Junkyard Tuned Port Injection For AMC 360

[ThisGuyUKnow]

So I have been flirting with this for several years and never seem to have a jeep long enough with the 360 to do it. I am in the process of acquiring an 85 CJ7 Laredo with a freshly rebuilt AMC 360 and I have decided that this will be the base and recipient of this experiment.



People have been doing junkyard throttle body injection (TBI) conversions on the 360s for years and this is a very similar process. For anyone who does not know TBI is computer controlled fuel injection that still uses a wet intake like the carb. This will be a Tuned Port Injection (TPI) which will actually inject the fuel to the intake runner, just like Multi Port Injection in modern vehicles except TPI is batch fired and MultiPort generally fires individually per cylinder.

So first off edlebrock actually makes an AMC v8 MPFI kit for a huge price of about 3000+ dollars. Howell and other companies make bolt on TBI kits for about 1000+ dollars. These are bolt on, come with instructions, and are definitely the way to go if you have the cash or do not have strong wiring abilities. My conversion is going run me about 800-1000 dollars and most of that was spent years ago when i first acquired the modded intake. The second big expense was the wiring harness.


Alright so the there are a two major hold hurtles for converting to TPI first off, unlike the AMC 6 Cyl which got redesigned to the 4.0 with MPFI, AMC v8s never saw fuel injection in production so no factory compatible Multiport intake exists for the v8 (not counting the edlebrock kit which appears to be a modified torker). So google searches and trolling through lots of threads alot of people who know way more than me when it comes to engine performance all seem to point to a single plane 4bb intake could be modified to accept fuel rails and injectors. This is exactly what it appears Edlebrock had done in their kit. Two intakes id'd as potential donors for such a project that were suggested were the edle torker and the holley street dominator. More of these internet "experts" seemed to lean towards the torker for whatever reason. I am not an engine builder nor pretend to understand the mechanics involved with getting proper atomization of fuel and air flow but basically once converted the intake would cease to be "wet" intake and the runners would serve only to supply air and the fuel would be directly injected into the intake runners on the head.



I have had the Modded intake for probably close to 3 years now. I ended up cheating a little on the intake. I did not acquire one and have a builder mod it to my specs. Instead, at the time, there was a guy in GA that ran a business called CustomeEFIs specializing in producing FI for AMC jeeps. The guy temporarily got into TPIs before he went under and I got the intake from a member on IFSJA.org who had started the project but due issues gave up on it. The intake is a holley street dominator and the throttle body in it is also a holley (I have not id'd the TB model yet) and the fuel rails are custom. The injectors and sensors are all based off a 90-92 Corvette 350 harness. I'll hit these in detail later but a good number of the sensors required for the TPI computer are actually located in the intake.

Another issue is that one of the major benefits of any FI system is the electronic spark advance. The AMC v8s use a mechanical advance and since they never made a FI v8 there is no boltin electronic distributor. I will have to either mod an HEI or stock distributor to lock out the mechanical advance and convert to spark control. The TBI pioneers have long since found a way to do this and I am hoping that when I get that far that the process will be similar. I will use the TBI write ups as a guide on that one.

I ended up joining a corvette forum and trolling around on there until a 90-92 vette showed up as a complete part out. I contacted the guy and we worked out a deal for 200 bucks he would carefully extract both the engine harness and the underdash harness with no cutting (with the exception of where the harness enters the doors, don't think there is anything potentially important in the doors) What I ended up with was this:



the birds nest at by the bike jack is the interior harness and very dense. The wrapped harness is the engine side and less intimidating looking... well slightly. I stretched it out around the intake to kinda see where everything ends up.

 

[ThisGuyUKnow]

I am going to start on thinning out the wiring within the next week or so. First I am going to ID all the essentials in the engine harness, i.e. all the sensors the ECU mus have to control the engine. Then work my way to the distributor and the diagnostic cable.

When I am done I plan to not only have the engine systems run through the Chevy harness but I also will run the starter, and the the charging system through it as well. I plan to completely integrate what is left from the CJ wiring (mostly stuff for the guages and the headlights) into the chevy harness. I will then rewrap it all and hopefully set it up so it doesn't look like a shade tree install but a somewhat professional kit.

Thats as far as I am now. I am kinda more making this write up so I can keep track of what I am doing and can look back on later to see how I did things incase there are problems down the road. There are a lot of wiring schematics that I have to sort through and a few issues I can already forsee (the Chevy has a vehicle anti theft system that tells the ecu to disable the fuel and spark if it does not sense the right key is one example) and will have to find work arounds for. I expect this write up to probably be ongoing from now to next spring, since I do not have a garage I probably will not actually install it on the engine until the end of winter. So stay tuned, prepare for very wordy posts, and grab some brews.

