Pulse Air Information

[MIPS]

Shamelessly cross-posting from the AMC forum because I think this was used in most late model AMC's and not in 70's/60's models.

I'm being very Experimenty lately and in a pile of parts I received the correct air cleaner, hoses, valves and solenoids to add the Pulse AIR system to an Eagle. My computer is a 1982 so it does technically already have the wiring and blindly operates the solenoids, but they were never installed. From what I can tell it uses pulsations of negative air pressure in the exhaust system to feed in fresh air either upstream just after the exhaust manifold flange or downstream at the catalytic converter. It doesn't use an AIR pump, a diverter valve or a bunch of extra plumbing so other than the rigid lines and check valves I already have all the "hard to get" bits to take a shot at trying and then cap the ports off and convert it back when I'm done. The system however isn't well documented other than it had a thing for rusting out and fouling the carb when a check valve fails so here's a diagram I drew up with how I think it goes together.




So here's the fun part. My car never had AIR management or a catalytic converter so I don't have upstream or downstream ports in the exhaust. My plan was to drill the hole in the upstream where it should be and a hole in the side of the resonator but I don't know if A) The correct pressures needed for it to work will be present and B) If the upstream port has something like an elbow inside the pipe to use the venturi effect to build negative pressure or if it's just a hole on the side of the pipe.
The other thing is that I'm missing the check valves. They look visually different from the AIR injection type with rounded tops and I can't find anyone that sells that style. Will the injection type check valves work in its place?

[MIPS]

Okay, so here me out. I'm crazy but I got ideas in my head so read on and listen to what I want to try and do.
I'm going to test this. The results might be obvious. This is very much pointless but I want to try an experiment with an Exhaust Gas Analyzer to see exactly what the reduction of hydrocarbons is without a catalytic converter as the earliest versions of air injection was simply "add air to exhaust and burning/partially burned exhaust hydrocarbons will more likely be fully spent before leaving the tailpipe". It might also not work at all because we're going to be comparing Apples to Pears.

So we know that the air management system for the Eagle was originally only available with the troublesome mechanical air pump. A diverter valve dumped air when the computer/throttle condition deemed the exhaust system to be unsuitable (EG. too rich and likely to cause a backfire) and an Air Control Valve directed air either upstream to the exhaust manifold via the air injection manifold or downstream to the catalytic converter but it could not block both at the same time, so that was left to the diverter. In all the system looks like this on paper:



Starting in 1983 AMC offered in some areas the option of PulseAir injection. No air pump and considerably less plumbing is required, including the omission of the injection manifold and relocation of the upstream port to after the O2 sensor which means it is entirely an emission control device that provides no feedback to the computer and requires no parasitic load on the engine. Through above mentioned magic that is negative air pressure pulsations in the exhaust system a pair of reed valves allow the exhaust to naturally aspirate fresh air. The Air Control Valve is replaced by two separate but identical valves that are either open, closed or both are closed, eliminating the need for a diverter valve. It's essentially two identical circuits. They just terminate in different parts of the exhaust system. On paper the PulseAir system looks like this:



The eagle has seen several revisions of the computer. In my case I'm comparing the 1982 CeC with the 1983 CeC (I've seen it alternately just called the MCU). Visually they are different computers with different connections, enclosures and harnesses but in reality the internals are relatively identical inside. Notably the wiring associated with the Air Management system (upstream/divert and upstream/downstream) use the same solenoids (12v 160mA at 80 ohms), the same OEM marked driver IC's (I have the footprint at this point reverse engineered but it's not helped considerably with narrowing down the list of 16 pin DIP driver IC's that existed in the 80's) and both pull the signal line to ground when they want to operate the solenoid because the subharness to the solenoid manifold is the same and B+ is present as long as the computer is powered.
The logical answer here is to upgrade an 82 Eagle to an 83 computer and wiring harness (which I can do as I have a complete spare MCU, sensor and harness kit) but that's an awful lot of work for just an experiment. The point I'm getting to is that because PulseAir doesn't have a diverter valve and there's only one other solenoid for controlling air direction when we need two, we can build an interposer that plugs in between the solenoids and the plug on the harness and converts the Pulse Air system to work on a mechanical air injection system, assuming the duty cycling between the CeC and MCU are similar enough.



=====The following is absolutely subject to change as I start testing things and find mistakes I've made=====
On the left are the four pins present at the harness, their wire color codes and their pin number locations on the diagnostic plug. (according to the 1982 TSM) The Vacuum Switching Solenoid is unrelated to the air system and is part of the Sol-Vac, so just disregard it.
I may still need to change my truth tables once I build and test a prototype (because I know I'm going to have things reversed) but normally the system in closed loop is sending air downstream to the catalytic converter, so the downstream solenoid is always open, allowing vacuum to pull open the valve. When the computer calls for air upstream we can translate one output into controlling two solenoids with a single pole-double throw relay. I'm using really cheap relays that draw 30mA at 12v. Using other relays will probably work but I would not want to go over 160mA in case you overload and blow the otherwise unobtanium driver in the CeC.
Anyways, the computer pulls the UPSTREAM signal to ground, the relay operates, the downstream solenoid vents and closes the downstream air valve while the upstream solenoid operates and vacuum opens the upstream valve. That I can tell there is no instance where both upstream and downstream will both open, so this works.
Regardless of upstream or downstream, when the computer wants to divert it pulls the DIVERT signal to ground and a second relay operates. We don't have a full-time ground available at the harness connector like we do for B+ and while DIVERT and UPSTREAM do get pulled to ground when signalled we cannot assume that path to ground will be available when we need it, or if the driver chip wants to handle a load like that but that I can tell the B+ is not going through anything inside the computer that would be disturbed by an extra 160mA. To get a ground we either have a tether attaching to the firewall or if you are like almost everyone else, you have a metal valve cover, so we can ground directly to that. So that relay opens the path to ground and both the upstream and downstream air valves as a result close, blocking the air path as if a diverter valve was installed.
==========

This leaves all the plumbing after the check valves. The rigid steel lines that would connect to a port upstream (below the mating flange to the exhaust manifold) and to the catalytic converter. (which I don't have, so lets just put a port in the side of the pipe after the converter support bracket and see what stupidity that does) Currently there is no affordable sources for these air tubes, so I'm relying either on someone to ship me theirs (even if it's rusted out that's fine because then I have a bending template) or I'm off to the Home Depot to buy some EMT conduit and spend an afternoon at the pipe bender. Likewise there is no specifications I can find in regards to how these ports are fitted to the exhaust pipe. Are they orifice devices? Are they just pipes welded in? Do they need to be placed at specific angles?
No data. Nothing! This is where I'm at so far.