NefMoto

Hello, everyone. I have been searching this forum but found different opinions about my problem, and I couldn't take a final decision what to do best.

I am currently running a Z20LET engine on an Opel Speedster, the setup involving 630cc inj and GT3071 turbo, on standard ME1.5.5 Bosch electronics.

Engine is pulling a nice 425hp figure at 1.5bar boost, but I am having terrible problems when letting the engine idle after a full boost dyno run. Initially the engine seems to die, if I press a little bit the accelerator pedal while revs go down, I manage to make it idle, but the idle is lumpy. If I unplug the MAF connector, the idle goes back to normal. Stopping the engine, plugging the MAF back and starting the engine solves the problem, but this only lasts till next full boost run. I suspect the engine stopping is caused by the fact that I am running an SSQV BOV, but why is the idle still lumpy once I get the engine running at idle after a full boost run?

My suppose is that the turbo pressure is maxing out the 80mm Z20LEH MAF which I fitted ( of course linearising the new sensor by copying the map from original Z20LEH Astra VXR file, as original one was 70mm ) and this causes the MAF to go crazy after the huge airflow caused by boost. Unfortunately on last dyno I kept my eyes on AFR and boost gauges, so did not have the chance to see the actual MAF voltage, and the sensor failed. Tomorrow will replace the new sensor, but I am afraid that one single run at 1.5bar boost will break it again.

Fortunately, I have a 90mm MAF housing from EVOMS. Initially this came with a different sensor than the one I use in my 80mm MAF, so I have no clue about how linearisation map should look like and the info from the European providers of this intake is kept "inhouse".

Question is: what does fitting the 80mm sensor into the 90mm housing involve regarding linearisation? Will I have to multiply the values inside the map with the ratio between the squares of both sections, or just with the ratio of both diameters? Or this approach is far from being precise and I need testing and testing?

Can someone confirm that 2.7T has 90mm MAF housing? Because this way I can get a sensor and fit it, so I know the exact linearisation map for it. Maybe someone can tell me Bosch code of inner sensor from this MAF...

More than this, actual measurements I made show that inner diameter is 72mm for the 80mm MAF and 85mm for the 90mm MAF, so does the 2.7T S4 have same size regarding the MAF - 85mm?

Any other opinions about my lumpy idle after high boost runs?

Sorry for asking so many questions but I have reached a dead end and I need help from others more experienced than me.

Get rid of the BOV, it's just sticking open.
Fit a bypass valve.

I took the BOV apart today, cleaned it and greased it, it looks ok. It's not causing lumpy idle problems, I am almost sure.

I told you what your problem is.
Get rid of the BOV and fit a recirculating valve, it's not rocket science.

I don't even see the point of these things, besides making noise.
They don't open properly on lower-transients causing compressor surge, and cause the airflow to play pinball in your pipes, and if the spring is weak enough for them to open, then they just open in normal vacuum/idle, which is what is happening to you.

Unless the MAF is on the pressure side post-bov it is not going to work right.
And even then you are just masking the problem, which is sucking unfiltered air through the BOV into your engine.

I have a tial 50mm BOV on my car and it is closed at idle. I do get minimal amounts of compressor surge while the valve is opening, but as soon as it opens there is no surge. The surge is present for maybe half a second. I am quite happy with it and have noticed no adverse effects, although I have ME7.5 with wideband primary o2. I have no interest in recirculating since I am MAFless, and I wanted to relocate the DV/BOV to the cold side piping right before the TB.

Mine is also closed at idle. I got your point, still don't think this is the cause but will leave it there as I am going MAFLESS because I cannot get rid of the MAF issues, can't calibrate the sensor in the new housing and anyway I use this car for drag race only. Thanks for answering.

Which is 99% of the operating conditions on a street driven car. That's why recirculating valves are used. So you can use a soft spring, and have them open on idle with no adverse effects.

This man knows what he's talking about - might be worth trying the solution and have a car with a chance of running correctly, no?

Why oh why do you think its a good idea to go mafless on ME ECU's and think that it will be a good idea? You are removing the ECU's primary load input...

I know it's foolish to cancel ecu's primary load input because of many reasons exposed here on the forum, I read about it, it's not ok to fool the ecu, BUT:

1. I don't have a suitable MAF, in fact I have, but I am unable to linearise it and the method of increasing in percentage by ratio of new housing/old housing diameters squares does not work, so it is equal to not have a suitable MAF.
2. A solution will be to fit a larger MAF from VAG, but this means that I have to also find a connector for it, as Opel connector will not work, cut the original loom, etc
3. Last night I measured the car with no MAF, with the same file that pulled 425hp and 535Nm, new figures being 462hp and 590Nm. In fact I had to lower the values inside the file as torque is too much for my sachs clutch.
4. Car is only used for drag racing and weekend round-the-city trips, so it's not a daily use car.
5. I am tired of buying MAF's.

