SAI Pump diagnostics with Durametric,Bentley and Multimeter - FAILED !
 

Failed but eventually resolved - see below.
Boxster S 2001 6 speed.
Engine starts instantly, runs smoothly but SAI pump will not run in any key position
. Bench test of pump, fuse, relay is perfect. Pump been given the Krazy K SAI Pump refurb. Thank you K !
This is a problem because I have just completed the engine rebuild and need to get Smog Test to re-register the car. The engine died 10 days after I bought it – 2 years ago !

Following are the diagnostics I did & I hope others with this problem will find them a useful reference. You just need Bentley, Durametric and a simple Multimeter.
I also hope someone way smarter than me will help and tell me what big mistake I have made !. It was a Failed connection - see Post 26 following
More diagnostics - basically a no power problem to the SAI pump fuse even when the engine is running or any key position:
1. The wonderful Durametric cable +laptop will operate the change-over valve for the SAI system. In a very quiet environment you can hear it cycle when commanded by Durametric. BUT it fails to operate the SAI pump when commanded in any ignition key position or when the engine is running. The SAI pump never runs so it will not complete it’s “Readiness” for the Smog Test.
I used - Bentley Service Manual Wiring Diagram EWD 109 Diagram 12. to check the circuit as follows :
2. The OEM Maxi fuse 40 A in the trunk is perfect. The 4 pin 40A Bosch relay in position #10 bench tests perfect . Part # is 0 332 019 156 .
3. In any key position, there is no 12v power at the 40A fuse socket .There is continuity between one conductor of the fuse socket and pin 87 on relay socket.
Do the following test first
But there is no power at pin 30 of the relay socket with KOEO, nor KOEOff,nor KOEon. So logically the problem is upstream from here ?There should be power to the #30 socket for the SAI relay with key off ,engine off.
http://www.12voltplanet.co.uk/relay-guide.html
For the sake of completeness I also did a little more testing of downstream wiring :
4. SAI pump wiring - the brown conductor shows continuity to ground (as it should). This is the uppermost terminal on the pump when installed btw.
The red conductor to the SAI pump shows no 12v power in any key position. Neither does it show continuity to ground.
5. There is 12v power at Terminal 30 the main, big "Connecting Point" with the flip top lid in the engine bay.

So here are the upstream tests from the Maxi Fuse to find the break in continuity.

From the wiring diagram EWD 109, Diagram 12 in Bentley, the terminal A2 on the 40A SAI Pump Maxi Fuse in the trunk is supplied from the 'generator'(alternator) Terminal B1. So this is where the problem must be ??
You can not reach the back of the alternator to check this connection ! Could this be the problem? I poked my fingers in between the Intake runners and groped to find loose wires – nothing. But you can’t see in there !
This suspect connection B1 is also connected to Terminal 30 & the starter ,so I assume the connections at the back of the alternator are good also.
The alternator warning light on the instrument panel behaves normally- goes out immediately the engine runs. So surely this means all is good on the back of the alternator?
The diagram shows a mystery connection(?) between the alternator and the Maxi Fuse - “14 X59” – is this the ignition switch ? I can’t find any explanation in Bentley of 14X59 – anyone know what it is?
This is both embarrassing and disheartening – all the equipment and some diagnostics yet a total failure to locate and rectify a simple continuity problem .
So I am hoping for some helpful advice please.
Unrelated but I’ll mention what else Durametric found:
Misfire on 1,2,3.
Camshaft retard of 2 degrees also on Bank 1. The Ahsai wire pointer mod would have prevented this minor adjustment .And of course Bank 1 on a Boxster is the worst for camshaft adjustment . I am tempted to leave it – at least until I can fix the SAI wiring problem.
Grateful for any help. Thank you.

Sounds like the SAI and its supporting circuitry work fine.

I would chase the sensors that are used to determine whether the SAI should be run. Its only activated when then engine is cold.

