Vacuum Problems Can Really Suck

Everyone loves a happy ending. Share your diagnostic success stories here.
Avdr
TwoWaves
TwoWaves
Posts: 157
Joined: Sun Jan 08, 2012 9:25 am

Vacuum Problems Can Really Suck

Post by Avdr »

The Vehicle:

Vauxhall Vectra 2.0DTI
2001 model with Y20DTH and Bosch MSA 15M management

The Customer Complaint:

Once slightly warm the Engine management light would come on once the vehicle had come to a stop and had returned to idle. This would result in loss of power upon pulling away, right when it was needed to clear a junction. Within moments the fault would clear and the MIL would go out, until the next time the car came to a stop and was allowed to idle.

My Analysis:

Read DTC’s – P0400 EGR FLOW MALFUNCTION
Attempt to re- create problem in the workshop, after extensive trying, the fault could not be forced. Went out and drove the car and confirmed the fault was as described, EML was set during test drive.
The detection of the fault was putting the vehicle into limp mode, which gave rise to the loss of power.
Noticed that while stationary with the EML on, a quick blip of throttle would result in light going off.

Egr System Overview:

A poppet valve covers an EGR port within the inlet manifold, by means of a diaphragm and vacuum the valve can be lifted, allowing the mixing of exhaust gas with fresh air in the inlet manifold.
The vacuum to the poppet valve is controlled by a solenoid which receives a VPW to ground, so regulating the vacuum allowed to the poppet valve.

As there is no direct way of measuring the position of the EGR valve, the ECU monitors the amount by which the MAFM is being offset by mixed in exhaust gasses. In this way the duty cycle to the solenoid is varied to achieve the target amount of exhaust gas.

Tests Carried Out:

Verify the diaphragm in the EGR valve was capable of holding a vacuum without leaking. OK
Swap around the vacuum solenoid for EGR with vacuum solenoid for swirl flaps. NO IMPROVEMENT

Hooked up the pico and went on a test drive;

CHA - VPW to EGR solenoid
CHB – MAFM Signal
CHC – APP Sensor – to indicate when vehicle was at idle.

I set up the above test to verify that the signal from MAFM was being offset in unison with the ECU command to the solenoid, and not by a random valve opening that the ECU had not commanded. I was a little pushed for time hence not connecting to the crankshaft sensor for RPM. Pedal position was more accessible. In hindsight I set the time base a little long and lost detail as a result – but I didn’t know how long it would take the fault to appear, as it was not occurring at every junction.
vec long 23m.jpg
vec long 23m.psdata
(2.41 MiB) Downloaded 751 times
The test proved that the valve was working properly and under full command of the ECU at all times.
Unfortunately I don’t have any form of pressure transducer so was unable to use pico to directly measure the vacuum at the EGR valve.

I’ve had no prior experience with vacuum faults, and had no idea what kind of vacuum I should expect a vacuum pump to create. In the course of testing the EGR valve diaphragm I found the vacuum at idle to be -16” of mercury. I began to compare this to other diesel’s I had access to and found a navara gave - 22” Hg at idle and a mondeo achieved a full -30” Hg.

I began to suspect the vacuum pump of being weak, which would tally with my finding that blipping the throttle seemed to clear the fault – this acted to increase the vacuum. It went some way to explaining why the fault only ever occurred at idle.

I sourced a replacement pump from a breaker – in unknown condition and fitted it, only to find the vacuum remained at a miserable -16” Hg. Feeling it was unlikely to have two duff pumps I had a brainwave, I had already verified no leaks existed in the tubing so far as the EGR control, Turbo waste gate and swirl flaps control systems. I had however over looked the brake servo, in a moment of inspiration I disconnected the servo and plugged the union on the vacuum pump, while monitoring vacuum on the EGR system (the vacuum pump has two fittings one for servo only and the other for all other vacuum systems) instantly the vacuum increased to -24” Hg.

I left the union plugged and took the car on an extended test drive, the fault did not re-occur, finally I had sussed it!

