

Also, some kind of side benefit I guess, I think the computer likes this better. I used to have to run 400ohms to get the pressure I needed. Now, I'm able to turn the resistance down to 300ohms because the computer increases the pressure faster because the signal is a single point in time change. It was probably getting confused seeing the circuit resistance changing really fast back and forth.

The speedometer doesn't go wonky anymore either, so I don't think I'll bother moving the power wires.

Thanks a ton you guys!

I think this circuit is perfect.

It's kinda funny. Everybody hates the computer/electronic control of cars because you can't "play with it" as much. But once you understand the electric systems, this is actually better. This system is cheaper than a rising rate fuel pressure regulator you'd use with a return fuel system. And nobody with a RRFPR has ****pit adjustable pressure!

[This message has been edited by P-51 (edited 10-05-2001).]

 

[P-51]

Yes, the detonation is completely gone, and I have NO drivability issues at all. It runs like it was factory tuned.... let me rephrase that, it runs like it was Toyota factory tuned!

(we all know how badly some of the Foci run.

)

I can try and get that for you guys... but I had trouble last night. I tried to get it after I soldered the cap on, but it wouldn't take a reading. I imagine the cap was not compatible with my multimeter. Must have been absorbing the voltage so it couldn't measure Ohms. I'd have to unsolder the cap I think. If it's real important for you guys I can do it to return the favor.

If I wanted it to hold on for a bit longer, could I put a 2000microF cap on instead? Any performance advantage to using a 1000microF cap rated at 16V instead of this 24V one?

I have an A/F gauge, but you couldn't tune with it. It definitely reads richer though.

The delay is fine, and desirable I think. There is no stumble or anything, so it's fine.

I could make a kit, or maybe a shopping list and instructions for other people. I think Jim might include it in the St2 kit if that's what anybody is thinking about.

It would probably cost about $1000 all told. From what I can remember:

Pressure switch: $35
Adjustable Pot: $15
Dial: ~$8
Cap: ~$2
LED: ~$2
Connectors: ~$5
Project Box: ~$15 (waterproof, overkill)
Relay: ~$5
30 feet wire...

Other little bits.

The fuel pressure gauge was the expensive part, but if all our cars respond the same, not necessary.

I think adding it to a 5th injector probably isn't a good idea if you don't have detonation. However, I think 36lb injectors, with a recalibrated MAF and use this to richen up the mixture for anti detonation is a better plan than the 5th injector.

Another thing: On second thought, the frequency of the flutter was probably closer to 40hz than 10... But I don't think it was 60. It didn't hum like a transformer, it sounded more like... remember old hard drives (way back when they were 20-100 Megs) they used to max this noise when being accessed. A rattly kinda noise. It was like that.

 

[P-51]

Ok, so other than size, I wouldn't have a difference? I was wondering if it responds quicker or anything with the lower voltage rating.

I meant $100US. I'd have to figure it out if there was real interest. I think the only problem with the kit idea is the soldering. It's tricky to get right, if you'd never done it, you will probably screw everything up. (not being condescending... as I've screwed up PLENTY of soldering jobs!)

Do you guys think it would be better to have all of the connections in the engine bay, with a fixed resistor? Or everything in the ****pit with adjustable pressure and an LED light to show when it's activated.

Actually... the LED is mandatory to tune the kick-in pressure.

I'm sure the injectors will be fine. We're not talking astronomical pressure here. Max 80psi.

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>If I understand you correctly yes. A larger cap will hold the charge longer extending how long it takes to fully discharge.<HR></BLOCKQUOTE>

Any disadvantages? It might be better for the average Joe so you don't have to shift like "The Flash" to keep the pressure up.

Not that you have to keep the pressure up... I'm just being a prefectionist. Just if you were all hot and heavy drag racing in 100 degree weather, you might want the reasurance the pressure is there when the boost hits in the next gear. For driving around normally, it doens't make a difference, the pressure rised VERY fast. It goes from 40-60 in very short period of time, on the order of 1/10th of a second.

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>I assume that includes "shipping and handling" if you were making that kit huh?<HR></BLOCKQUOTE>

Uh yeah, that $1000 includes flying me out to hand deliver it!