Do you have any maps (damos) for Z30LET ecu?
I can help you for the new linearization but for the beggining, you can use RS4 map.
Have you change the TEMIN (minimales TE),TEMINVA (minimales TE bei VA)maps?

Isn´t your idle too rich ? (see LTFT e LTFT)

when you make a pull on dyno and release the gas pedal the engine will decay revolutions with lambda system disable (no correction).

It happens very often with me when calibrating on 100% Ethanol with very big injectors... put a little more advance (torque reserve in idle) and play with TEMIN TEMINVA TUVB to see what happens

The problem is he has a post-maf BOV as I understand.
So of course it goes to crap after a WOT pull, since measured air is vented to atmosphere and it goes rich.

Also, with the BOV, I wonder if someone ever logged rlroh_w or raw MAF signal during low transients, when the valve does not open.
It is quite erratic ... huge zigzag.

I´m considering he veryfied BOV and it didn´t gone open after the pull.

As I said, when I beggin calibrating with big injectors and Ethanol there is a plenty of modifications to make idle well, and allways involve TUVB and TEMIN... sometimes because I use Ethanol there must be some "torque reserve" map modification

Have you logged your duty cycle on that also?

What pressure are you running 3 bar or 3.3 bar? 3.3 bar only gives you 660cc from what I've seen duty is very high on 60lb's.

Sean cause that's what other Vauxhall tuners do who don't know how to do 90mm MAF despite it being a simple process and just takes a bit of dyno time..

P.s... You can just buy a 90mm MAF housing (check out courtenays website for example) and install your existing 80mm element into it and then just map it in.

New conclusions: car runs like crap at idle no matter if MAF is connected/disconnected, after it heats up everything is fine, I assume this is because of shitty 630cc deka's installed. Even raising idle at 1300rpm to avoid low revs running does not seem to cure this behaviour.

With new MAF sensor installed yesterday the engine will die after a full boost run because of the reason you explained, the BOV will release air and mixture will go rich, if I rev the engine while the revs go down after a dyno run, the engine does not stop anymore. So problem solved - seems old sensor was damaged.

Now, let's get back to my main issue - last night tests maxed out the 80mm MAF from first run so it is obvious that this MAF cannot cope with 1.6bar of boost.

I have an EVOMS 90mm housing. I obviously can take 80mm sensor and move it inside new housing, but as I said, I cannot make the new linearisation map by simply multiplying old linearisation map with the ratio between new diam/old diam squares. It does not work, at least for me it doesn't.

I also cannot use an RS4 linearisation as I assume sensors are not the same. My sensor is F00C2G2063, this is bosch code.

I have dyno, if someone is kind enough to point me in the right direction to calibrate this new MAF housing, I appreciate and will follow the advice.

Thanks.

And yes, I have Z20LET damos exactly for my software version, so I can make all necessary mods to make it work.

I have a tough time believing you are maxing the 90mm MAF at your HP levels.... are you sure you aren't running into another limit? Have you logged MAF voltage?

an 85mm MAF should be good up to 700ish HP.

Also the reason it does not work on cold start is because injector tuning is much more than just KRKTE and TVUB!

Might be a long shot but deka 630s don't really have good QC in regards to the spray pattern. On mine, the spray pattern was not consistent between any of the injectors and was not even close to stock. I marked the injectors to be able to index them upon reinstall. After I installed them and verified that they were aimed correctly, my cold starts were much better and I could drive immediately after starting the car. Just a thought...


You will not be able to scale in a new MAF 100% unless your fueling is dead on.

@nyet: in fact it all comes to marketing purposes...because the Z20LEH ( Astra VXR ) MAF quoted to be 80mm MAF has 72mm section. The EVOMS 100mm housing sold by EDS has indeed a cone shaped intake of 100mm, but slowly goes to 90mm external diameter of its section, and the inner diameter is 85mm.
And yes, I am maxing it out, first run at 1.6bar boost retrieved "MAF high input" DTC, so apparently it goes till max voltage.

@prj: indeed you are right. But as far as I've read I made a bad choice with Deka's. So it's not going to be an easy job to make the car idle as it should.