I'll do a quick check of the Porsche repair manual in the morning.

Thank you for your offer of research.
Others may find this instructive:
https://www.youtube.com/watch?v=98W38PDAteE

The ECU commands the SAI only after engine has cooled down to certain temp. I also had a MAF sensor go bad earlier in my ownership that prevented the SAI to startup. The engine ran smoothly but no SAI. Replaced MAF and ok. I have confirmed also that you can force SAI once it comes on to go ready without driving the car. Many have doubted this to be true and stated car needs to be driven.

Have you checked the pump itself by jumpering 12v directly to the pump? Think it's only 2 wires to a motor.

I know the sai is an emissions device and technically needs to work to pass smog. If you bring a warm car to get smog checked, how are they going to know it's not working? Is it throwing a code?

That refers to pin 14 of connector X59.
X59 is one of the large connectors in the left firewall in the trunk. The white one I think.

http://986forum.com/forums/uploads02...1485955662.jpg
http://986forum.com/forums/uploads02...1485955725.jpg

http://986forum.com/forums/uploads02/x591485955677.jpg

Any chance that your DME is ROW?

I checked the repair manuals and there's desperately little on the SAI there. Something that I did notice was a reference to a ground strap to the secondary air pump bracket. If that's missing you may have a break in the circuit that drives the pump.

When the battery is disconnected everything clears including Emissions readiness flags. The SAI has to run at least once to set its "Ready" flag, that's how "they" block people from disconnecting the battery right before going to the emissions check.

Thanks for all the replies.I very much appreciate the help fellow Forum members.
Sooner or later it seems many of us get this SAI problem so bookmark this and the links for when it is your turn !
An ROW reflash would be my last resort.That eliminates the whole SAI system from the Readiness test.
Some people say the SAI is more trouble(=cost to repair) than it is worth in emissions benefit. For that reason, the Rest Of the World do not use SAI. The SAI runs for about 60-90 seconds immediately after a cold start. The SAI system is getting a command from the DME but there is a break in the power to the pump somewhere.
78F350 - thank you for the location. Yes ,it is the white one. The adjacent connector is black. I had checked that white connector block because I just traced the red conductor from the relay tray(40A Maxi Fuse) back there. There is no 12v + power from the appropriate pins in the 'feed' side of this connector. So basically the entire rear relay tray is 'dead'. I also checked both sides of the 40A fuse connectors for a fault to ground - no continuity to ground.
BTW anyone trying to separate this type of connector - be careful ! Pull out the big tab on the face. Wiggle the tab and pull toward you firmly until the top of the connector moves upward. Then it is free to be removed. Do not try to force it. Just keep doing push-pull on the big tab until the top of the connector jumps up a little. The tap has little ramps alongside the connector and these will lift the top of the connector when it is fully released. If it does not pop up,it is not fully released yet .I hope this helps.

Clearing the Readiness codes with a code reader or Durametric also does not work - they remain Pending and are detected when connected to the SMog Test machine=automatic Fail !

Yes, this is a huge ground connector right next to the SAI pump with a 13mm nut on it.You probably have to disconnect and move this out of the way when R&R the SAI pump.

Late last year for reasons unknown my Maf went bad when chasing down a vacuum leak cause I had the resonance flap and SAI side vacuum hoses reversed. Car ran fine but SAI pump never came on but had P1128/1130. After Maf replacement all was well. Don't understand the relationship. Just throwing it out there.

Thanks for the anecdote. If you had been able to use a Durametric in your situation it would have been useful.
For those trying to follow all the acronyms and code & with the same problem, here is a great link to Ahsai's photo of the system components.
MkII 996 vacuum line routing and related components - Rennlist Discussion Forums

Several months ago I changed my SAI Pump because it was whining. It worked fine, but it was obviously dying a slow death. I still have the whining one and can send it to you to see if the pump itself (even after the bench test) may be the issue.

Thank you for that kind offer Chuck. I have sent you a PM.