In Summary:


I have certainly learned from this job, and in future will think to check the servo for leaks also. The replacement vacuum pump cost £20 so it’s not too hard to bear.
Having got to the bottom of the fault I was in a better position to understand the reason for the intermittence – I concluded it was due to how often/forcibly the brakes were applied to bring the car to a stop, so affecting how much air the vacuum pump had to remove from the servo in a short space of time while the engine was doing only idle rpm.

It also left me in a better position to understand the relationship between EGR solenoid duty cycle and amount of exhaust gas actually admitted, this was something I had studied through serial data early on – but with no technical information it was pretty meaningless data.

Below are two screen shots;
The first taken when the servo was still connected to pump, and having just applied the brakes; it can be seen that a pulse ratio of 65% was required to produce enough opening of the EGR valve to achieve the target EGR amount.

The second, with the Servo union plugged required a greatly reduced duty cycle of 28% to attain the same amount of EGR. Bringing the duty cycle within tolerance for the ECU.
du02 fve servo connected.JPG
du02 fve servo blanked.JPG

User avatar
Robski
Advanced User
Advanced User
Posts: 598
Joined: Tue Sep 14, 2010 6:36 pm
Location: Yorkshire

Re: Vacuum Problems Can Really Suck

Post by Robski »

I was a little pushed for time hence not connecting to the crankshaft sensor for RPM
Can you elaborate a bit more on that ?
In hindsight I set the time base a little long and lost detail as a result – but I didn’t know how long it would take the fault to appear
Not exactly true, not enough sample rate as well, detail was lost when zooming in.You would have been better shortening the time base & adding more sample rate & dialing in on the properties tab.Then rather than all on one screen you would have had buffers.
I tried dialing in at that time base & the lowest sample was 34us iirc & Pico crashed :?
I left the union plugged and took the car on an extended test drive, the fault did not re-occur, finally I had sussed it!
With no servo assistance to the brakes :shock:

Buy a smoke machine :wink:
Bringing the duty cycle within tolerance for the ECU.
Was there a 'desired' PID ?

Avdr
TwoWaves
TwoWaves
Posts: 157
Joined: Sun Jan 08, 2012 9:25 am

Re: Vacuum Problems Can Really Suck

Post by Avdr »

Car belonged to a builder from the firm next door. I had spent all morning with it in the workshop with no success, it was getting close to 3pm when I hooked scope up for the road. The plug for APP was right there, in the car easy to see, the CKS was somewhere else, I had not yet discovered where. As I said, it was quicker to connnect to APP.

Sample rate was 1MS/s. As already stated I couldv'e set the scope up better. Isn't hindsight wonderful.

Yes with no servo assistance to the brakes - I go to the gym so find my right foot still capable of pulling up in a hurry.

There was no specific PID re. duty cycle. This statement was my own take on the situation. Had there been a PID for EGR duty its unlikely there would've been a case study!

This being a case study I would have thought it open and shut, I went about the job in the method above, I have already stated things could've been done differently and that my initial pump diagnosis was incorrect. I didn't take the time to write it to invite people to question my method - or to help me decide how to spend my money.

User avatar
Robski
Advanced User
Advanced User
Posts: 598
Joined: Tue Sep 14, 2010 6:36 pm
Location: Yorkshire

Re: Vacuum Problems Can Really Suck

Post by Robski »

As I said, it was quicker to connnect to APP.
Why were you going to probe CKP that's what i questioned really ?
Sample rate was 1MS/s. As already stated I couldv'e set the scope up better. Isn't hindsight wonderful.
I saw that & that's why i mentioned what i did, i set the scope up before i go so that it's ready to catch at fault & also when saved i know i can zoom in & analyse without too much loss.
There was no specific PID re. duty cycle. This statement was my own take on the situation
I did wonder as i can't remember seeing a 'desired' ever , thought you might have got that info from Global TIS or the like.
This being a case study I would have thought it open and shut, I went about the job in the method above, I have already stated things could've been done differently and that my initial pump diagnosis was incorrect. I didn't take the time to write it to invite people to question my method - or to help me decide how to spend my money.
Not once have i mentioned your method, i have questioned a few statements & also offered advice on scope set up & thought it a tad dangerous doing an extended road test with no servo assistance, I've needed an under-cracker change towing before & that's at 30 MPH & less !
I'm not trying to spend your money, i think you would have found the fault with a smoke machine & no road test required without servo assistance, is the pedal any different now the fault is fixed, was it harder before ?

