05 BMW 325 Failure Code P0491 and P0492

Hopefully this will help the next person having to deal with the problem .

State inspection was due on the BMW last November .

  When I took it in they failed it on account of an oil leak on the VANOS line that feeds the mechanism that controls the variable valve system . A Royal pain to replace from all I'd read . So I really did not want to face that can of worms  and chose to ignore it as I had other stuff on my plate at the time and we don't drive that car in winter anyhow . But I can only put it off so long and had to face my fears inevitably .

The line in question is about a foot long with a banjo fitting on each end .

The front end is readily accessible ,

But the feed end

is attached way under the intake manifold and absolutely inaccessible without taking half the stuff out of the engine bay and still nearly impossible to get your hands in there . Eventually , with some colorful language , I managed to get it replaced and it no longer leaks . Put the car up on ramps , pulled the bottom plastic valences and de-greased the engine . That was actually the easy part of the job .

Found the intake boots broken too . Amazed it was not throwing a failure code on that .

So I replaced them as well .

Hey I did not mess it up and the engine still starts and runs well .

One of the persistent issues we have had with the car since we bought it two years ago was a "service engine soon" light coming on every so often . When I plug in my code reader to the ECU it tells me failure codes are P0491 and P0492 " Secondary Air Injection System Insufficient Flow" .
Further digging on the internet tells me the "secondary air system"  operates only on cold start , and pumps air into the exhaust manifold , for a total of ninety seconds , in order to fool the ECU into thinking it is running leaner than it actually is , precisely when it is giving it extra fuel to start . It only operates for ninety seconds when the engine temperature is below 120 deg F. Strictly a People's Republik of Kommiefornia requirement , so of course they put it on all BMW cars sold in the US . European market BMWs do not have this system and it is not critical to the proper operation of the car . I have been using my code reader to clear the malfunction code when it appears , which is only every three cold start cycles when the secondary air system fails to operate as intended .
But , since last November the State of Maine has changed its rules and inspection stations now must check that the Malfunction Indicator Light (MIL) is operating and has not been disabled in order to hide a malfunction , and all sensors must be "ready to read" with actual operating data .  If you clear the codes and drive it long enough without correcting the failure the MIL reappears and you do not pass . So when I took it in having just cleared the code. . .  it failed on because the ECU was "not ready to read" . . . So that leaves me no choice but to go down the dreaded rabbit hole again .

(Picture blatantly pilfered from the web somewhere)
This is the general schematic of the secondary air system , simple enough in principle but clearly fraught with multiple possible failure points at every turn . Vacuum is drawn from the intake manifold via an 1/8th inch line through a non return valve , an electric vacuum valve opens when power is applied , and said vacuum then opens the air pump valve on the exhaust manifold that allows the Secondary Air Pump to feed air into the exhaust manifold .

 But when you open the hood , this is what you see .
NOTE: the electric vacuum switch as shown in the above photo is now zip-tied in an accessible location . Manufacturer location is inaccessible under the intake manifold .

 To start diagnosing the problem first remove the cabin air filter and its housing , as well as the driver's side shield around the brake master cylinder to gain access to all vacuum lines and components that are still almost inaccessible underneath the intake manifold . Vacuum lines are likely rotted so they must be replaced . Easy enough if you can actually see where they go or even get to put your fingers on them .

This is what it looks like looking forward from the firewall which is about two inches away , under the cabin air filter housing you just removed but still can't get your hand on . After much frustrating contortions with a mirror I manage to disconnect the electric vacuum switch from its mount under the intake manifold , and found that the last person to work on the car had (likely accidentally) bypassed the electric vacuum switch completely and had plugged the vacuum line going to the air pump valve directly in to the intake manifold port . Additionally a second vacuum line going to a small vacuum reservoir to operate some flapper in the exhaust was badly rotted . And the cap on the 1/4 inch blind port was split . Simply put it was impossible for it NOT to throw a code . I replaced the 1/4 inch cap and all vacuum lines , and connected them as they are supposed to be . Yet when I start the car cold , the system still does not operate . Hmmmmm????
I test the air pump valve with a hand held vacuum pump and confirm that it opens when ten PSI of vacuum is applied .  It closes as it is supposed to, when vacuum is released . I bench test the electric vacuum valve and that operates as it should , opening when twelve volts are applied and closing when there is no voltage supplied . So we know those two components are sound and as I just replaced the vacuum hoses I know I am good there too . So what gives ?

