Auxiliary Air Valve (AAV) disassembly and rebuild guide

As promised I’m making a detailed guide for rebuilding AAV.
First step is to remove thermo bulb.
You can do that by pressing in the other end (air outlet).
Spray decent amount of oil inside the valve to make disassembly easier.

Don’t worry you will not brake anything inside.

Keep pushing till the bulb falls out.

When the bulb falls out, take out the piston and spring and then push the remaining part the other way back.

Theoretically you can just press it back to its previous location, but in this case I’m pressing it all the way out as I want to use different air hole configuration (more info on different air holes will follow)

After full disassembly all parts go to ultrasonic cleaning.

To be continued…


Here are all the parts of AAV:

And here is why you have to change the thermo-bulb when you rebuild AAV:

Oil from the bulb probably leaks with age and even though the valve is not stuck it will no fully close anymore.


More about different AAV configurations:

There are 4 different part numbers for AAV and they all have different air hole configuration.
Here is a picture of different hole configurations:

GREEN: EAC4438 used on 5,3l engines with cats and air pump (A emissions) I would guess it is designed for 88c t-stats in mind as it will be open for longer period then other configurations.
BLUE: EAC2273 used on 5,3l engines except A emissions. I would say this configuration should be used if you dump air pump and use 82c t-stats. I use this configuration on my 5,3l engine with lucas ignition 82c t-stats, air pump removed. It idles ~1100 to 1200rpm when cold and gradually goes down to 750rpm when temperature reaches ~ 60-70c.
RED: EBC1198 used on 5,3l engines from engine no.: 57762 up to introduction of facelift (probably introduced with Marelli ignition??) and then on all 6.0l engines. I personally tried this valve on my 5,3l engine with lucas ignition, but cold idle was a bit too high for my taste. It would idle ~1500 to 1700rpm and then gradually drop to 750rpm when warmed up.

Then there is 4th configuration EBC3819, But I haven’t found any pictures or any other info on how it looks like. It was used on facelift 5,3l engines from VIN179732 to VIN 226645. If anyone has some info on how this valve looks like, please share.


EBC1198 John :-)). Stock option for my 6.0L

From my experiments with AAV, what part number one uses hardly matters if everything else is properly set up. Cold idle seems to be dependent on too many factors

1 Like

Thanks :wink: typing mistake :wink:

I have tryed myself 2 different valves EBC1198 and EAC2273 and there was a noticable difference in cold idle. EBC1198 when cold was idling ~1500 to 1700rpm and EAC2273 was idling ~1100 to 1200rpm. Both fully closed at ~ 70c and then idle settled at 750rpm.
Just out of curiosity how high is your cold idle with EBC1198 on 6.0 engine?

My car is a garage queen and rarely gets out of the garage when the weather is not nice and the sun is not out.
The cold idle seems strongly dependent on how cold it actually is, as the fuel enrichment governed by the CTS apparently plays a very big role too.
I have never seen cold idle at 1700 rpm – this is in the range of the injector cut-off by the way, and suggests something is not right.

For full disclosure – I have done a number of experiments in the close to 8 years of XJS ownership, but to this date, have not achieved that magic V12 idle that lets you balance a coin on its edge.
This seems to be possible only when I force the ECU to run in open loop (lambda sensor feedback inhibited, FF44 and FF45 flashed on the display).

PS I am looking at one EAC2273 and EBC1198 right now – the top part of the former (brand new unit) appears to be pressed at least 2-3 mm deeper into the body than the top part of the EBC1198 (my original unit, do not believe it’s been split).
Not sure why.

I beleave this is part of calibration procedure. The deeper top part is pressed in the sooner it will fully close. On the ones I had it was also pressed in a bit differently on each valve

Thats why I switched to the valve with smaller orifice. Reg silky smooth idle I’m still in pursuit for that as well :)))

I have sequential injection and electronic ignition control via a Megasquirt. The only way I can get a silky smooth idle is to richen the idle mixture to about 13.5:1. Running 14.7:1 or close to it…results in the original “stumble” I had on idle with the original gear.

Do they all use the same bulb?

Bulb is the same. I have seen 3 of the 4 different part numbers.
Also, as discussed in a different thread, the bulb is a Mercedes/Bosch design used in millions MB (and BMW cars).
Apparently the MB guys knew it before we, the JAG community figured it out:

Post #7 from July 2010 here:

Kirby’s Book was cited in their forum. Well done!

1 Like

I have found at least 2 different bulbs, but the difference is only in bulb supplier or manufacturing changes, as they both are marked with same temperature rating -30c

That is correct from all of my experience, best, JW

Because it was mentioned here I’ll ask, if you don’t mind? I’m a new-ish owner, about 2+yrs. 1990 XJS convertible, 5.3 v12…

About leaning out or riching the fueling mixture. I’m assuming folks are talking about adjustment with use of the fuel pot on the side of the ECU…?

With the ECU is sitting in its place, fuel pot adjust port on the outward RH side, which direction (to front of car or back) is the make shift handle I’ve fashioned to be turned??? Should I click my little rigged up handle towards the front or back of the car for say, a leaner mix?

I’ve counted about 13-14 clicks. I’ve simply put mine in the middle since I’m having a hard time figuring out which way is rich, which way is lean. Any any any help would be greatly appreciated…



I would suggest that you start a new thread.

To answer your question about the Base Fuel Map Setting: CCW: Rich - CW: Lean
It’s more than this though, you need to hook up a potentiometer at the feedback monitor socket and measure voltages. Yours being a later model you might also need a capacitor in parallel. Otherwise it’s just guesswork.
More details at Kirbert’s book and at Bywater’s site

1 Like

All great pointers and leads… I’m on it!
Thank you very much!