Xantia V6 broken exhaust (update: and rough low rpm running)

[addo]

You'll have to make one. Buy a male O2 bung and then tap it to 1/8" NPT.

[lexi]

With no shelter and a wind from Scandinavia that has been blowing here for weeks I don't blame ya! 2 deg. and windchill feels like -2 ........springtime

Set to continue into next week.......
Newsflash!........A "thing" has appeared in the sky today from 96 million miles away, WTF is that?

Subjects of Scotland, never mind this Pope stuff......get down on yer knees and worship the Sun God RA, before the beggar goes for forty days and nights again

[Mandrake]

You'll have to make one. Buy a male O2 bung and then tap it to 1/8" NPT.

I don't have any tools to drill and tap a thread properly, so that's probably a non starter.

This looks promising though:

http://circuitse7en.net/fantcube/index. ... oductId=14" onclick="window.open(this.href);return false;

[Mandrake]

With no shelter and a wind from Scandinavia that has been blowing here for weeks I don't blame ya! 2 deg. and windchill feels like -2 ........springtime

Set to continue into next week.......
Newsflash!........A "thing" has appeared in the sky today from 96 million miles away, WTF is that?

Subjects of Scotland, never mind this Pope stuff......get down on yer knees and worship the Sun God RA, before the beggar goes for forty days and nights again

Tell me about it. Wind chill indeed... and its predicted to snow tomorrow.

[addo]

Good grief - no vice, no cordless drill? Don't you get lonely at night?

Won't that piece of kit be £20 or more by the time it arrives?

I still believe you'll have the most flexibility with a separate boss for the pressure, as you can drive with all systems operating "normally". A weld-in boss and matching plug (that I'm sure you could get help with drilling and tapping of) is about USD$16, posted to the UK.

[Mandrake]

I have a small cordless drill but no vice and nowhere to put it either. Likewise no welding plant or place to connect power tools within reach of the car. Such is life in a 2nd floor apartment which is not a bachelor pad...

I'll email them and see what they quote, I'd estimate around £20 ish including shipping. I can't get it until the end of the month anyway so I'll keep looking to see what else I can find meanwhile. Seems perfect for the job to me, steel is much better than copper in this application as copper is too good a conductor of heat while steel is a poor conductor, its also annealed so I should be able to bend it such that it points upwards behind the back of the engine to attach to a flexible hose.

£20 might sound like a lot but it beats £120 by a large margin, which is the cheapest purpose built back pressure gauge with adaptors that I've seen so far.

Edit: Just noticed that this kit also includes a threaded boss and blanking plug if I did want to go the route of welding a separate one in, so I would have all bases covered with this one.

[addo]

If it includes those bits, it's a goer. The welding will be cheap enough and even if the cat is rogered you can transplant the inlet pipe (with added outlet) onto a new cat body.

[lexi]

If anyone local is making stainless exhausts near you, they will weld that together for a bung. It would cost you that in fuel to come over here to weld it. I am out of gas mind for Mig Tig and only have the stick Inverter (thank god) ATM. If it is thick I can weld it..........I have some nice 316 rods. If it is light, nah!

Snow and blizzards here today...........even at sea level.

[Mandrake]

Thanks alex, I should be fine to just do the test through the oxygen sensor port as the oxygen sensor needs replacing anyway, but its good to have the threaded boss and bung as well for future use...

In any case I ordered the adaptor on thursday, but it will be 10-14 days delivery. (No matter, its going to take that long before the weather is fines up enough to be climbing around underneath! Brrrrr )

[Mandrake]

Quick update, I've ordered the previously discussed exhaust system from eBay (middle silencer, back box, connecting pipes and fittings, basically everything from the cat tail joint back) which should be here middle of next week.

I don't know yet if the cat is blocked or not but the rest of the exhaust is pretty stuffed and is getting noisier by the week, (and is possibly blocked) so it needs replacing regardless of whether I'll also need a cat.

[RichardW]

It will need to be modified to fit your cat of course....