After further frustrations and testing I find I have an intermittent working air pump . Some times when it stops on the right spot it will not start back up when you put power to it .

Remove the back cap on the pump and connected my multi-meter to the terminals and find the dead spots when I rotate the pump motor . If I put 12v to it when it is on a dead spot it will not run . Rotate the pump manually a couple degrees and it runs again .

So we have an intermittent working pump .  It's junk . A new one is 330$ at NAPA .

So figuring I have nothing to lose , I put a 3/16s drill bit to the aluminum rivets holding it together .
Don't forget to scratch a mark on all parts of the layer cake as if you get to reassemble it you want it all in the same orientation .

Top cover is off in no time flat , and we see the impeller . Take a 3/16s punch stick it in the hole in the hub of the impeller , grab a two pound maul and beat the living snot out of it to drive the motor shaft off the impeller . Yea you'll cringe while you are doing it . Never mind , its junk anyhow right ?

It will come apart , and this is the line up of components

Pry the front cap off the electric motor

And the back one too . They come of easy as they are lightly staked on those dovetails

When I pull the armature out of the stator it is evident that the commutator is severely and very unevenly worn where the carbon brush assembly contacts it . Chuck it up in a drill and hit it with a file and some sand paper to even it all up . You'll also need to true up the end of the shaft where you beat in it with the punch with some sand-paper till the bushing slides on freely .
Blow out with compressed air , lube the bearing on one end and the bushing on the other and reassemble .

Works like it was brand new , and starts every time I hit it with twelve volts .

Reassemble all the parts in to the pump housing .

Yea the plastic keeper that holds the motor in the housing is likely split from the beating you gave it .

Never mind that , hold it all together and stuff it back in it will stay together by virtue of the external pump housing .

A brass hammer held in my vice as an anvil to prevent deformation of the shaft . . .

. . . when we reinstall the impeller using an appropriate socket and a hammer to tap the impeller hub back on to the shaft . Reinstall the housing cover with new screws .

Bottom cover with the plug connection goes back on . Bench test and it works and blows more air than it did when it intermittently worked before .
Reinstall the pump in the car . . . . aaaaaaand it still don't power up when I do a cold start .
When I check the plug that feeds 12V power to the pump on cold start , I get zip . Additionally the plug that feeds the electric vacuum switch also provides no power on cold start .
WTF? what is wrong with me ?
So further down the rabbit hole we go .
Check al fuses in the drop down panel inside the glove box , as well as those in the engine bay compartment on the driver side next to the brake master cylinder , and confirm that they are all good .

I conclude that there must be a relay somewhere in the system . Mr Google tells me it is located behind the glove box .

Remove the glove box and the housing it sits in , and this thing falls out , about inch and a half in diameter coil of copper wire in a paper sleeve . Don't have a clue what it is.  An antenna of some kind?  one side of the packet has some mild adhesive . . . .  comment below if you know....

. . . and we find this salmon colored relay that powers the air pump .

Which  when removed and bench tested operates as intended . It audibly clicks as well as providing continuity between terminal #30 and #87 when power is applied to terminals #85 and #86 . Furthermore when I jump terminals #30 to #87 of the relay plug in the car , I am able to power the air pump telling me the system and associated wiring is sound and operates as intended .
But when I put it all together I still can't get the secondary air system components to play nice and operate on cold start....aghhhhhhhhhhhh
Been at this crap for four days straight now
In frustration I simply start the car and let it come to temperature for ten minutes while I clean my work bench , close shop for the night , have a beer and ponder the situation . I am yet no smarter about it all .
Next morning , just for yucks , I try starting again and miracle of miracles , I hear the air pump come on and no MIL .
OK . I am still no smarter but I can only assume that the ECU needed to find itself again with some real world operating temperatures to load all the fields I had cleared .
Drove it twenty five miles up to mom's and back so the ECU could get some more operating data and reload all its fields . Three cold starts more and the MIL has yet to reappear so I figure we are good .
Took it in next day for a re-scan at the inspection station and got my sticker .

 Enabling us to legally drive it up for my friend Clint's daughter's wedding yesterday .