Overboost after wastegate porting!

[vi11ain]

Right then. I've replaced the hoses between the turbo, the wastegate actuator and the boost control frequency valve (BCFV), so I'm pretty certain that they're not leaking now. I've also studied my photos of the turbo before removal etc and the compressor housing appears to be in the correct position.

I've just taken the car for a test run with ElanScan and (unfortunately) managed to provoke fuel cut a couple of times due to boost creep. I can see that it's regularly boosting to around 0.75 bar, which ought to be impossible anyway, but going at it harder makes the boost inexorably climb above 0.9 with the inevitable result.

The thing that's puzzling me at the moment is that the wastegate duty cycle never seems to rise above 25% - in fact, at around 580 seconds into the scan the duty cycle is 0% even though the boost is over 0.9 bar! Can anyone shed any light on this, or am I misunderstanding the data in some way?

I'm attaching the zipped ElanScan file so you can see what I mean.....

Over the weekend I plan to disengage the actuator to see what happens then.

Ben

[esprit888]

Ben,
Would a bad MAP sensor or leaking hose to the sensor give you the symptoms you have ? If the map sensor reading is not sending the correct voltage to the ECU then it will not be telling the boost control valve to cycle sufficiently to control boost at 0.6. I.E if it senses less or more vacuum than required to cycle the valve.A defective MAP sensor would show a code 31, 33 or 34 but if the hose is off, partially plugged or the filter clogged it may not. I am not sure if the TPS return voltage is also tied into the ECU for BSCV output but i would check the MAP sensor and hose first.

I've just taken the car for a test run with ElanScan and (unfortunately) managed to provoke fuel cut a couple of times due to boost creep. I can see that it's regularly boosting to around 0.75 bar, which ought to be impossible anyway, but going at it harder makes the boost inexorably climb above 0.9 with the inevitable result.

The thing that's puzzling me at the moment is that the wastegate duty cycle never seems to rise above 25% - in fact, at around 580 seconds into the scan the duty cycle is 0% even though the boost is over 0.9 bar! Can anyone shed any light on this, or am I misunderstanding the data in some way?

I'm attaching the zipped ElanScan file so you can see what I mean.....

Over the weekend I plan to disengage the actuator to see what happens then.

Ben

[vi11ain]

While in the shower this morning (all my best thinking is done there!) I realised that 0% wastegate actuator duty indicates that the ECU is trying to REDUCE the boost as it's allowing the full bost pressure to bear on the actuator, thus opening the wastegate. Activity above 0% on the BCFV indicates that the ECU is trying to increase boost above 0.4 bar (or thereabouts).

I must've been staring at ElanScan figures for too long last night - my bad!

Unfortunately that still leaves me with a problem. Right now I'm thinking it's either:

a) Insufficient wastegate port size

b) Faulty (not moving enough) wastegate actuator

c) Faulty (leaking) BCFV

The next (cheapest/easiest) test is to try wiring open the wastgate, I think. Sadly that'll have to wait until Friday or Monday at the earliest now.

Ben

[GeoffSmith]

While in the shower this morning (all my best thinking is done there!) I realised that 0% wastegate actuator duty indicates that the ECU is trying to REDUCE the boost as it's allowing the full bost pressure to bear on the actuator, thus opening the wastegate. Activity above 0% on the BCFV indicates that the ECU is trying to increase boost above 0.4 bar (or thereabouts).

Correct. Duty cycle will increase to ~25% as boost ramps up then ramps down to 0% to maintain boost. As yours is not being brought under control, I would do as you suggest and try with it hard open to rule out wastegate size issue now you've de-restricted the flow with a PCE and Piper.

[vi11ain]

I've just tried running the car with the wastgate actuator disconnected, so the wastegate is free to open fully. The result is that the maximum boost I was able to achieve was about 0.6 bar, although I'm going to do another run with ElanScan to be sure. This seems to prove that the port is adequate when the wastegate is wide open....