I questioned the CKP hook up as you might well have taught me something or given me another angle to look at when i'm faced with similar problems along with other users.

No need to get cranky mate, when you do a case study or any public thing like this you are open to cross examination !

Just as i knew the first reply to my ignition one would be about NGK spark plugs, the car is still running faultless on NGK plugs & i will stand by & argue about my experiences with NGK plugs.

Avdr
TwoWaves
TwoWaves
Posts: 157
Joined: Sun Jan 08, 2012 9:25 am

Re: Vacuum Problems Can Really Suck

Post by Avdr »

My apologies for weighing in heavy with my reply. Your comments tend to be short in the first instance and can easily be read as a bit of a virtual slap down for doing something in a manner you wouldn't have used yourself.

I only mention CKP/APP as an indicator of where I had taken my foot off the throttle, so when I went back through the data later on I could pick out the points at which the car would be at idle - as it was only these times that were of main interest.
I have yet to fit the replacement servo, and will gladly get a CKP trace if you are in need of one.

The trace gave me just enough info to see the increased areas of duty cycle did all correspond with decreased MAFM signal. Had it been inconclusive in some way I would have had to re-set up the scope for a better picture. In future I will know first time out to set the scope up differently.

It honestly wasn't that hard to brake, I must have big quads or something.

RE: NGK plugs, I don't know what other people's take on them is, but given their company is the size it is. It would seem a fair assesment that they must be doing something right.

User avatar
Robski
Advanced User
Advanced User
Posts: 598
Joined: Tue Sep 14, 2010 6:36 pm
Location: Yorkshire

Re: Vacuum Problems Can Really Suck

Post by Robski »

and will gladly get a CKP trace if you are in need of one
I don't require one ?

I asked why you wanted the CKP captured so you would explain what you were expecting to see/do with the waveform, as like i said you could have educated me & others to something we might be over looking or could use in the future !
Your comments tend to be short in the first instance and can easily be read as a bit of a virtual slap down for doing something in a manner you wouldn't have used yourself.
I didn't see my first question as being in this manner ?
Can you elaborate a bit more on that ?

Avdr
TwoWaves
TwoWaves
Posts: 157
Joined: Sun Jan 08, 2012 9:25 am

Re: Vacuum Problems Can Really Suck

Post by Avdr »

I got the wrong end of the stick there about CKP trace, nevermind.

Granted in this thread your comments were not strongly worded, but in other threads (mine and others) they can be read in the manner I mentioned - by me atleast.
This is beginning to look like I'm trying to do a demolition job on you - which was never my intention.
Perhaps the biggest flaw in the internet (in text form atleast) is the complete loss of context within a conversation. We are all just reading plain text with no indicators by which to gauge the context, ie. a tone of voice or facial expression when compared to the phone or a face to face conversation.

Andy

GRUSS
TwoWaves
TwoWaves
Posts: 119
Joined: Tue Dec 27, 2011 11:16 pm

Re: Vacuum Problems Can Really Suck

Post by GRUSS »

Nice little case study Andy, I never had a problem driving my Elise with no servo assisted brakes :D

How do you find the texa idc4? When your looking through the live data screens is it slow to change pages? I think my laptop needs more memory.

User avatar
Robski
Advanced User
Advanced User
Posts: 598
Joined: Tue Sep 14, 2010 6:36 pm
Location: Yorkshire

Re: Vacuum Problems Can Really Suck

Post by Robski »

I never had a problem driving my Elise with no servo assisted brakes
Maybe VOSA need to restructure the MOT & brake tests then if they're not so important :|

Alan
Pico Staff Member
Pico Staff Member
Posts: 264
Joined: Thu May 25, 2006 8:02 am

Re: Vacuum Problems Can Really Suck

Post by Alan »

Thanks for the case study! Hope you do not mind but I added a screenshot of the waveform to your post.

I guess this is way off topic now but the S1 Elise did not have a brake servo but the slightly heavier S2 from 2002 did.

Post Reply