[Mandrake]

Yep

A relatively easy modification though, and one that has proved to be strong and air tight... I still have 3 more of the same type of clamp we used for the modified joint as well, from the joint kit I had that we didn't use...(apart from the clamp)

[Mandrake]

I've been doing a bit of research recently on how speed-density (MAP) engine management systems work, and what the possible effects of exhaust restriction might be on the calculation of air flow and have made an interesting discovery backed up by a practical test performed today...

A quick recap on MAF vs MAP systems, as the two respond very differently to an exhaust restriction from the research I've done.

A MAF (mass air flow) sensor is usually a hot wire/resistor in the air intake before the throttle plate that directly measures the flow of air into the inlet manifold in grams per second, from this the ECU can calculate the precise amount of fuel required to achieve the desired 14.7/1 air fuel mixture. An exhaust restriction on a MAF system will reduce the volumetric efficiency of the engine since it can't draw in as much air/fuel per stroke, but the MAF sensor will correctly indicate the reduced air flow (since it measure it directly) and reduce the injection pulse as well to keep the mixture correct. The car will be down on power but will otherwise run ok(ish).

Not so with a speed-density (MAP) system. The ES9J4 V6 like many petrol engines is a MAP system meaning it measures manifold absolute pressure but does not directly measure mass air flow - instead it estimates it based on engine RPM, pre-defined volumetric efficiency of the engine, manifold air pressure and intake air temperature - all of these must be correct for the calculation of mass air flow to be correct and therefore the correct fuel injection to achieve 14.7/1 to be correct.

The way it calculates it is to first measure the manifold absolute pressure and air intake temperature - these two figures allow the the density of the air in the intake manifold to be calculated, it then looks at the engine RPM and based on a lookup table of volumetric efficiency vs air density and RPM it works out how many grams of air are drawn into the engine per revolution - thus mass air flow in grams is indirectly derived.

An additional correction comes from the TPS (throttle position sensor) to allow for the fact that the MAP sensor output is low pass filtered (approx 100-200ms delay) and thus has a slightly delayed response to sudden throttle changes - if only the MAP sensor data was used, opening the throttle suddenly would cause the engine to take a gulp of air but there would be a delay before the MAP reading increased and the injector pulse increased, thus you'd get a lean stumble. The sudden opening of the TPS triggers the ECU to give the injectors an instantaneous but temporary increase in pulse width for a few hundred milliseconds to richen the mixture while the MAP reading "catches up", at which point the MAP reading then takes control again.

Therefore the MAP reading has a fairly direct effect on the injector pulse width - in normal operation for a given RPM a higher MAP reading (higher pressure, lower vacuum) corresponds to a greater throttle opening and therefore more air flow thus more fuel is required so the injector pulse widths are increased. The MAP combined with RPM is also used to calculate the "engine load" figure, because a higher MAP reading (and throttle opening) while maintaining the same RPM means that the engine is under more load. The engine load figure is passed to the gearbox to help it decide on gear change strategies, and is also used by the ECU to avoid detonation - under high load at low RPM it will start to over richen the mixture and retard the timing to avoid detonation. (pinking)

So what would exhaust back pressure do to a MAP system ? From the reading that I've been doing and also thinking it through myself, exhaust back pressure would cause a lower than normal vacuum under partial load conditions - eg not full throttle (which is always zero vacuum) and not closed throttle (which is always high vacuum) but in the normal working range in between with partial throttle openings. Thanks to addo for pointing this out many many pages ago as well...

A lower vacuum reading is higher pressure, in a MAP system the higher manifold pressure under partial loads is falsely interpreted as an INCREASE in air flow (and increase in engine load) when there is actually a REDUCTION in air flow - the reduced vacuum and the partial blockage in the exhaust are obviously going to reduce the amount of air flow through the engine - the volumetric efficiency goes down because its no longer as good of an air pump...

However the speed-density system makes an ASSUMPTION that the volumetric efficiency is what it is when it leaves the factory when it calculates the air flow, a blocked exhaust makes this assumption invalid. This assumption is in the form of a lookup table of air flow versus engine RPM versus air density which matches the characteristics of a new engine and exhaust system.