 

[Jimbos76cj]

Looks good.. Can't wait to see how it looks at the end.

 

[ThisGuyUKnow]

Today I isolated all the wiring for the ECM and engine control and cut out all the unnecessary systems. I labeled all the sensors and everything that I cut for and made sure that I left enough wire to solder to should it turn out later that I removed something that I need later. In the second picture you can see a made a much more confusing mess of the main engine harness, but most of that is going to be tossed anyway. I need to get my hands on detailed CJ wiring to continue because I still have not decided whether I am going to use the 350 harness for the charger/alternator circuit and the starter circuit or if I am going to use the CJ wiring in those areas.

The corvette donor lacked a true power distribution center for housing relays so my next step will be further cleaning up the ecm wiring and routing everything where it needs to go for the AMC engine setup and I will have to decide a location under the hood to cleanly add a bunch of relays such as the cooling fan relays and the fuel pump relay. An underlying goal for this whole project is I don't want this too look like a slapped together hack job. Once the wiring harness is in I want it to look like something that was done factory.

 

[ThisGuyUKnow]

Further organization of the ecu wiring. I can already tell I am going to have to shorten alot of sensors due to differences in the 360 block vs the 350 block.

 

[ThisGuyUKnow]

No turning back now. I completely removed all the wiring from the cj harness. The only stuff I left un molested were the lighting harness, the wiring for the fuel sender, the harness for the tranny and tc, the main power feed and the acc hot wire.

 

[ThisGuyUKnow]

I removed the Edlebrock intake and holley carb today.



Its my first time swapping a v8 intake, I opted to use edlbrock gaskets with no valley pan since the holley intakes are supposed to not require the valley pan gasket. I was not expecting the gap at the front and back of the block to be so big so I ended up with a large air gap and had to remove the intake and then laying on a thicker layer of rtv.



So maybe there is some sort of trick I dont know about but I could not get a socket and torque wrench into most of the intake bolts. I did discover that the boxed and of the wrench fit in the torque wrench so I improvised. Surprisingly this worked.


I got to mocking up the wiring around 5 oclock but I quickly got frustrated and called it a day.

 

[tarry99]

So you have a Corvette Harness going on your AMC motor...........what ECM did you use?.........If it is the Corvette one the factory ECM's do well talking to the same motor & components it came from , But with a different motor and a different tune no so very well........there are other aftermarket ECM's that have a better ability to make that switch......but the tune is where the problem may be without spending a bunch of time on a dyno.................Do you have a solution for this?........If so... I'm interested...................Correct me if I'm wrong.

 

[ThisGuyUKnow]

So you have a Corvette Harness going on your AMC motor...........what ECM did you use?.........If it is the Corvette one the factory ECM's do well talking to the same motor & components it came from , But with a different motor and a different tune no so very well........there are other aftermarket ECM's that have a better ability to make that switch......but the tune is where the problem may be without spending a bunch of time on a dyno.................Do you have a solution for this?........If so... I'm interested...................Correct me if I'm wrong.

Well the ecm is a 727 and you can rewrite the tune. I havent made it to all the tuning tech but from what I understand the ecm takes into account the injector size and all the other sensor readings then adjusts the fuel based on the o2 in the exhaust. I did a 4.0 head and injection swap on my old cj so I used a 4.0 computer on a AMC 258 i6 / 4.2l and it drove very well and that is a jump of 16 CIs. So I am hoping that it will be able to compensate for the extra 10 CIs in displacement but we will see.. I have a moates autoprom setup so should be able to do the tuning from home if need be.

 

[ThisGuyUKnow]

Well progress is very slow going. I spent literally my entire day installing the fuel pump above the gas tank. I modified a plate to mount through two of the bolts for the tank and mounted a MSD external fuel pump and a clear fuel filter to it. I was a fraction of an inch off on one of the bolt holes and had to extend it out, of course my drill died so That took up alot of time. Then I has to make multiple trips to the parts store for hoses, clamps this and that.

Once I finally got everything In i tested the pump and before I knew it the top side of the intake was flooded and pouring gas onto the ground. It was leaking through the injectors where they meet the fuel rail and a slower leak from the tank to the filter So another trip to the parts store for new orings. Hopefully by lunch tomorrow I will have he fuel delivery knocked out and can shift to the ignition system.

 

[ThisGuyUKnow]

The flooded intake issue was easy enough to fix. It was just one injector that needed a new oring.