So this is your first starting of engine after rebuild? Is SAI the only one that won't go ready? Was it functional before rebuild? Sounds like a wire or connector loose if you had to pull the engine. Is car drivable? Think O2 sensors need to go ready before SAI too. Yeah, SAI =PIA

The engine runs fine after the rebuild.
The only obvious problem is the lack of 12v + power to the rear relay+fuse try. That prevents the SAI pump from running so an SAI readiness code is flagged.
The 1,2,3 Misfire codes are puzzling but one thing at a time.
I have a known good used SAI pump coming from Chuck (Thank you !). So that will eliminate one variable.
Meanwhile I have the alternator out to check wiring for continuity between the back of the alternator and the rear relay tray +40A fuse
When/if I find the wiring fault .I'll cut that conductor out and overlay a replacement as a test.The problem is access.
The looming issue is the 15mm,hex head bolt behind the tensioner pulley that is not staked.Common problem.Thinking of a mod to use a grade 8 or better carriage bolt instead but it is left hand thread !
Really need a hex head on it also(?) ,so may modify a hex head shoulder bolt .
.

Still easy to rule out a bad SAI pump would be disconnect plug and feed 12vdc directly to it momentarily. Should be 2 wire plug I see in all pics. Good luck.

Bench and insitu tests already done on the current Pump as noted above.
Problem may be too high current draw or some other obscure problem. Hence the simple substitution plan.I agree ,not logical but no harm trying.
The Power Probe Continuity test with a transmitter/receiver is next. Prime suspect is the huge fat harness & specifically the conductor between the back of the alternator and the rear relay tray+40A fuse holder. Don't want to unbundle that harness needlessly ! May be multiple issues.For example - until I get power to the relay tray , I can't test the rest of the circuit.Lots of poking and probing to do.
Thanks for your positive thoughts.

Reminds me on something Mercedes related. I know Porsche is different, but have you checked the relay contacts and the tray contacts of the relay. There is a lot of currency running when these SAI air pumps starts. So often the contacts in the relay melt which leads to misfunction.

Concerning Mercedes the SAI relays are located in a module called RCM module. The relays are mounted on a circuit board and the solder joints brake. But this is Mercedes specific.

Regards, Markus

Thank you Marcus !
I am hoping to have more diagnosis and tracking done this evening.I suspect a break in continuity deep in the wiring harness.

Thank you Chuck W for the SAI Pump
 

A few more hints for those who follow.
HOW TO FIX A NOISY SAI PUMP.
To test an SAI Pump you need to run it for 2 minutes continuously. In normal operation it runs for 60 -90 seconds. Chuck's pump was quiet for 2 minutes after I tweaked it .Mine was not.I had only tested mine for 10 seconds and presumed that was enough.
After you do the Krazy K rebuild of the pump it may run perfectly for 30 + seconds . Then it may start screeching. STOP ! If you are quick, you can save it.
Remove the discs and run just the motor.If it still screeches, you need to replace the bearings - as described in previous links.
If the motor is quiet ,it must be the discs rubbing. The prime suspect is the last disk to fit.The center lip in the middle of the disc rubs against the fixed disc in the final part of the housing. Look carefully and you should see the rub marks. Note the raised lip in the center ? Use a Dremel tool to grind away part of that lip. Make it perfectly smooth. add a little grease to the suspect area, try again.Yes, it is dentistry. RInse and repeat as necessary until the pump runs quietly. Test for 2 minutes.
Feel the blind end of the motor - is it hot or just warm? Hot = need new bearing.
Feel the volume of air being blown .It should be powerful and the air should be not much above ambient. If the flow is weak, check the air filter.in the end cap. Be very careful removing the tiny push nuts. I used a set of picks from HF. More dentistry ! If you loose one of the push nuts - they spring out wildly - you may be able to substitute an internally serrated washer of the right/tight i.d.
I mention this noise issue because I suspect a noisy pump is often condemned as having defective bearings when actually ir was the center of the last disc rubbing. I hope this helps others.