The question now is how far does the actuator actually move to open the wastegate when it's functioning correctly? I'm going to carry out a pressure test on the actuator to see how far my one moves, but I have no basis for comparison. Unfortunately the Service Guide only mentions that it should initially have moved 2mm when it has attained a pressure of between 9.8 & 10.2 psi.

Ben

[esprit888]

Glad thats the case as i am about to go the same route (Piper PCE And 2.25 pipes with ported wastegate) plus an everest chip. Checking the full motion of the capsule & rod will prove its capable of full travel, what will not be known is if it moves at all under boost conditions so the MAP and BCSV are still suspect. I had a similar problem on an Esprit so i tie wrapped a 12vdc microswitch to the rod and used a small hose clamp to attach a piece of metal to the compressor housing approx 1/4 of an inch from the switch. Connected the switch to a power source and a bright red 12volt bulb and taped it to the rear window Went for a spin and at boost the light would come on & off (as i let my foot off) crude but for $8 it proved to me that the wastegate was moving fully, and when it should. Thought about making up a permanent indicator light for 'wastegate open' in the dash but hit a deer before i got that far.

I've just tried running the car with the wastgate actuator disconnected, so the wastegate is free to open fully. The result is that the maximum boost I was able to achieve was about 0.6 bar, although I'm going to do another run with ElanScan to be sure. This seems to prove that the port is adequate when the wastegate is wide open....

The question now is how far does the actuator actually move to open the wastegate when it's functioning correctly? I'm going to carry out a pressure test on the actuator to see how far my one moves, but I have no basis for comparison. Unfortunately the Service Guide only mentions that it should initially have moved 2mm when it has attained a pressure of between 9.8 & 10.2 psi.

Ben

[vi11ain]

So how much movement of the actuator rod did you have on your Esprit to trigger your switch then?

I don't think that the MAP sensor is suspect because it is providing the readings that show the overboost on Elanscan and thus cause the ECU to cut the fuel giving message 31.

[GeoffSmith]

I don't think that the MAP sensor is suspect because it is providing the readings that show the overboost on Elanscan and thus cause the ECU to cut the fuel giving message 31.

Makes sense to me.

[vi11ain]

I don't think that the MAP sensor is suspect because it is providing the readings that show the overboost on Elanscan and thus cause the ECU to cut the fuel giving message 31.

Makes sense to me.

Thanks for the vote of confidence there Geoff - it's always useful to have a quick sanity check from people on here!

I've just completed a few pressure tests on the actuator and BCFV. Both of them hold pressure nicely and the actuator seems to be moving correctly. I don't have a DTI (yet - one is on order!) to check for the 2mm movement at 10psi, but it looks to be similar to that amount. Further increases in pressure cause the actuator to move quite a bit further.

It appears that the maximum actuator movement is something slightly over 12mm - this is limited by its attachment to the wastegate arm, which also means that the wastegate is unable to open absolutely wide open as it can when disconnected from the actuator.

So, this leads me to believe that I need to further port my wastegate . One thing that I didn't do was chamfer or radius the edges of the wastegate port - I left them "sharp" at 20mm after grinding with the Bosch conical stone. Did anyone else here put a chamfer or round off the edges?

Mercifully I think I can remove the wastegate housing/downpipe without removing the rest of the turbo, so I should be able to carry out this extra porting without too much trouble - once it's all cooled down enough to touch, that is!

Ben

[esprit888]

If it makes sense to Geoff then it makes sense to me I have yet to see the results of an Elanscan, but i will do a 'before and after' run once the PCE and pipes get here. Geoff, just so i understand, what is the MAP voltage to the ECU under normal boost (0.4 to 0.6 ish) and what is the voltage at overboost which commits the ECU to fuel cut off? My reasoning for a possible bad map sensor was based on the EMK 2 list (below) and i wondered if a bad MAP ground could result in this scenario? Is the MAP sensor the only method that the ECU uses to determine overboost? (i am guessing it is)

I don't think that the MAP sensor is suspect because it is providing the readings that show the overboost on Elanscan and thus cause the ECU to cut the fuel giving message 31.

