AFR versus exhaust smoke

[ollykiwi3]

this kind of follows on from my previous thread about fluctuating AFR.
I still have a fluctuation with AFR reading, but using lag factor of 30 has helped it.
Im thinking the injectors(bosch 40lb green 4 hole) are either too big or are cheap knock offs that aren't working correctly(bought new from High flow dynamics in case someone has had dealings with them). and yes I still have to check flow rates physically, but for now i need help with the following:

stable idle around 900rpm and AFR reading of 14.5 , however the exhaust pipe is pumping out a fair bit of black smoke, and smells of raw fuel.
the VE table is around 20 at this point, and inj time 1.3-1.4ms(bit low I know in non linear) but if I take more fuel out(15-16 VE) it starts to run rough and tries to die but still has black smoke.
how is possible to have raw fuel coming out, an AFR of 14.5(sensor output voltage seems correct) and if I remove fuel it starts to lean out and still have black smoke?

if im in the non linear region and try to pull fuel out, will it act like it normally would or would it be unstable and not do what you ask of it?
I took the advise of some of the other members and used the bigger injectors for future proof, but they have no issues with the idle tune and control. so not sure why I do?

the engine has only been running for a couple of hours and I haven't driven it yet as I want the tune to be better first.
the LH exhaust pipe does pop a bit also, like a misfire. just wondering if it might be a fouling plug. I have pulled them and they are all the same, very black and sooty. ties up with the exhaust. did a compression test and all are between 178-185, so don't think I have a valve issue. KPA is around 55-60 at idle.

have just updated to the latest firmware which helped with the AFR fluctuation from previous thread.
thanks

[DaveEFI]

My first reaction is your O2 sensor (or software setting) is wrong. 14.5:1 wouldn't normally produce black smoke or black plugs.

[billr]

Check the injector flow at your operating pressure, using your pump and FPR arrangement; just a simple flowing one into a bottle for a timed period will be better than nothing. Your VE values are so far off that the req_fuel has to be wrong, What fuel are you using with a stoic of 14.8?

[ollykiwi3]

Daveefi: I have checked,recalibrated, and changed sensors to confirm the settings(please read my previous thread for more info). all seems to be ok

Billr: I was just plying with things prior to posting that msq, and forgot to put back to normal. gas here is 95 octane unleaded. I changed back to 14.7, but nothing changed as far as required fuel is concerned. its how it was and what TS has dialled up using the calculator 9.75/4.75.

lets go back to basics- if we put aside the AFR readings, injector flows etc, and just look at the exhaust. black sooty smoke=rich right? so if I take fuel out the smoke should go away, but the car starts to RUN lean(not look lean according to AFR), but still has smoke.

Billr: what makes you say my VE table is out? that's where autotune put it, based on the AFR required and what its seeing. I guess if I have big injectors, then it would want less fuel, so VE table at idle would be low would it not?

I will take an injector out and have it tested. but regardless of what it actually is, its bloody strange to have an AFR of 14.5 and still have sooty smoke either way.
how rich would it need to be to geta rich misfire so it thinks it's lean? Just another thought.

cheers

[billr]

First, the WB (AFR) sensors are valuable tools, but not omniscient or infallible. If the engine is belching smelly black smoke and fouling the plugs black, the AFR is most certainly not 14.5 (at least on all cylinders), regardless of what the sensor is telling you. VEAL (what your are calling "autotune"?) uses the O2 sensor for one of its prime data inputs, so if the WB reading is off, from wiring errors, physical location errors, exhaust leaks, calibration values, etc. then you get a bogus VE table set. Simply the classic "garbage in, garbage out " computer syndrome. Again, the WB is not a deity to bow to, question its readings freely...

My interpretation of "VE" is that it stands for Volumetric Efficiency, how well the cylinder fill with the air/fuel mixture prior to the start of the compression stroke. If the engine was turned slowly and the intake valve timing was such that the intake valve is fully shut before the piston leaves the BDC before compression, then the VE would be 100 in all cells, regardless of MAP or baro. The actual density of the mixture in the cylinder may not be "100% of standard-day", of course, but the "multiply on" and MAT table corrections take care of those issues. VE is primarily a parameter based on the engine's ability to "breathe". Areal engine will not have primitive valve timing like that, of course. The valving and intake/exhaust systems are optimized to get as high a VE as possible. In fact, it is common to get a VE higher than 100% at the engine's torque peak, it is actually being "supercharged" a bit from inertia effects in both the intake and exhaust. Point is, if your max VE is really only 80, that means your engine is breathing very poorly, like an over-revving flat-head lawn mower engine. Likewise, it is rather unlikely it is only achieving 20% VE down in the lower rpm ranges. Those low values, for both the low and high end of VE, suggest that the req_fuel is set incorrect (too low, I think). Since you used the req_fuel calculator correctly, I'm questioning the injector flow value you used. Injector dead-time also affects the "effective" flow-rate, but that's more complicated to check. Just do a simple check of injector static flow first, see if there is any obvious problem there...