Rather than reducing the fuelling to match the reduced air flow the response from the ECU is actually to over fuel the engine causing it to run very rich under partial loads, (killing the MPG) but at the same time the unusually high calculated engine load causes the ECU to RETARD the timing to protect the engine and avoid detonation...(killing the performance)

When I put my foot down to 50% or so throttle at 2000rpm in a higher gear the ECU is calculating an extremely high engine load because the vacuum is not nearly as high as it should be under those conditions, thus it over richens the mixture to the point that the engine is choking and stumbling on a rich mixture and at the same time retards the timing 20 degrees or so from what it should be.

All nice in theory, but how to prove that the back pressure is upsetting manifold vacuum and confusing the ECU ? What would happen if I disconnected the MAP sensor I wondered, would the car even run ? I've done it by mistake once before and it did run although it stumbled a bit when first opening the throttle, so I gave it a try.

At first it wouldn't start but after 2-3 tries the ECU realised the sensor was not working and obviously went into a backup strategy, letting the car start and idle, so I took it for a drive. The backup strategy usually applied from what I've read is simply that it estimates the air flow based on engine RPM and throttle plate opening, there will be a lookup table that provides a "simulated" MAP reading for different combinations of RPM and throttle plate opening. (again making assumptions based on a known engine volumetric efficiency)

Well, what a revelation. The very laggy unresponsive throttle was gone, if anything its a bit too abrupt and responsive now, when first taking off it pulls away very sharply and abruptly, likewise if you have moderately wide throttle at higher rpm and you suddenly shut the throttle it backs off very abruptly causing a bit of a lurch. It also starts a bit funny - it doesn't lift the RPM above 1000 while starting then let it drop, it starts at around 500rpm and then slowly climbs up to idle after starting, so cold starting might be a bit of a problem without some throttle.

Apart from that it behaves quite well, the idle is very steady, and apart from a tiny stumble when the throttle is just opening the throttle response is very sharp and there is LOADS more power across the whole RPM range. It is far more responsive between 1500-2000 where it was really struggling before, it now actually has low down pickup. Power at the mid and high rev range is also very good. I still don't think its performing to full potential but its overall far more responsive and has more power than before, and more importantly its far more drivable.

I also notice that the gearbox is far more willing to change down a gear without excessive throttle openings, which I think goes to show that the incorrect engine load figure that is sent to the gearbox ECU is indeed the reason why it has been reluctant to change down.

I still have the original MAP sensor that I replaced, I'm pretty sure that neither the original one or the new one are faulty - I could try swapping it back but I suspect it won't make any difference, so I think the fact that its got so much more power with a known good MAP sensor disconnected than when its connected shows that the intake vacuum under partial load is out of the normal range and that its probably due to an exhaust restriction as I don't think there are any vacuum leaks on the inlet side.

Although the exhaust back pressure would cause a loss of power due to lowering the volumetric efficiency of the engine (it can't breathe properly) it appears that the major portion of power loss is in fact retarded timing and possibly an over rich mixture to the point of choking the engine, all due to the incorrect MAP reading caused by the probable exhaust blockage...

There's obviously not enough of a blockage there to strangle the engine completely as it had a LOT of power today even at wide open throttle at high RPM... but theres enough to upset the vacuum reading.

One other thing I noticed is that previously under wide open throttle below 3000rpm the oxygen sensor was reading lean - its now reading rich under the same conditions, as it should under wide open throttle, although I'm not entirely sure how that fits in with my diagnosis of MAP pressure readings...

I'm tempted to just leave the MAP sensor disconnected until I've dealt with the exhaust - the drivability and performance is so much better...

Although it logs a fault code it doesn't light the EML on the dashboard - it seems that short of catching fire nothing will make this ECU light the EML.

[addo]

You can permanently illuminate the EML by unplugging the canister purge solenoid. (An item that has nil impact on performance.)

[CitroJim]

Very interesting indeed Simon, still digesting it

Something else that will light the EML is a duff coil.

A duff oxygen sensor doesn't illuminate it either...