EFI fuel pump ?

[FastWoman]

Mike, I suppose I never noticed whether my carbon deposits (both plugs and tailpipe) were fluffy vs. harder. However, I am 100% certain they were from an extremely lean condition. When I richened the mix, the engine ran much, MUCH better/stronger, and the carbon deposit issue went away. My plugs are now a pretty mocha.

I, too, calibrated the AFM per Atlantic Z, and my CTS was functioning in spec. All electrical connections were new, injectors new, all other components individually cleaned/verified or replaced if needed, rebuilt distributor and correct timing, properly adjusted valve lash, everything verified to OEM specs, zero vacuum leaks (new tubing, accordion boots, intake/exhaust manifold gasket) -- and VERY lean running. The problem was the ECU, which had drifted lean. The fix at that point was the potentiometer. My Z is a '78 with a Hitachi ECU. If your '75 has the original ECU, it's probably a Bosch. I think it's the '78 (and '77?) Hitachi ECUs that have the biggest lean drift issue, but maybe also the Bosch.

Check to see what mixture your engine "likes" by fingering the AFM counterweight while your engine is running. If the engine runs at a higher RPM and better vacuum when you open up the AFM a tiny bit, it's running lean. Easy test. Ideally, your RPMs should drop off in either direction, whether you finger the AFM richer or leaner. In fact you should be just a tiny bit towards the rich direction.

As for the altitude question: According to the FSM, you should inject about 6% less fuel above 3675 ft vs. below. Up at 9000 ft, you might need an even leaner mix still. (I know that I lean a lot when walking around at that altitude!)

[Captain Obvious]

Anybody got a "best" method for priming an EFI fuel pump besides filling up the tank and cranking the motor?

The FI Bible (page 3) warns against running the tank out of fuel because it may be difficult to prime the system after that happens. They even used capitol letters and everything. I suspect that may be the source of the apprehension about priming the system. That said...

I didn't run my car out of gas, but I did the same thing you did and took the pump out. When I disconnected the line between the pump and the tank, gas came out of the tank and didn't stop. I had to cap the fuel tank outlet nipple to stop the flow. I think I had about half a tank and the head pressure was enough to keep the feed line to the pump full.

When I was done messing with the pump, I pulled the caps off the fuel tank nipples and hooked all the hoses back up. Then I pulled the starter solenoid wire off the starter and hit the key. The engine doesn't crank because the starter is disabled, but the fuel pump will run. I could hear the pump run and I could also detect a change in the sound once the pump fully primed. Then I could also tell when the bubbling stopped and liquid fuel had gone full circle back to the tank through the return line. Whole process took less than thirty seconds.

Connect the starter solenoid back up, hit the key, and vroom. Maybe I was lucky and mine primed itself easily.

[mjr45]

Sarah, you crack me up . I've tired fingering the AFM and it seems to like a much leaner mix, at one point prior to calibrating the AFM I had turned the "black" wheel 10 teeth clockwise which should have really leaned it out, but didn't or at least very little, maybe I'll try to enrich the mix and see what happens.

Cap'n, I may try to fill the pump thru the intake hole prior yo attaching the intake line or maybe just use your approach which would probably be safer than dumping gas all over the place. I'm going to use your the 539 check valve(your idea) with an Audi fuel line after I put new O rings in the pump.

Mike

Edited April 8, 2014 by mjr45

[FastWoman]

I'm not understanding, Mike. When you open the AFM vane very slightly (maybe 1 mm), what happens? RPM increases (indicating your engine runs too lean)? RPM decreases (indicating your engine runs too rich)? Not much (meaning your engine runs with about the right mix)?

Note: I'm not talking about adjusting the clock spring. Rather, I'm talking about pushing the vane slightly farther open than it "wants" to be.

Edited April 8, 2014 by FastWoman

[mjr45]

When I finger the AFM vane counter weight CCW the RPM goes up, by going CW it drops a little. If I'm understanding you correctly its running to lean?

If that's the case, I should probably go the potentiometer route rather than adjusting the AFM clock spring.

