Z3M Secondary Air Pump - Bad Ground???

[Carnut68]

I have a 1999 Z3M with 36k miles. I am getting P1421 and P1423 which indicates a problem with the secondary air pump system. The pump is good...runs fine when power is applied but it is not getting power on cold start. The SAP control valve functions as it should...only passes air when vacuum is applied. No evidence of moisture in the system. Vacuum hoses appear in tact. The SAP relay...which oddly enough doesn't appear to be in the position indicated by my Z3 shop manual...appears to function properly when you test it outside of the car. I have not tested the SAP solenoid but at this point I don't think that is the issue either. I am thinking it could be bad ground but I am not sure where to go from here. Any advice where to look for a bad ground or what else it could be?

[rf900rkw]

If it's showing the wrong location for the relay, you have the wrong diagram. If it's not showing where the ground is located, you have the wrong diagram supplier. Toss the Bentley and get the factory diagrams.

But even with that, grounds are easy to check. Do you have power on the ground wire? If so, ithe ground is bad.

- - - Updated - - -

Ground is X6455, which is on the back of the RH strut tower. If the ground point was bad, the engine wouldn't start.

[Carnut68]

Randy - I am actually using the BMW Z3 service manual. Page 610-31 Figure #15 shows the SAP relay as item F in the E-box...a green relay closest to the positive battery junction. However, I was under the impression that SAP relay was the salmon colored one in the second row because the hot wire to the SAP runs to that relay. This is marked as the main DME relay in the service manual. Is that was actually provides power to the SAP? If we removed that relay and jumped the pins (I forget which ones) the pump functioned fine at start up. You would think the relay would be the problem but it made the proper clicking noises when we pulled it out and applied 12V and it tested OK for continuity. So I am not sure about the bad ground theory (that came from my brother in law who was poking around in the E-box in that relays position and said something wasn't right). All I do know is that the pump is good, the SAP diverter valve is good, the relay appears good but the pump still done run on cold start up. And of course the registration is due for renewal and I can't renew until I get it through emissions and I can't get it through emissions until I resolve this. Great timing.

[rf900rkw]

The pump runs only in a limited set of parameters. Basicly a cold start between certain ambient temps. The fact that you set codes means the computer tried to turn on the pump and did not see a corresponding lean surge in the O2 sensors. If the conditions were such that the computer did not command the pump on, it would not set codes.

m-dme-box.jpg

m-sap.pdf


Pump power to the relay is supplied directly from the battery (no fuse) via the jump-post. Relay winding power is supplied from the DME relay. The DME supplies relay winding ground (ground side switching). With the key on , you should have power on two of the relay legs. If you manually pull the DME side of the relay winding to ground with the key on, the pump should run.

[Carnut68]

OK I misspoke earlier when I said I was using the "service manual" it is indeed a Bentley service manual. And it clearly has mislabeled the relays according to the picture you sent me. Thanks I think you have sent me some very helpful information. Now if I can only figure out how to use this information to resolve the problem. Chasing down electrical gremlins and reading schematics is not my strong suite, I would much rather bust my knuckles on rusted bolts in hard to get at places. I am surprised I made it this far figuring out what is NOT the problem...and before spending money on unnecessary parts. I have to study this to figure out what next steps are. Very much appreciated.

[rf900rkw]

Electrically, the M is VERY different from the Z3. Bentley seems to not acknowledge this.


I don't mind electrical at all. But I don't like doing it remotely. One wrong move can let lots of smoke out, and this is an unfused circuit.

[Carnut68]

Ok, just when I was starting to feel more confident you had to throw that in there to spook me. I took a crash course on relays and some of this is starting to make sense. So if the ground for the SAP relay comes from the DME relay then the DME relay could actually be the culprit. Oddly enough, that green relay has a red slash drawn on the casing. The identical green relay next to it ("unloader relay" in your diagram) has a black marker slash. Could be factory markings I suppose but I am thinking maybe someone did some prior troubleshooting before and swapped them. The DME relay seems to make the right clicking noise when I apply 12V volts to it but I noticed I do not get continuity across the winding poles 85 & 86 like I do on the SAP relay (which is slightly different). I thought that was indicating there was a problem with it but then I checked the green unloader relay next to...same thing, no continuity.

[rf900rkw]

Not quite.