[EspeNS]

-155-968 injectors?
In spring this year I helped a mate measure deadtime and flow on 155-968 injectors. Lucky we had 8 injectors to the 6cyl BMW. Because one of the injectors measured about 20% more than the others, I think it was. And that was not knock offs. The other injectors measured exactly the same with nice spray pattern.
On 3.8 bar they also measured not 44.25 lb/hr / 465ccm as usual rated, but 47lb/hr / 495ccm, and deadtime of 0.4 at about 13v.
All injectors had previously been used without problems, until it started flooding the oil with fuel.

[ollykiwi3]

Billr.
The top end hasn't been touched yet.is just the old map from when it was running with the old engine. I'm just concentrating on the idle right now.
If I get that sorted and working right then 'veal 'can do the rest.
If my VE is this low at idle and is not req fuel issue where would I start looking for engine problems to solve poor flow?
Intake is clear heads and all other parts are new.
Maybe cam timing or lift issues?
I will do the flow test this weekend .
Regardless of that,why if i take more fuel out would it start to run Leaner and rough but still have black smoke?
Maybe an injector not closing properly or wrong flow like the other post suggests?
I will go back through software for the wide band and see if something is out there.it obviously had a problem still.
Thanks

[billr]

"why if i take more fuel out would it start to run Leaner and rough but still have black smoke? "

Honestly, I can't tell you. I could make up some theories, but would primarily be guessing. However, if you still have black smoke, then it sure seems to me that it isn't running "lean", at least on all cylinders. It may be going leaner when you "take fuel out", but not leaner than stoic. My focus on the injector flow and dead-time is only because that is where people seem to get in the most trouble trying to get a nice idle. Like you say, a leaking/sticking injector is also a possibility, and even a poor spray pattern. Some simple flow checks can give a clue as to gross problems like that, even if flow-rate and dead-time aren't measured super accurate.

[kaeman]

on the exhaust smoke issue, I made the mistake of doing a bunch of tuning on a good running engine by just calibrating the 02 sensor and then driving and datalogging and using megalog viewer to analyze my ve tables... got them all tuned and then started having a misfire..... decided to do a tune up... plugs cap rotor ....before the tuneup, my idle afr was 14.3 and my cruising was 14.8 to 15.2 right where I wanted it. power was at 12.8 when under hard acceleration... things were great....after tune up idle was 12 and cruising was 12.2 power was 10.... all because the ignition spark was a little weaker with the other plugs. now I get to retune the darn thing because the properly firing plugs are making all the extra fuel burn and read real rich. my old plugs were black and sooty....what I learned that hard way was that rich misfires read lean.....and if the plugs are worn out then every cylinder has a slight misfire and will read leaner than it really is..... could be you need to check ignition and timing.....both could create the leaner readings.

[ollykiwi3]

Kaeman
thanks for the reply.
I have new plugs, as I just rebuilt engine. running NGK6 series v groove plugs which are one step colder than originals, with .050" gap. also running MSD coil(coil only), so would have thought it should be enough. but will take on board what you have said, seeing as you have first hand experience. I have tried timing from -30deg and although it does increase rpm with more adv, its not a huge change to AFR. what about dwell time? maybe need more? should I reduce plug gap?

I need to find a way to stabilise voltage output or at least what TS sees, as its pretty crazy oscillations. rapid, not just wandering, even with car not running.
I plugged in the STIM and outputs are stable, so its not the MS unit playing up.

does anyone know of another way to buffer the AFR out put or the voltage TS sees? I have lag factor on 20. yes grounds have been checked, and ive tried jumper cables to other ground ponts so I don't think its an offset issue.

thanks

[kaeman]

I have been reading some info about plug gap..... its like a magic trick to get any good info. but in speedtalk and a couple other forums I found that most of the serious tuners set plug gaps at .030... and less if it was a boosted engine. they said the extra gap was only for smog engines from the 80's....because of the low compression and lean fuel mixtures.......
they said that large gaps promoted bad ignition of the fuel because of the extra time it took to jump the larger gap. just some helpful info I picked up ... I lowered my plug gap to .030 from .035 and the engine ran better.... it may help in your case also.