I do repairs like this almost weekly. My guess is that you either forgot to hook up the power wire to the 40 Amp fuse from the alternator when you did the rebuild, or you have a break or bad connection at the connections from the engine harness to rear harness. Wires pretty much never break in the middle unless they get pinched, etc., and generally you will have multiple issues and shorts and often smoke... That doesn't match your failure descriptions. That wire has no splices, so the likelihood of having a break in the middle of the run is virtually nil (barely a physically damaged harness of course).

Personally, I would first inspection the connection from the engine harness to the rear harness. If there is no power there, then I would disconnect the battery, and get to the back of the alternator and inspect. If you have to pull the alternator to get to it, so be it. My guess is that your problem lies there.

As for the SAI pump, if it is at all noisy, it must be disassembled, the bearings should be replaced, and the commutator turned and polished. Use the best quality bearings with good seals, not just skate board bearings. They don't last long.

One of the big issues with those pumps is that the carbon from the brushes gets into the bearings, first causing noise, then they seize and finally windings burn. If the motor has stopped or doesn't run full speed or has trouble starting, check continuity and shorting of the windings (I do on all of them). Most that have seized have damaged windings IME. If you don't polish the commutator (and brushes for that matter) and remove ALL the carbon dust, the carbon will get into the bearings (especially the one on the fan side) and cause the same problem again. These pumps are not a very good design, and they don't last particularly long. There are a couple of other issues, like exhaust getting into the pump, that is common on Audis that use the same pump, but I haven't seen it on the Porsches.

Thank you for the suggestions.
The fuse is fine. No power to the fuse holder terminals.
There is no power to the relay tray in the rear of the vehicle where the 40A fuse is located.
I removed the alternator and there is power there.I have started to trace for break between the back of the alternator and the rear relay tray.
Odd thing is the trunk light is still working .
Just plodding through the circuits with the help of Bentley.
Interesting comments about the pump motor- mus be a lot of them that are just thrown away when they could be repaired.
Interesting comment about the inferior quality bearings inadvertently used for the motor refurb. Do you have a spec & part number to suggest for others who do this diy ?

While you had power to the alternator, did you have the 2.5mm red wire along with the large cable that goes to the connecting point, and was it intact? You should check continuity between there and X59 (white connector) pin 14. On the other side of the connector it goes to a 4mm wire (because the Maxi fuse terminals are only available for that size) to the fuse holder. Like I said, in most cases you will find the breaks at the ends. That said, also check for chafed or pinched spots in the harness, or melted spots. I still will bet on the ends, either at the alternator, large multi-pin white connector in the back (check both sides of the connectors). or fuse holder (check from behind).

The ONLY thing that wire powers is the secondary air pump, so it is no surprise that everything else works.

Honestly, I haven't bothered to rebuild these pumps in a couple of years. If you want a core, I have a big box of them. For a while you could get new ones for less than $100, so it didn't make sense to rebuild them. That and people were throwing pumps at cars with leaking combi valves that killed pumps in a matter of weeks by filling them with exhaust vapors and expected warranties for that. I had made a prototype replacement with a brushless motor that eliminated the issues, but I lost interest in that project.

Thank you Qmulus ,Miles and others who Pm'd me. In the middle of the night I realized what the cause was and wrote myself a note in case I forgot . Yes, I know !
In Post 1 "...(alternator) Terminal B1. So this is where the problem must be ??"
Correct but why?
Qmulus must have been dreaming about this also because he also got the right answer. There are 3 conductors that connect on the back of the alternator : 1 blue that goes to the instrument cluster(IIRC) -it has it's own connector.
The other large terminal with the 14(??) mm nut has the starter wire and one red conductor. This red conductor with a ring terminal was not connected.Exactly what Qmulus predicted. Thank you ! I can still see the i.d. tag "ALT"I made for it dangling.
Qmulus made another good point about carbon contamination of the bearings and the need to clean out the carbon from the motor to prevent the fan discs sucking it into the front bearing of the SAI pump motor.
But how to free the stakes on the face plate of the motor.?? I used a Dremel with a grinding wheel but it was a tedious process and lots of abrasive debris.Maybe Qmulus has a better way to dismantle the motor?
Thank you ALL who helped me !