The DME relay supplies +12V power to many components, fuel injectors included. The DME itself supplies the ground for each component. If the DME relay was faulty, the car would not start.


There should be continuity across 85 and 86. Since most relays have a flash diode incorporated, you will get a different reading depending on which way you hook up the test leads.

[rf900rkw]

m-relay.jpg

These are the pin numbers that match the diagram I posted earlier.

[Carnut68]

At the risk of sounding really dumb, what is the orientation in the engine bay of Fig 4 that you posted? And when you say "with the key on , you should have power on two of the relay legs" which two?

[rf900rkw]

Can't tell you which way it is turned. You just have to match it up.


Relay number 85 and 86 are the windings. They go to pins 8 and 4 in the socket. Pin 8 will have power with the key on. Pin 4 goes to the DME, which supplies the ground to energize the relay.

Relay 30 goes to the pump motor, pin 6 of the socket

relay 87 and 87A go to power and ground. Socket pin 2 will have power always. Socket pin 5 will have ground always. With the relay at rest, the power side of the motor is cponnected to ground. Thus both sides are connected to ground. (*there is a reason, we won't get into). When the relay is energized, the power side of the motor is switched to battery power, and the pump runs.

[Carnut68]

Thanks for your continued help on this. I didn't realize until I looked closer that the relay socket pins are actually labeled. Everything checked out except pin 4 which goes to the DME. We started checking all the other circuits connected to the DME according to the diagram, by checking the pins in the DME connector to make sure we were getting good grounds. We didn't get through all of them but did note we were not getting a ground at #56 and were getting resistance on this. At that point we gave up, I put everything back together including putting the DME back in, cleared the codes, started the engine and by some miracle everything seemed to work. I am going to cross my fingers and start it up today and hope that is still the case. If it does, I will drive it just enough to reset all the monitors and then get it to the emissions testing center. I still don't know what might have fixed it...if it is fixed...other than disconnecting and reconnecting the DME.

[BimmerBreaker]

"If unplugging it and plugging it back in doesnt fix the problem, try pressing the Any Key." -Helpful IT Guy

[rf900rkw]

How many programmers does it take to change a light bulb?


You should only show ground on pin 4 when the DME is commanding the pump on.

Pin 56 from the DME goes to one of the ground lugs on the RH strut tower. But again, if that was bad, the engine shouldn't start because that gorund is used for mission critical components.

[MRFAST]

At that point we gave up, I put everything back together including putting the DME back in, cleared the codes, started the engine and by some miracle everything seemed to work.

Back in the day... When encountering memory errors on PCs the most common 'fix' was to pop all the DIP chips on the memory board and then reinstall them. Galvanic corrosion over extended periods causes just enough of a voltage drop that TTL logic was unable to distinguish the correct logic signal. The same effect happens in connectors and other electrical interfaces.

[Carnut68]

Well I drove the car again today and the SAP fired up again and no CEL. I haven't driven enough to reset all the monitors so I can take it to emissions but will this week. I don't really know for sure what fixed it which kind of sucks after all the effort but I learned a lot throughout the process and really appreciate the assistance Randy. BTW, MrFast I also happen to be a current 07 RS4 owner;-) I spend more time on mye28.com and Audi forums but I will have to spend more time on here. With this hopefully behind me I can focus on the broken seat belt guides. That should be fun.

[MRFAST]

BTW, MrFast I also happen to be a current 07 RS4 owner;-) I spend more time on mye28.com and Audi forums but I will have to spend more time on here.

Completely different animals are they not? I have to say the RS4 is far more costly to maintain.

[Carnut68]

No comparison between the two other than they are both fun to drive. Roadster is a bit better on gas that's for sure. I haven't been hit with any maintenance bills yet. Had it about six months now and it has 63k miles. Probably due for carbon cleaning soon.

[MRFAST]

No comparison between the two other than they are both fun to drive. Roadster is a bit better on gas that's for sure. I haven't been hit with any maintenance bills yet. Had it about six months now and it has 63k miles. Probably due for carbon cleaning soon.

Don't price out brake rotors. You'll quit using your brakes.

[Carnut68]

Haha. I researched the car for years before I jumped in. I knew what I was getting into.

[Carnut68]

And of course the CEL came back. What a bummer.