 

@nyet: I am reffering to LEH 72mm inner diameter MAF that is maxed out, probably the 85mm one should do the job, as long as it is well calibrated.

Huh......almost 40hp and 55nm ??? From where is that difference ? Different boost, fueling or timing ? Are you use stock intake manifold and throttle body ? I am interested to see dyno graph if thats possible ?

I am using upgraded intake manifold but stock throttle body. Boost went 0.1bar higher on MAFLESS than with MAF installed, AFR went a little bit leaner, but still in an acceptable range- 11.2-11.5. There was a problem on this MAFLESS run, a certain drop in torque at boost peak rpm range, boost went at aprox 1.6bar for a little time, I suppose ecu made a little correction as I have seen on other similar cars that ME1.5.5 does not like more than 1.55bar boost and makes a little correction if original pierburg valve is governing the turbo pressure and this is not done via separate boost controller. My experience on this ECU is limited, still not found what could possibly cause this correction once the boost exceeds 1.55bar. As long as boost is under this value, which is also the stock MAP sensor max value, everything is fine. I thought the key to this problem is to install a 2bar boost MAP, which I did (including calibration via gradient and offset), but seems it does not solve the issue. I still have a lot to learn about this ecu, please do not laugh at my mistakes.

The dyno run plot I am talking about is called softv2, and is blue. The other plots are several attempts at lower boost and lower injection advance, but the car heated and the other plots are not so concludent because you can obviously see that performance is poor in comparison with first run.
(http://postimage.org/image/a1euykprf/)

The dynoplot here, seems I could not attach it as an image.

http://postimage.org/image/a1euykprf/

Why your torque peak is that late i.e after 4700 ? On what gear you made dyno runs and with or without eddy brake load if you have such ?

Btw about 85mm MAF scaling.......if you already have real 72mm MAF curve, try to up it whole by 39% and then put MAF offset 280kg/h instead of 200 like in stock file. You need to have around 3-3.5g/sec airflow on idle, if not try to +/- offset value a little bit ;)

Torque peak is late because of GT3071 and because intention is to protect clutch and gearbox. Dyno is made in 4th gear, on a F23 diesel gearbox, which means aprox 240km/h in 4th gear at 7000rpm.

Will try your method about new MAF, thanks.

The run was made in inertial mode, due to long gear ratio I did not brake the dyno, although I have this option.

Yes, but if its not BOSCH HFM5 :)
85mm internal diameter HFM5 will finish on around 500hp on crank if not die many times before that :)

With 3071 turbo on that engine, torque need to be at least 800rpm lower........when you have full spool ?

Why would it die? Air velocity is air velocity, regardless of diameter.

It will die because its BOSCH, thats simple.......its made that way by factory i.e to have limited life, sometimes its really "limited" :) Anyway regardless of diameter if HFM5 stays around 50-60cm from turbo it can last some time........

I have seen cars tuned by famous companies in Europe, even 600bhp rated ones, once this engine goes to GT30 or over, I think it is impossible to have max torque at less than 4000revs, usual power figures show max torque at 5000rpm and beyond.

Mine is building 1.5-1.6bar of boost at 4500-4800rpm.

600hp are not in GT3071R power range, they are suitable for GT3582R, and you cant compare spool between 3071 and 3582 so torque curve will be very different between these turbos. Anyway with 3071 you need to have full spool before 4000rpm, if not, something is not right on that engine.........

how far is this Maf curve from yours , i have known this one to work . Unsure what housing it was in tbh but if it helps you out then bonus
Id say altering ignition low down would help with the cutting out possibly ,

I wouldn't say Seimens Deka 60lb's are shitty, infact even tuners who don't properly map in TVUB still manage to get them running pretty sweet at low idles such as 750-900rpm and still pass emissions tests. Heck I've even had Seimens Deka 80lb's idling perfect with cams on a Z20LET.

In regards to not having a MAF curve or not being able to make one, if you have a dyno and a wide band you can make one. Disable fuel trims set the desired fuelling and measure the error and correct the MAF curve easy.

But what I would add is on the Z20LET's running 11's at 450bhp on a Z20LET with Seimens 60lb's will probably be at about 100% duty on Z20LET's hence I am going 80lb's after about 420.

Dunno. I avoid bosch and always use hitachi, regardless of MAF diameter, for the reasons you state :)

If using a known maf (hitachi or bosch) just scale MLHFM to theoretical using area (or dia) ratio to get in ballpark. Then use KRKTE, KFKHFM and KFLF to fine tune. Fine tuning individual points on MLHFM is a huge waste of time.