[ollykiwi3]

great. thanks again. will change gaps tomorrow.
I just put my old 19lb inj back in again as I know they used to work,and ran it up. absolutely no difference, in fact I think its worse...
its stupid rich.. no idea what actual AFR is right now, but I don't care..its so rich its not funny, and I left the VE table as it was, and I tried to pull fuel out again, and it just died.
I tried pulling a couple inj leads off while running, but no major difference. which seems a little strange.
I must be missing something here. its like all injectors are firing at once. I checked with stim again, and itall seems ok.
is there a way t o see if inj are alternating? and not firing all at once?

[DaveEFI]

An LED with series resistor - or a noid light - will show the injectors firing at low revs. At higher revs it will tend to look always on.

A dwell meter (was used for checking point ignition) will also show the mark space ratio. So could check all the injectors are receiving the same signal.

If an injector was 'leaking', the fuel rail wouldn't hold pressure with the pump stopped.

[steveh66]

You nose tells you what it happening! Its running rich!

After a run and its warm...pull the plugs and see which ones are firing which will black and the wet one that are not. A problem with the ignition will cause the extra O2 in the exhaust and since you are tuning based on the O2 sensor you will add more fuel to compensate.

If you had one injector that is running too much fuel you may have the same problem if it causing the that cylinder not to fire. If it is firing then it will cause the other to run lean based on the O2 feed back.

[ollykiwi3]

Plugs are all black and quite even.
I'm not tuning with the afr right now.its unreliable so I'm just tuning by ear and sight.
Still hold rail pressure when stopped. Was thinking more about it holding open more during running.

Are the led for inj on the stim controlling the inj or displaying what the ecu is doing?
They looked about right at 350rpm using the stim if that is what theecu is actually doing.

the pw for the 19lb has gone back up to. 2.5 ms which is where it used to be but it's running the same as the 40 lb. It used to run fine with the 19
.

[Jagboi]

Out of curiosity what wide band O2 sensor are you using? I have not had good luck with the Innovate meters, I had readings all over the place. Then I switched to an AEM unit and had no further problems ( and no more calibrations). Have you ensured that the unit you are using is calibrated for gasoline, and not some other fuel?

[DaveEFI]

Plugs are all black and quite even.
I'm not tuning with the afr right now.its unreliable so I'm just tuning by ear and sight.
Still hold rail pressure when stopped. Was thinking more about it holding open more during running.

Are the led for inj on the stim controlling the inj or displaying what the ecu is doing?
They looked about right at 350rpm using the stim if that is what theecu is actually doing.

the pw for the 19lb has gone back up to. 2.5 ms which is where it used to be but it's running the same as the 40 lb. It used to run fine with the 19
.

Using a LED across (not instead of) an actual injector with the engine running just helps to prove the wiring etc to that injector is ok - which a stim doesn't. It just checks the MS itself - but doesn't provide the same load as the injectors. You'd be looking for obvious differences from one injector to the next.

If the fuel rail is holding pressure with the pump not running it means there isn't an injector stuck open or weeping.

But I still think there's something odd with your O2 reader side.

[ollykiwi3]

AAARRGGGGGhhhhh............
holy crap. what is going on here.
I re checked all the injector wiring, found one on the wrong bank, but don't really think its an issue. changed it to where it should be.
batch fire, 1674, 2358. as suggested and as was running for past 12 months no problems.

re set plug gap to.030".

started car, and no different. req fuel at 21/5.28 per calculation, pw 2.5(remember now back on 19lb inj). couldn't rev engine at all and lots of backfire still with black smoke, but seems too lean also. confusing.
as a quick change I adjusted req fuel up to 35/8.75 and It ran a little better, minimal backfire.
was able to rev a little but still clouds of black smoke.
pw was around3.5(inj 1 &2) at idle with req fuel as above.
pulled WUE down to minimum to get rid of some fuel.

remember all I have done is rebuilt engine(new cam /higher comp also) and put 40lb inj in it. it was running fine before that.
it was running better than this on ms BETA version, but had the strange AFR(previous thread), then downloaded latest firmware 3.3 something, and AFR got better but im sure its getting worse for running it.

could it be an injector driver sticking, and if so how would I check it. im not sure where to find them ,and I looked in the build manual but cant find specific item.
I built it myself, and it has been running fine for a year. but I have no idea what the driver looks like.

I need to get this thing going again as it now appears the lh bank is starting to smoke(oil) like the bore is glazing over from idling for too long

Im not using any AFR feedback as its a waste of time right now.

it seems like injectors are firing too often or too long but not showing up as that on TS dash. what are the injectors triggered from?? could it be a signal issue triggering twice?

A carb is looking pretty good right now.

[DaveEFI]

I'm surprised it ran at all with such a big change in the req.fuel.

Have you tried 2 squirts per cycle?

You're using low impedance injectors with no series resistors?

[Jagboi]

Intuitively, your pulsewidth seems too large at idle, and the black smoke confirms it is rich. Try cutting back your req fuel and see what happens.