No problem.

For disassembling the pump motor, I squeeze the "ears" that were peened over back up, then use a flat chisel to work the end plate off of the motor housing. Doing it this way you can then peen the ears back down when you are done with another sharp chisel and it looks like no one was even there. If you grind away the ears, you have no material to keep the end plate on.

The actual process of rebuilding these pumps correctly is an involved and messy process. First you disassemble and clean everything (an industrial washer and or ultrasonic cleaner is ideal), then test the armature windings, turn and properly finish the commutator and brushes, replace bearings, reassemble and test for some amount of time (note current and airflow) . There are also some other steps that I consider "proprietary" that I learned from experience on these, but that is basics. This all assumes the motors are in good shape, when in reality about 50% have damaged windings and/or are badly corroded from exhaust gases. I don't bother with those, but they could be repaired. Luckily the orientation of the pumps in the Porsches (horiontal) is more favorable to these pumps than the vertical orientation used in other cars as the carbon from the brushes is not as likely to settle on the bearings. A brushless motor in this application would be ideal.

Anyway, glad to hear that you found your issue.

No, YOU found it- thanks very much .
I hope other Forum Members note you have a stash of cores that they may contact you about buying if you are interested.

Glad you got it figured out. I'll be up on the Crest or GMR waiting fer ya. See ya on the road.

Thanks guys !
So what did I learn and why did I make the mistake? I'll confess in the hope it spares others similar problems.
1. Durametric is an invaluable tool for diy ! It eliminates a lot of speculation.
2. I need to print a good copy of all the electrical component locations that 78F350 posted Is it from PET ?
3. When you do a rebuild, take and annotate lots of photos and organize them. In addition label the components and mark their orientation. You may have a great memory but if life hits you in the middle of a rebuild project (my unexpected 2 year hiatus problem), you will forget the details and the writing on the labels fades,the labels fall off . The best ones are the flag type with the wire tag on them and permanent marker. Zip lock bags all marked too. This was my downfall !
4. Read everything ,bookmark and organize it. This makes it easier to help others. I have almost 1500 posts here and on R'list that are almost all answers using links I culled for my own projects - a few clicks and you can share the wisdom and experience of others . Unlike Qmuluus , I have little proprietary content to offer. Thanks Q.
5. Access - unless you have "keyboard fingers" ,skinny arms and Kevlar-arm gloves this car is hard to see/access. A powerful ,skinny Maglight is invaluable.I just wish they had a "gummy"version you can bite on. Yes, sometimes you need 2 hands and a light so the flashlight is held by your teeth! (Ignition switch r&r for example) I also used a Kobalt flex-neck flashlight with a tiny ,powerful LED bulb. For under car inspection I use a home-made Miners helmet with a powerful light on it.
An oddball suggestion I gleaned from an unrelated post by Member San Renso - remove the trunk lid and the top cover, kneel in the trunk to work on the the engine from above. It isn't as mad as it sounds if you have a lot to do there and maybe the lift struts need replacing anyway.
Hope this helps others who may follow and thank you for all the help given.

Glad it is useful. The component locations are in the back sheets of the wiring diagrams. The one I posted from was 'Sheet 19 - Connectors and abbreviations'.

Correction/update:

The component locations I posted were from a PDF I have that is a partial copy of the Porsche Service manual. It only contains the wiring diagrams from 1997 - 1999. The link I downloaded it from is dead now, but I'm sure it's out there on the net still. The actual page was:'Model 1999, Sheet 19 - Connectors and abbreviations'.

