Graphite Green 9-5 Aero rescue

[carrera]

Yesterday I took the car for an aircon regas. That worked, so now the ACC is working correctly, we'll have to see how long this lasts, there was 100g of gas in there, but still getting 3.5 bar pressure.

On the way I noticed that the engine was pulling better, the boost gauge went into the red more easily and it seemed to be running smoother.

Am I imagining things, or could the old CPS have been on it's way out and causing a reluctance to boost at lower speeds. If so I'm trying to understand the failure mode

The breakdown may have just been the CPS then ?

[sgould]

CPS triggers the spark.  If it's failing, it might miss a few or send the signal "late".  In which case the various sensors might not see a complete match of all the parameters.  But I'm guessing, as usual.

[carrera]

She's being a bit of a tease at the moment.

Running fine, but not wanting to deliver the full beans.

I read the ECU a couple of days ago and enabled opensid so I could see more of what was going on.

Airmass request is up at around 1130mg / combustion, which is about normal, but the amount delivered is no more than 900 mg /c.  Boost is only about 0.55 bar, which isn't much more than base boost. The PWM for the control valve was increasing at high airflow request, so that appears to be electrically functioning. The dump valve is holding vaccuum.

The fuel filter was changed yesterday, and though it was not fresh, there were no signs of blockage and it made no difference. I tried the "bolt mod" on the pipe to the cobra with no change and also bypassed the control valve so boost was purely mechanical. At least I think that was what I did. Again no change.

I don't think I have an intercooler leak, the pipes seem OK and I have no reason to suspect that the turbo has suddenly lost the ability to provide flow.

I'm scratching my head a bit on this now, she drives perfectly well with no fault codes, just a bit short of breath. 

[fka]

must be a boost leak somewhere or something's opening that shouldn't. BPCV mechanically not working despite PWM appearing to be ok?

[carrera]

Well that's encouraging as that's exactly the areas I was thinking about. I tried to bypass the BPCV with no difference, in fact I did go to swap it out and then found it is help onto the metal plate by rivets not screws. Why, oh why oh why Saab ?

I might just chuck the whole spare engine loom into the engine bay for a test

[carrera]

There was not enough room to chuck the spare BPCV in with the loom so I removed the rivets and swapped it over. No change, so my BPCV is good.

I removed the MAF in case there was a blockage, and noticed that the lower rubber section where the outlet pipe connects to the cobra was bent over. Ah, potential for unmetered air. Fixed that, mReq increased to 1200 but Mair remained at 900. Little tease, more potential, but no delivery.

I also put an airline onto the cobra having blocked off the intercooler pipe to the plastik intake pipe. I didn't detect any leaks.

More thinking required, and I probably need to learn how the T7 valve works

[fka]

Are you talking Bypass Control Valve or Boost pressure control valve, APC? I meant the APC.

[carrera]

 

So many things with similar names where the descriptions people use are different to Saab WIS. I'm easily confused.

I was talking about the control valve on the bulkhead, that manages over run situations and according to WIS can cause low Mair if it is failing.

The "APC" / T7 control valve on the top of the cobra I have not swapped, but I did remove the top pipe and block the connections with a bolt on that pipe and the cobra.

No difference so I assume it's working correctly.

I need to re-remember the bolt mod I saw on Saabscene about 15 years ago with the one way valves, one of them is on the right hand wing. I had that problem on the blue car. I thionk that's more related to the EVAP part of the system, but it has the same effect

[sgould]

The bolt mod on the APC is a bodge that tries to mimic the removal of the second EVAP non-return valve.  Saab removed the second valve for 2006 onwards.  It won't improve things if the remaining valve is leaking...

the picture shows the pipe to remove.

[fka]

The "APC" / T7 control valve on the top of the cobra I have not swapped, but I did remove the top pipe and block the connections with a bolt on that pipe and the cobra.

It's been a while since I've looked at one but isn't the top port the return, going to the cobra? You'd need to remove either the collector (bottom, that goes to the compressor housing) or wastegate (middle, that goes to the actuator).
Then in theory you should hit overboost fuel cut when you give it some welly.
0.55bar is close enough to be base boost.

[carrera]

I really need to get my head around how it all works.

Yesterday I was reading other threads for suggestions, I'm fairly sure one had contributions from Karl Noobtune, while the other was over on Saabcentral, so was just following instructions really.

The illustration above is one I recognise, I thought that is what I had done, but I am not so sure now.

It seems there are any number of bolt mods too 

[fka]

Again it's been a while since I tried the bolt mod on my 9-3 (I ended up removing the hose highlighted in Will's pic as it's a slightly more elegant solution). What you did is probably correct, assuming the boost leak is due to a failed non-return valve on that line to the EVAP system.

The APC is normally open (C port to the W port) - i.e if it's not working at all (When on boost it should be diverting the pressure from the compressor away from the wastegate, to the return port) you'll get base boost.
As you know, the signal to it is PWM, the higher the PWM number on openSID the harder the APC (and ECU) are working to to try and divert pressure away from the wastegate. You'll see PWM hit 980 when you floor it and then it should slowly level off. If you're not achieving MREQ and PWM stays at 980 then the APC is incapable of controlling the actuator/ achieving target air mass . Which could be a weak actuator, or at higher target air masses - that the return port is too restricted or the the C port not restricted enough.. or a boost leak..

[carrera]

No chance to do any work on this yet, but I did notice something on the way home.

Only in certain circumstances do I see a PWM of 980, and it then instantly drops down to something around say 660-710. It'll then bounce around a bit and move upwards as the engine speed gets to 4000 rpm.

Idle the value is 20, and light throttle to maintain speed at 55mph is about 120.

I also spotted the Mair tracks Mreq at demands up to about 750, after than Mair lags.

It seems as though the demand to increase airflow by allowing boost is being given, but then something is pulling it back.

I have already done another map to make ECMStat available so I can see what limit is in operation, but I have not had the opportunity to upload it yet.

If there was a leak from the intercooler, then would the MAF record the increased airflow, but then the leak would cause lower airmass in the intake and the MAP sensor would correct the fuel quantity delivered, or would it just run rich.

[fka]

You should see PWM = 980 when you WOT, then settle down (according to Karl) to around 700>800. Then climb to 980 again at the top end of the rev range. So yours sounds about right.
I'd be interested to see what it does if you reset the ECU and try and go through adaptation again. I'd wonder if it would just stay at 980 and not be able to adapt. Although it's probably quicker and easier to just chuck a known good APC on it.

I don't think it's too odd that MAir tracks MReq to 750, you'd probably see that behaviour if you connected the actuator direct to the compressor outlet.

Just out of interest, have you watched the knock count when you put your foot down?

[carrera]

Ah, I pulled fuse 17 a couple of days ago after swapping the bulkhead valve over, just in case it had adapted to my recent gentle driving. No change in performance.

I don't think I have seen any knock count, best go looking for where I can find it and see what it is doing