Mike

Edited April 9, 2014 by mjr45

[Zed Head]

Playing with the vane at idle only changes idle fuel mix. There's a screw adjustment on the side of the AFM to adjust idle mixture. It's described in the FSM.

Have you checked coolant temperature sensor resistance at the ECU plug yet? That needs to be right before messing with the AFM. There's a chart in the FSM, Engine Fuel chapter, that will tell you if the ECU is seeing what it needs to see. Check at the ECU, if it's wrong then check at the sensor. If it's wrong at the ECU but right at the sensor, check the wiring in-between.

Best to get each individual item correct before modifying any. They all interact, and the ECU expects them all to be right.

[mjr45]

Zed yes I've checked the CTS per FSM and the EFI Bible and is all good, I've also checked the thermotine switch and the AFM electrical with the AFM resistance being off about 12% from normal values. It was my understanding that fingering the AFM vane would tell if the engine wanted a leaner or richer A/F at idle, which should if I understand correctly be the same thru out the RPM as long as the ECU is operating correctly to enrich the mix at higher RPM, or am I totally wrong on all of this?

Mike

[FastWoman]

Yes, it's (much) too lean, at least at idle.

First thing to check: Throttle position sensor. See the FSM.

Next, check for vacuum leaks, and make certain every other fuel component is working properly. Check timing, valve lash, etc. Calibrate the AFM clock spring per Atlantic Z. If you've ruled everything else out and still have a lean running problem....

Yes, you want to add the potentiometer to richen the mixture. You don't want to richen the mixture with adjustments to the clock spring, because that will leave you with a leaned-out engine at WOT.

Zed is correct that there's a separate idle mixture screw. However, in MY experience (not to say other people would experience this), it doesn't really do much. I have mine adjusted 5 full turns open (CCW) from fully closed. This is an arbitrary position, but it allows at least some flow-through in the event of a backfire. Anyway, with my idle mixture screw adjusted 5 turns open and my AFM calibrated per Atlantic Z, I've found my mixture under load is approximately correct when my idle mixture is correct (adjusted slightly to the rich side of the peak RPM / vacuum). I judge my mixture under load entirely by the "feel" of the engine and my reading of the plugs, currently lacking any more sophisticated method. But plug readings are a pretty good indicator.

Edited April 9, 2014 by FastWoman

[siteunseen]

If you do decide to do the potiometer route a volume knob at Radio Shack is pretty much what I got. Here it is, 10K-Ohm Linear-Taper Potentiometer : Potentiometers | RadioShack.com. You put the bullet connectors on then plug it in between the original bullet connection on top of the intake. I ran mine inside the car so I could play with while driving. It worked wonders without doing any kind of damage to the existing wiring for less than $5. If it didn't help I could just unplug it and plug the old wires back together. That was for testing, now I want 1 like Fastwoman has that goes under the dash with the ecu. Her's is a lot more precise.

Edited April 9, 2014 by siteunseen

[siteunseen]

Fastwoman and ZedHead, a question. I'm thinking of puttting flat top pistons under my N42 fuel injected head. With the potentiometer's adjustable air/fuel ratio benefits do you think it would run ok? I've got triple Webers but I like FI way better and have a good system, it's all new.

[EuroDat]

I have been playing with my AFM the last couple of days. The air temperature sensor was shorted, so I fitted an alternative thermal couple. After that I calibrated the spring using the procedure on the atlanticz site and it ran very lean after that. So lean that I couldn't correct it with my varible 2000ohm resistor in the water temperature sensor circuit.

In the end I took it out on the road and kept decreasing the tension until it started running to rich and worked out a point in between.

The spring is about 125grams compared to the 148grams on the site. The resistor is now set at about 650ohms.

Ill post some photo later if you want. Also have a video of the spring working with engine reving. See my thread http://www.classiczcars.com/forums/fuel-injection-s30/51993-faulty-air-temperature-sensor.html

Chas

[EuroDat]

I'm thinking of puttting flat top pistons under my N42 fuel injected head.

I don't know if the temp resistor will compensate enough, but with flat tops you're going to be in the mid 10's CR depending on head gasket thickness.

It won't like standard pump gas much.

Chas