I would disagree you don't have to perfectly scale all 512 values but you can get a percentage error for a range of the curve and correct on that percentage.

Applying one equation over the whole thing from old size to new size hasn't ever come out anywhere near what I've seen in use on other vehicles which the larger MAF's actually fitted.

Keeping it monotonic and smooth is a huge pain.

Doing % corrections in KFKHFM is exactly what that map is for. Why not do it there? Plus you can correlate it with load/rpm, generally shows more consistent error patterns than simply vs MAF voltage.

But hey, it if works for you, it works for you. There's more than one way to skin a cat.

Regardless if it sticks open or not, it is irrelevant.
What is relevant is that he dumps measured air to atmosphere, so the fueling will always go very rich when you lift the throttle.
And as I said before, even funnier when the valve does not open at lower transients and causes the MAF value to zig zag. You can guess what happens with timing and fueling at that moment...

But prj, there are 1000hp + cars running BOVs! That means BOVs are good.

 ;D

They are good when they are new. For 20000-30000 km.
After that their spray pattern on low pulsewidths goes to shit and they start misfiring on idle.
I wish I took a video of the injector bench, but those injectors at low pulsewidth were not spraying, they were taking a piss.

BTDT a million times now.

No, BOVs are not good at all for Motronic cars even with mafless tunes they are trickly, the only way they can work is like prj said with blow thru maf if the BOV is before the maf i.e turbo ---> bov ---> maf ---> throttle

Will receive car back from bodywork shop these days as it will be painted and will restart adjusting it. Will also swap to separate boost controller as I suspect the pierburg valve of the car is far from being precise. Will also move the 72mm sensor into the 85mm housing and try to adjust it, will keep you updated. Thank you so far for the good piece of advice posted here.

Not sure if serious or trolling.

First you fit a BOV on a ME7 car.
Then you say that you can't tune a MAF, so you want to remove it.
And now you say you don't know how to tune boost, so you want to put a MBC on?

jfc, give it to someone who has an idea.

Boost is tuned, little 1.6bar spike then constant 1.5bar boost. It's not that I can't tune it, it's the fact that the turbo solenoid valve seems to not be so precise on higher boost values.

ME1.5.5 it's a hybrid ecu, narrowband, indeed operating principle may be close to Me7, but this does not mean that it has same behaviour and precision.

And yes, I don't know how to tune the new MAF, no one was born a genius, I mentioned my experience is limited on these ecus. All I said about it is that I tried the diameters square ratio method, which is described all over this forum, but no one except one user here suggested MAF offset correction also, which sounds pure logical now once it was explained.

I would give it to someone who has an idea if I lived in a civilised country and not at the end of the world. In my country there are 2 cars like this, one stock and the other is mine. Closest Opel tuning company is in Germany, 2000km away.

Parts take from 1 to 3 weeks to arrive, no matter what I order.

Having these great opportunities in this wonderful country that I live in, I am forced to find myself the hard way a solution to make this car work, no matter if it has MAF or not, if turbo goes via stock solenoid or via MBC, etc.

I thank you all for precious piece of advice, don't want to bother anymore with my actions on this car, considering both lumpy idle and MAF calibration subjects received a certain answer, topic can be closed. Will open a new one once the setup is ready, of course I will try to keep both MAF and N75 active, if that will be possible.

Regards,
Alex

It suppose to be for 90mm (outside diameter) housing ... but it is scaled only by 20% in the range 1-5V ... Dan, did You used it without any other scaling factors?
I made a comparison with graph and included real Bosch 92mm(82mm ID) AFM characteristics from Bosch site (http://rb-aa.bosch.com/boaasocs/;jsessionid=025023972E38073B649A9BB6E6E44B46.sundoro2?template=page_product.vm&ccat_id=72&prod_id=341&lang=en).

The ECU does not see more than 2550mbar. But you don't need an MBC, it can just be tuned to a fixed WGDC value.
Nothing to do with "being precise" or not. The only thing a MBC is good for, is to put it in parallel with the N75, to make sure the boost does not spike over a certain value.
It is basically ME7. Yes it has pretty much same behaviour, and I don't see why you are talking about "precision". I don't think you quite understand what this word means in this context.
Fit the stock MAF, then tune the injectors to be perfect.
After you are done, fit the new MAF, and tune it again to be perfect.
Like this you can tune completely unknown injectors and completely unknown MAF, for which you have no specifications at all.
So what country is that?