Makes sense to me.

[vi11ain]

If you take a look at my Elanscan results, you'll see that the MAP sensor pressure is varying along normal lines - I'm only getting high pressure readings when you'd expect to be seeing them. Looking further through the engine management guide (EMK3 - D) I see that the voltage output from the MAP sensor should vary between 0 and 5 volts - 2.5v being normal atmospheric pressure (1bar absolute) and 5v being 1 bar boost (2 bar absolute).

The basic statement in the page you've attached is that the ECU takes a one off reading from the MAP at ignition on, then subtracts that figure from the ongoing MAP readings to determine boost. I don't know how the ECU senses overboost if the MAP sensor fails - I'd imagine that in that event, it'd go into "limp home" mode, so it wouldn't really be a possibility.

I've managed to remove the wastegate housing/downpipe now (it wasn't as easy as I originally thought!) - the wastegate actuator and arm are quite clearly visible now, so I'll take some photos of it at rest, under pressure etc. before I grind and reassemble again - they might be useful to someone!

[Paul.P]

Ben, be careful if your gonna grind the whole any larger, the flap only covers 25mm & you should already have 20mm, so that only leaves 2.5mm either side, not a lot to get a seal on, especially if your gonna make it smaller ?

[vi11ain]

Ben, be careful if your gonna grind the whole any larger, the flap only covers 25mm & you should already have 20mm, so that only leaves 2.5mm either side, not a lot to get a seal on, especially if your gonna make it smaller ?

No worries - that potential problem is in the forefront of my mind at the moment - I'm not doing this lightly as I really didn't want to go back in there!

Fortunately I don't think I'm going to have to go very far:

Increasing from 16mm to 20mm represents an increase in area of ~56%, increasing to 21mm is an increase in area of ~72%, 22mm is ~89%, and so on....

I'm thinking of increasing the port diameter to somewhere around 21.5 - 22mm maximum - I'm going to have a go at widening the inner areas as well to aid gas flow....

Ben

[vi11ain]

I managed to remove the wastegate downpipe (so much easier now it's been recently removed!!) and I could just about get to the wastegate port for some more grinding, although it wasn't very easy!!

Anyway, to avoid too many of the grisly details, I think I managed to remove about 1 - 1.5mm material on the diameter of the port. I also smoothed out the interior of the port so the gas could flow around the corner more easily - there were a couple of sharp corners before that took a bit of removing. I finished putting all the heat shields back in place this afternoon and took it for a test run with Elanscan - result.....


No more boost creep or overboost!!!! Hurrah!

I can't really express how happy I was to see the boost gauge stop rise to around 0.6 bar and just stay there, back under the control of the ECU!! It's really great to be able to drive using full throttle again after two years of enforced 1/2 throttle - not to mention the past couple of months of fiddling around with the turbo and grinding ports etc! I've finally got my old car back - but with a better exhaust and PCE.

Elanscan confirms that the max boost was 0.66 bar - nearly perfect!

As a further bonus I managed to remove the pesky O2 sensor blanking plug from the manifold after the test run, so I will be able to fit a sensor and hopefully increase my mpg once the appropriate Everest chip has been fitted

So, tomorrow I'm going to celebrate by fitting the new Hispec brakes - I'm hoping for no "complications" this time!!

Ben

PS It was great to meet a few of you at Donington today - I managed to resist spending loads of money at the show and it was good to see so many Lotii in one spot!

[esprit888]

Great news Ben ! But scary to me as i am going down this same road. So the opening size is critical to within 1mm to prevent creep & OB ?

I managed to remove the wastegate downpipe (so much easier now it's been recently removed!!) and I could just about get to the wastegate port for some more grinding, although it wasn't very easy!!