I just checked through my hard copy manual (15 binders 2 of which are pull-out wiring diagrams) and saw that it's laid out a little different each year: Sheet 18 for the 1997, Sheet 19 for 98-99, Later model years it's called "Plug Connections, abbreviations" and ranges between sheets 21 - 23.

I Just looked through the Bentley. It has the information, but it is presented differently. It has the X-connectors and their locations stated on "EWD-8", on the page before the actual wiring diagrams start. There's some more useful info on how to read the wire diagrams in the pages before that, "EWD-3", "EWD-4", etc....

Thank you for finding it. Bentley certainly did hide it well.
I have been trying to get a CD version of the FM with the Supplements/updates but the best I can find is a 1997 manual with updates as far as #16 .There is no date for any of the updates so am clueless as to if it would be useful.

A few notes to others who may try this .
I found it very difficult to remove the motor faceplate gently.I did squeeze the stakes as Qmulus kindly suggests below. It went easier when I ground back the obstructing 'ears' and edges. Remove only as much as is necessary. There are 2 Torx screws that help secure the front plate so I was not worried about reassembly with slimmed down stakes.
I was worried about damaging the carbon brushes so removed the springs- that was fun replacing them !I used an Exacto blade to hold them compressed in position while bending back the 4 retaining tabs.
The bearings are(?) 608 2RS .2RS Not ZZ ! I measured the o.d. as 22mm, the i.d. as 8mm and the thickness as 7mm.As Qmulus suggests, this is a popular size for skateboards but better quality are available- even ceramic on Amazon!
The bigger problem is the stakes around the mid point of the circumference of the body of the motor. They are very thick.I had no way of bending them without risking damage & debris inside the motor. Perhaps they could be weakened with a Dremel wheel and then bent out of the way? Q was right ,there was carbon debris so I removed what I could with a powerful shop vac.
My suggestion is that if the motor is noisy - find a replacement motor - but where? It has no part number or manufacturer name on it. Must be available because the same motor is used on many other vehicles.
WRITEUP: 2.7T Secondary Air Injection SAI Pump rebuild - AudiWorld Forums
Maybe the VW/Audi?Mercedes Forums have solved this issue??
Reassembly - you may need pushnuts - the 3 that retain the filter housing? Here is a possible source Aaronspushnuts. FYI ,they are approx. 12mm o.d. and 2mm i.d.:
** PUSHNUTS, push on nuts for bolts or unthreaded rods **
There is also an External circlip 15mm i.d. I do not know what function it has other than to make a scraping noise. Chucks Pump was missing this part but worked fine w/o it.

Great write up Tony. Nicely done. The pump I sent you had about 28,000 miles on it. It was making a whining sound for a couple months before I replaced it. I had read the various threads regarding rebuilding it and some of the rebuild failures. Audi uses the same pump and they had posted some rebuilds too.

Bosch sells them new in the $300 range while the Porsche dealer sells them for $1700.

Then I looked at qualityporscheparts on eBay. They had (have) over 50 pumps for sale (Porsche Part No. 99660510400). Their prices vary depending on the mileage of the Porsche they came out of. I paid $65 for one with 49,000 miles on it. For me, it was worth the gamble of buying a used one for $65 verse trying a rebuild.

One thing I noted is the top comes off very easily by removing three compression nuts. Once removed you can see and remove the air filter. Both, mine with 28,000 miles on it and the donor one with 48,000 miles were filthy with fine black soot. When I cleaned the original one it ripped it. They are very fragile. When I cleaned the donor one I used Dawn dish-washing soap I was very careful and it came out well.

I agree with Chuck's recommendations . Half the battle is knowing how far to go "down the rabbit hole" with these diy projects.
Until we find a source of new or rebuilt electric motors for the Secondary Air Pump, our options are very limited.
Someone in The Audi and VW world must have a source for these little electric motors?