Holley Pro-Jection Installation on an AMC 401

Contributed By: Ray Drouillard

I installed Holley Pro-Jection 4Di on my '89 Grand Wagoneer this weekend. I had to do more fabricating than I thought. It's still not totally done. I'll give y'all a quick rundown.

The project is at least three bananas. I don't consider it to be as difficult as rebuilding an engine, but some might. There is no precision mechanical work to do.

Note: You MUST get rid of the air injection system (awwwww). It'll throw off the O2 sensor and cause an extra rick mixture. If you must, you can inject air after the O2 sensor. There is a tube that goes into the side of the OEM cat. That might be a suitable injection point.

The O2 sensor should be mounted after the "Y" pipe so that the system can read the oxygen from the whole engine. We had the sensor mounted at a muffler shop before we started on the rest of the installation.

ECU:

You'll want to mount the ECU first. The manual has the fuel system listed first, but you'll make less trips under the vehicle and will avoid putting the wires on the pump while on your back and fighting the limited clearance

After much deliberation, we decided to mount the ECU vertically just in front of the front passenger door, as close to the firewall as possible. We peeled back the carpeting and bored a 2" hole through the firewall right behind the fenderwell.

The wiring harness is a big thick thing with two wires coming off within a couple inches of the ECU connecter. One wire is a DB9 that goes to the serial port of the laptop, and the other is for the optional non-computer calibration module. Those stay inside the vehicle.

After about two or three feet, the big cable splits off into four single wires and about ten smaller cables, and eventually into 16 connecters:

• Fuel pump relay & ignition relay (same cable)
• Throttle body group - one cable with connectors for:
  • Fuel Injectors
  • IAC Motor (high idle)
  • Air Charge Temp sensor
  • TPS (throttle position sensor)
  • Coolant Temp sensor (on a longer cable)
• MAP (manifold absolute pressure) sensor
• Oxygen sensor
• Distributor (GM, can be adapted to Ford) - for use with spark control
• Knock module (GM) with single wire and connecter for a knock sensor.

Feed the cables one by one through the hole. There is very little clearance between the hole and the fender well. This, by the way, helps protect the wires and minimizes the amount of gunk that the grommet and whatever you use to seal the hole will have to stop.

Start with the biggest ones and feed the connecters through the hole one by one. We used a coat hanger with a small hook bent in the end to fish the wires up. After about the third wire, I fed the wire down, the crawled under the Jeep to pull it through. Then, my dad slid the coat hanger down and I hooked the connecter to it and he pulled it up. I saved the O2 sensor for last and hooked it up and tied up the wire while I was under the Jeep.

We mounted the relays on the fender well and the MAP sensor to the firewall near the passenger side next to the heater core box.

Once the engine was cool, I removed the radiater cap to relieve any pressure, then reinstalled it. Then, I removed the spark CTO valve (in front of the carb). I put my finger over the hole to stop the leaking, then I lined up the coolant temp sensor, removed my finger from the hole, and quickly pushed the sensor in. I used no teflon tape, since that might prevent a good electrical connection.

There is a ported vacuum line on the throttle body that goes to the vacuum advance. No CTO is necessary. Eventually, I am going to use the spark advance feature of the ECU (I need some connecters). Then, I won't need the vacuum advance at all.

The next thing to do is to mount the throttle bodyt. Mount it like you would any carb. The throttle linkage and cruise control linkage were a challenge. If you are removing a 4V carb to install pro-jection, you will probablly have less difficulty. I removed the stock Ford 2V carb and adapter plate.

Route the cable that contains all the throttle body connecters and plug them in. Route the cable with the coolant temp sensor up to the coolant sensor and plug it in.

There is a red wire with a fuse holder. That goes to straight 12V. If you have an ammeter, make sure you connect it on the side of the meter opposite the battery.

Finding an ignition source (12V when the ignition switch is in the "run" position) turned out to be a major pain. I finally ended up stripping a little insulation off of one of the wires that goes to the two-wire connecter that goes to the ignition module. I used a meter to see which one hag 12V when the ignition was turned on. It turned out to be the red wire on the module. The wires in the Jeep's wiring harness are not the same colors as indicated in the manual, by the way.

Crimp a ring connecter onto the green fuel pump wire (it comes from the fuel pump relay connecter) and run it along the firewall, behind the fender, to the underside of the vehicle along the frame. I also ran a ground wire, since I don't trust a frame ground. If you have a bad fuel gauge due to a bad ground, you know why.

Go through the instructions and hook up all the wires. We forgot to hook up the wire that goes to the coil (tach wire). The ECU didn't know that the engine was turning over, so we would get a brief run from the gas that is initially squirted in, but the ECU wouldn't supply more fuel because it didn't think the engine was running.

Coil the excess wire neatly and tie it down to the fender well. My dad will laugh when he reads this because the wire on my fender well is anything but neat. I have an engine guard oil pump and the associated wire on the fender as well as the fuel injecter wiring.

Cover any unused connecters with a baggy to keep the conducters from corroding.

Fuel System

I used the stock fuel lines. So far. so good. There seems to be no problem with using a 1/4 inch return line.

Before you start, get about 4-5' of 1/4" ID fuel line and 4-5 feet of 3/8" ID fuel line. I got 3' of each and had to do some less-than-optimum routing. Also, I used some extra line that I had laying around. I bought four extra small hose clamps and still had to scrounge one up.

Mount the pump inside the framerail in front of the gas tank. Eventually, I'm going to mount a mini skid plate under mine.

They give you a little plastic filter to mount between the gas tank and the pump. This gets out the big chunks so that they won't damage the pump.