Anyway, to avoid too many of the grisly details, I think I managed to remove about 1 - 1.5mm material on the diameter of the port. I also smoothed out the interior of the port so the gas could flow around the corner more easily - there were a couple of sharp corners before that took a bit of removing. I finished putting all the heat shields back in place this afternoon and took it for a test run with Elanscan - result.....


No more boost creep or overboost!!!! Hurrah!

I can't really express how happy I was to see the boost gauge stop rise to around 0.6 bar and just stay there, back under the control of the ECU!! It's really great to be able to drive using full throttle again after two years of enforced 1/2 throttle - not to mention the past couple of months of fiddling around with the turbo and grinding ports etc! I've finally got my old car back - but with a better exhaust and PCE.

Elanscan confirms that the max boost was 0.66 bar - nearly perfect!

As a further bonus I managed to remove the pesky O2 sensor blanking plug from the manifold after the test run, so I will be able to fit a sensor and hopefully increase my mpg once the appropriate Everest chip has been fitted

So, tomorrow I'm going to celebrate by fitting the new Hispec brakes - I'm hoping for no "complications" this time!!

Ben

PS It was great to meet a few of you at Donington today - I managed to resist spending loads of money at the show and it was good to see so many Lotii in one spot!

[vi11ain]

Great news Ben ! But scary to me as i am going down this same road. So the opening size is critical to within 1mm to prevent creep & OB ?

Well, bear in mind that in my case the initial port to 20mm eliminated about 90% of the boost creep - beforehand I could reach fuel cut in 3rd gear without any problem and I was basically restricted to 1/2 throttle all the time - overtaking was dangerous and flooring the accelerator for any length of time on the motorway resulted in fuel cut. After boring out to 20mm I could only induce fuel cut in top gear at high (road-illegal) speeds and after several seconds of fully open throttle.

It seems that small changes in this area make a big difference when you're close to the "tipping point" as boost creep is a feedback kind of problem - more boost = more power = more exhaust gasses = more boost etc. Once you've just got on top of it, eveything's fine again.

From my experience I'd recommend boring out to 20mm with the Bosch conical grindstone, then open it up maybe another 1mm or so (should be easy on the workbench) and smooth the inner corners off to provide a nice easy route for the gasses. If you make the port slightly larger than necessary I'm sure it won't matter as the wastegate is controlled on a closed circuit by the ECU. Your only real risk is going larger than the flap, but you have to go quite a way for that - I sprayed some paint over the closed flap to show how far I could go. Here are some pics of the various stages.

tn_1 before porting with grinding stone.jpg

tn_2 Nearly through.JPG

tn_3 almost.JPG

tn_4 Through.JPG

tn_5 paint sprayed on closed wastegate.JPG

tn_6 Wastegate overlap.JPG

tn_7 second grind .JPG

tn_8 second grind progress 2.JPG

tn_9 second grind progress.JPG

tn_10 Finished.JPG

[Blue_Bullet]

Maybe a silly question but.....

Did you do all that with the wastgate still in the car?

or are you just undoing a nut and bolt with that 14.4 Volt Drill?

[HJ2]

That's exactly what i wanted to ask as well
so:

[vi11ain]

Maybe a silly question but.....

Did you do all that with the wastgate still in the car?

or are you just undoing a nut and bolt with that 14.4 Volt Drill?

I did the main bit of grinding (from 16mm to 20mm) with the turbo out of the car - I'd highly recommend doing all your wastegate grinding on the bench!

However, I did grind the last little bit that got everything completely working properly again while it was attached to the car - I just couldn't face undoing everything and removing the turbo all over again, especially as I could just about access the port with the 90 degree angled drill and a flexible drive shaft attached to my trusty Black and Decker Wizard!

Ben

[Blue_Bullet]

OH Ok Thanks for all your help.

My cars int he garage at the moment, I told them to port to 21mm.

I have recently had the cybox peco fitted and I am running on A Everest chip with a drop in K+N Filter.

Hopefully i will have the car back next week some time I will post and let everyone know the results!