Also, if you already know how to log the ECU and how to flash it, then I am sure you can be helped online for a fee.

My objective is not to diss your efforts, but you just did the 3 biggest no-no's that apply to tuning these cars:
1. Fitting a BOV.
2. Removing a MAF.
3. Fitting a MBC.
Add to that the fact that you make bold claims about "precision" when not fully understanding what you are doing, and you don't come over too well, that's the thing.

If asking for help you need to be more specific, for example:
I changed the MAF, then changed map X and map Y. The car was doing Z and I logged values P and Q to confirm.
Also, first thing will be to fit a recirculating valve if you want to get this sorted, as well as fitting back the stock MAF and linearizing the injectors correctly. After that actual tuning can start.

I've got a number of larger than LEH MAF cuves used on these engines. I've found quite a considerable varience with it between different tunes I've seen.

Really depends what actual MAF you're using as the term 90mm MAF seems pretty loose to which MAF the tuner's been using judging by the huge difference in max air flow.

Here is one Astra G Coupe Z20LET with stock M1.5.5, stock head, stock intake manifold with stock throttle body, GT3076R turbo 0.63 exhaust A/R, 830cc DEKA injectors, MAFless tune, removed both lambda sensors, with fuel economy better than stock car, idle is smooth like stock but on 950rpm because of more good idle oil pressure with hot oil. ECU controled boost and as you can see full spool 1.45bar on 3700rpm and flat torque curve from 3900 to 5000rpm. For reference only, on that dyno fully stock Astra H OPC(240hp) is 150kw on wheels

(http://s1.postimage.org/xrlsldhor/Astra_G_Coupe_Z20_LET_GT3076_R.jpg) (http://postimage.org/image/xrlsldhor/)

This is nice for driving but not so nice for the well known weak F23-gearbox. But it is good to know, I never thought the Gt30 could full spool under 4k.

Ok, PRJ, will take it the way you told me to, following the steps you mentioned. More feedback at the end of the week. Country is Romania, btw.

I might go there in Spring.
If you don't have it dialed it in by then, I'll sort it ;)

Oh come on Romania is not that bad. If you need some Opel hardware or other stuff that you can't get in Romania just ping me and I can probably help you out from Hungary. I've pretty decent connections at the local dealers here.

Whats wrong with Romania i dont know ?........In your country you have one of the world fastest street Nissan R35 GT-Rs (8.9sec 1/4mile), 8.8sec Nissan Pulsar, several 9sec cars and many 10sec cars, performance tuning in Romania is not bad at all.

Btw I am from Bulgaria, but as you see on dynograph, its not the problem to tune Z20LET engine here ;)

F23 ain't weak it's a beast when you compare it to the M32 box  ;D

Kelesha, there is a romanian phrase people use:one tiny flower does not bring spring. we are still far behind in means of tuning.  even with the gtr and pulsar...those are extremes...but majority suffers from lack of performance parts that usually arrive from Us in min3 weeks, very few tuners know what they are doing in this country, etc.

I am a in happy situation: i have my own dyno, i have tools, i have this opel speedster...but others have to travel hundreds of km to get to a dyno, for example...

Anyway, good to know that help is available if this project does not come to an end.

News update:

- MAF issues are solved: seems the thing that caused the engine to stop at idle once the MAF was plugged was the fact that because of too little space in engine bay, the MAF was located very close to the turbo, so the turbulence caused at idle by the turbo spinning at low revs was messing up MAF readings; it may sound easy now, but it took me a while to figure it out
- I moved MAF 1m away from the turbo, everything is fine
- I calibrated injectors after the 80mm LEH MAF installed, when everything was fine, I went for calibrating the MAF for larger housing, and here everything went fine also
- I stopped the dyno adjustment session because valve seals started to leak again, and I got a decent power figure of 425hp at 1.5bar boost, 11.4-11.8 AFR
- total adjustments took 500km on dyno and 104runs, but I am happy with the results. Boost control must be improved, it's not 100% precise, but this I will take care of once I get back to adjustments after fixing mechanical problems.

Engine will be taken apart for replacing all gaskets and seals, I also ordered ACL race bearings ( main + rod ), so I guess next adjustments will be made after driving 1000km break-in distance.

Thanks for help.

Well done sir  ;D

Which MAF curve did you use for the 90mm in the end? Theres several I've come across and depending which one you choose will either cause the car to use the right or wrong load values in other tables.

What was the top value in MLHFM you used?

Obviously any can be made to work by scaling the injectors.

1600kgh top value in MLHFM.