I modified a one wire HEI module to accept the 7 Pin TPI Ignition module. I removed the vacuum components and locked out the mechanical advance by drilling a hole through the top of the diz and putting in a sheet metal screw to keep it from moving. Alot of people when doing the TBI conversions mount the Ign module to the fender or fire wall. I found that the new 7 pin ignition module was a similar size to the HEI 4 pin module that mounts under the cap so I mounted Module in the same place under the rotor. Only one pin lined up but I i figure this is tight enough to hold it in place as this part of the distributor remains stationary and it should have enough contact for a solid ground.

Here is a pic of the the dist with mech and vac advanced locked out and the 7 pin wiring harness installed


This weekend I tried to rig up the harness and start the engine and was unsuccessful. I spent the whole weekend setting and resetting the Dist to TDC and the engine seemed to be sparking on the exhaust stroke. Finally I took Gerts advice and set it to 10 * BTC and it fired up the first try.

there is no coolant, exhaust pipes or belts on the jeep so it is definitely idling pretty rough but It IS IDLING! I was so exited I could have peed my pants.


Now I need to pull the wiring harness back out and put conduit on it to reduce some of that engine spaghetti I have going on under the hood and start putting things back together.

 

[CABBOTT]

Awesome the thread! The only prob I have with it is that you didn't make a duplicate for me while making this one

 

[ThisGuyUKnow]

Awesome the thread! The only prob I have with it is that you didn't make a duplicate for me while making this one

Get me a wiring harness and i will set it up for you.

 

[CABBOTT]

Haha I'll need more than the harness... That's what I like about this site good people always willing to help and lend a hand... And on a side note if you need anything just ask me as we'll ill try my best to help you out as well

 

[jdarg]

Well the ecm is a 727 and you can rewrite the tune. I havent made it to all the tuning tech but from what I understand the ecm takes into account the injector size and all the other sensor readings then adjusts the fuel based on the o2 in the exhaust. I did a 4.0 head and injection swap on my old cj so I used a 4.0 computer on a AMC 258 i6 / 4.2l and it drove very well and that is a jump of 16 CIs. So I am hoping that it will be able to compensate for the extra 10 CIs in displacement but we will see.. I have a moates autoprom setup so should be able to do the tuning from home if need be.

Good thread. You have a long road ahead of you though. Few things...

I'm surprised you didn't rework a PROM just for initial firing given how far the timing table is off. 10 CIs is negligible, and displacement is just one tiny variable in a system of many. The biggest problem I see for you is that the spark tables will be way off for your AMC cast iron heads vs. the performance (at the time) aluminum heads on Corvettes of early 90's vintage. If the injectors aren't stock for a 350 Chevy or close, your fuel tables will be off as well.

The vette timing table is probably not quite aggressive enough at high loads (the aluminum heads need way less total timing) but WAYYY too aggressive at medium/low loads, i.e. where you drive the most. Lots of newbs to tuning think they can swap in vette tables only to discover they get almost constant knock counts and retard under any sort of engine load. So the current table needs to go.

Try to find a spark table from a bin from an application with non-swirl port, non-fast burn, cast iron heads as a starting point. like the 84 vette maybe. IIRC they started going aluminum in 86 (there was no 85) and thats when the tables started changing a lot vs. the stuff from the early 80s. Even the 84 table will probably need to be dialed down a bit but it should be much much closer than what you have now.

Now this timing issue is being exacerbated because you are currently at 10BTC as you describe it. I assume static w/ the proper lead disconnected, and not while the motor is running under ECM timing control at idle, right? The bin is probably "expecting" 0-6BTC, so all your timing is up to 6 degrees more advanced now, and the timing was already way over the top for your heads. The ECM can't measure timing, except relative to where it thinks you have the distributor bolted down. It expects the static timing configured in the BIN matches that of the distributor and makes all its timing decisions based on that assumption. You need to fix this first. If you need any tuning advice shoot me a PM. I'm pretty sure I have some iron smog head bin files around you can use as a starting point.

 

[ThisGuyUKnow]

Good thread. You have a long road ahead of you though. Few things...

I'm surprised you didn't rework a PROM just for initial firing given how far the timing table is off. 10 CIs is negligible, and displacement is just one tiny variable in a system of many. The biggest problem I see for you is that the spark tables will be way off for your AMC cast iron heads vs. the performance (at the time) aluminum heads on Corvettes of early 90's vintage. If the injectors aren't stock for a 350 Chevy or close, your fuel tables will be off as well.

The vette timing table is probably not quite aggressive enough at high loads (the aluminum heads need way less total timing) but WAYYY too aggressive at medium/low loads, i.e. where you drive the most. Lots of newbs to tuning think they can swap in vette tables only to discover they get almost constant knock counts and retard under any sort of engine load. So the current table needs to go.