Hook a piece of 5/16" ID fuel line (perhaps 4-6 " long to the output of the pump and clamp it with a hose clamp. Tighten the clamp until the rubber just starts to bulge up in the screw slots of the clamp. Slip another hose clamp onto the hose. Use a short piece of hose clamp to connect the filter to the input of the pump. Connect a short piece of hose to the input of the filter. Don't forget to add the extra hose clamp. Attach both wires to the pump (+ and ground) and secure with the nuts. Cover the connections with grease, gasket goop, or something like that. Drill and tap a hole in the center of the frame rail. The center has the least amount of force applied, so we do put the hole in the center. Test fit the assembly and remove a large chunk of the gas line. Cut it with tin snips or something, then cut the ends off clean with a tubing cutter. Make sure that you cut the supply line, not the smaller return line or the brake line. Make sure that a lot of the hard line is slipped into the hose. This is a high pressure connection, and there is no bead on the line (unless you have a beading tool). Mount the fuel pump and tighten down the two hose clamps that you had slipped onto the tubing. Remove the excess slack from the wires and tie them up.

The stock fuel pump isn't as difficult to remove as I had thought. Remove the output tubing so that it'll be easier to manouver out. Cut the input tubing off. Remove the two bolts and slip the unit out. Cover the hole with the provided cover plate. Mount the big metal fuel filter that Holley provides. I mounted it with the nut that mounts the cruise control diapram. Not pretty, but it is effective. Attach a piece of 3/8 hose to the tubing that was going to the input side of the pump, and run it to the input of the filter. Run a long hose from the output of the filter to the input of the throttle body. The fuel return line of the throttle body won't fit a 1/4" ID hose. We removed the barb and screwed in a 1/4" barb. Then, we ran a line along the manifold down through the same routing as the stock return line. We were tempted to use the origional hose, but it was getting soft and we were switching to a higher pressure system.

At this point, you are ready to go through the initial test and startup procedures. When you turn on the ignition, the fuel pump will run until the lines are all full of fuel. Then the injecters will buzz and the initial priming charge of fuel will be injected.

I couldn't get my laptop to talk to the ECM, so I couldn't do any of the calibration.

The system is acting stupid whenever I change the throttle position, but it goes back to closed loop as soon as it figures out what is happening. Getting the correct maps into the ECU will make a big difference, I'm sure.

I couldn't get the spark advance control hooked up. Below is some information that I sent to Michael Baxter. If anyone has any experience with ESC systems, please send me some mail.

The 4Di is advertised as having a programmable advance curve. There is a 3D graph that you program. The independant variables are engine speed and throttle, and the dependant variable is the spark advance in degrees. It'll advance the spark up to 45 degrees. If I want more, I suspect that I can set the base timing at 12 BTDC, which would net me a maximum advance of 57. I doubt if I'll need that much advance, though.

I had assumed that they were going to use engine speed and MAP as the independant variables. Go figure...

Anyhow, there is a connecter that hooks into a GM knock module and a knock sensor. Actually, the connecter that goes into the knock module has a wire coming from it that goes to a knock sensor. The knock sensor is therefore connected only to the knock module, and the knock module is connected into the Holley ECU. I suspect that the spark module retards the timing when the knock sensor is activated. It has to be more than a simple amplifier, or Holley wouldn't have bothered to use it.

There is another connecter that goes into a GM distributer. I have an adapter that allows it to hook to a ford (non-signature TFI) distributer.

Hmmm... I wonder if the box is routing the signal from the relucter through the box (to advance the timing), and back to the ignition system. If that's the case, a little creative wiring should do the trick.

The instruction manual that Holley sends with the kit has big gaping holes in the info. Sort of like the factory manuals.

By the way, I can breadboard an amplifier or attenuater or whatever. Unfortunately, getting the use of a 'scope could be a challenge. Maybe I'll put a ad in the Tech News (campus newspaper for Lawrence Technological University) and let some bright young students do it as a senior project. I got my BSEE there in '90, so I should be able to do this.

By the way, I did a fair amount of fabrication to get the kit installed. I haven't been able to program it, though. I can't get my laptop to talk to the ECU. It's acting stupid whenever I change the throttle position, but it goes back to closed loop as soon as it figures out what is happening. Getting the correct maps into the ECU will make a big difference, I'm sure.

Ray Drouillard
'89 GW with freshly rebuilt 360, Engine Guard, RV cam, Edelbrock performer manifold, three inch exhaust system, and Holley Pro-Jection

Epilogue

I got the computer to talk to the ECU a couple days later. I called the place that sold me the used laptop and he said to bring it in so that he could enable the commport. Duh! I went into setup and found that the commport had been disabled. I tested it by using a nullmodem adapter and a serial cable and having it talk to my computer at work.

I fumbled around with the map quite a bit. I wanted to use the O2 sensor voltage as a guide for programming it. That voltage was oscillating all over the map, though.

I called tech support and they said that the voltage is supposed to vary. I don't think it should vary that much, though. They also said that I am supposed to use the offset as a guide, and to set it slightly rich. I suspect that the variance in the O2 voltage will go down when I get everything programmed right.

I found that I was right on that last thought this morning. I moved the whole map up and the voltage variance settled down. It still stumbles at the higher MAP (manifold absolute pressure) regions of the map. I guess the AMC360 can suck the air-fuel mixture in better than the GM350 that the map was made for.

Today, I ordered a Jacobs ignition system. I am going to have to get a Ford non-signature TFI distributer to make the Holley's spark timing control work. More fabrication ahead... It's a good thing I'm a glutton for punishment.

By the way, I'm going to make my final map available as soon as I am done. It's just a file that I can send via email.

Oh yes... don't even think of trying to program the ECU with a macintosh. You need a real computer to run the Holley program.

Ray Drouillard