Try to find a spark table from a bin from an application with non-swirl port, non-fast burn, cast iron heads as a starting point. like the 84 vette maybe. IIRC they started going aluminum in 86 (there was no 85) and thats when the tables started changing a lot vs. the stuff from the early 80s. Even the 84 table will probably need to be dialed down a bit but it should be much much closer than what you have now.

Now this timing issue is being exacerbated because you are currently at 10BTC as you describe it. I assume static w/ the proper lead disconnected, and not while the motor is running under ECM timing control at idle, right? The bin is probably "expecting" 0-6BTC, so all your timing is up to 6 degrees more advanced now, and the timing was already way over the top for your heads. The ECM can't measure timing, except relative to where it thinks you have the distributor bolted down. It expects the static timing configured in the BIN matches that of the distributor and makes all its timing decisions based on that assumption. You need to fix this first. If you need any tuning advice shoot me a PM. I'm pretty sure I have some iron smog head bin files around you can use as a starting point.

I'll hit you up when it comes time to tune it. I have a lot of stuff to do in between now and then to get this beast ready for that.

The injectors are stock.

Not sure what you mean the ECM can't measure timing? There is a knock sensor that senses knock so it can adjust the timing to spark at the correct time. At least I understand it. that way. I set the timing to 10 tdc just to get it to fire up.. once I get further along I will unplug the ESC bypass and set it to 5 TDC at idle but right now there is no coolant or accessories or exhaust on the engine and I dont want to damage anything.


I have a moates Autoprom. Admittedely I know very little about tuning an ECM. But like I said that is step 300 and I am on like step 5.

 

[jdarg]

I suggest researching how the GM ECMs manage spark before you start tuning. You have some misconceptions about how it works.

You have to keep in mind the "best" advance at any given RPM/load is NOT necessarily just before it starts knocking. That is why the motor can't adjust timing based on a signal from that sensor. It only knows when its gone too far.

The knock sensor is just "insurance" against a problem that would cause detonation in conditions where it would not an engine that is running properly. Some TBI/TPI applications do not even use one. If the ECM gets a signal from that sensor, it simply starts backing off timing, a few degrees at first, and then it ramps the retard up more aggressively until it no longer gets a knock signal. How that retard decays out, etc. is even more complex. On most bins the spark retard action is pretty tunable - how many degrees it pulls at first, decay rates, max. retard, etc.

Don't just blindly set the static advance to 5 degrees. You need to set it to what's actually in the BIN, or adjust the bin to match the actual distributor. I've never heard of a stock application where 5 degrees is the expected static timing. 0 or 6 are the most common for TBI/TPI era motors IIRC.

 

[ThisGuyUKnow]

I suggest researching how the GM ECMs manage spark before you start tuning. You have some misconceptions about how it works.

You have to keep in mind the "best" advance at any given RPM/load is NOT necessarily just before it starts knocking. That is why the motor can't adjust timing based on a signal from that sensor. It only knows when its gone too far.

The knock sensor is just "insurance" against a problem that would cause detonation in conditions where it would not an engine that is running properly. Some TBI/TPI applications do not even use one. If the ECM gets a signal from that sensor, it simply starts backing off timing, a few degrees at first, and then it ramps the retard up more aggressively until it no longer gets a knock signal. How that retard decays out, etc. is even more complex. On most bins the spark retard action is pretty tunable - how many degrees it pulls at first, decay rates, max. retard, etc.

Don't just blindly set the static advance to 5 degrees. You need to set it to what's actually in the BIN, or adjust the bin to match the actual distributor. I've never heard of a stock application where 5 degrees is the expected static timing. 0 or 6 are the most common for TBI/TPI era motors IIRC.

Makes sense. I am not trying to be combative. But to be honest I am just not ready to start on the tuning portion. I have alot of other stuff left to do. And I need to get the project into a final configuration before I tackle the tuning. If 6* is normal for GM engines thats too easy, once I get the dist back in Ill get the light on her and line it up before the 5* mark.

Once I get to where we can tune it I'll hit you up. I should have all the necessary stuff to write new chips. I believe people have ran TBI upgrades without retuning the ECM but like you pointed out the performance heads of the Vette will call for different tables than what was used in the pickups.

 

[gert]

i wanna know how well this turns out

 

[ThisGuyUKnow]

i wanna know how well this turns out

Me too gert me too. Right now I have to cover all the wire and get it in permenant situations, Finish the serp set up so I have accessories, Wire in the cooling fans (controlled by TPI harness), get an exhaust run and remove a broken stud from the intake so I can install the throttle linkage. And it seems the outboard fuel pump has shorted out already so I have to see what is up with that.