A Narrow Band o2 sensor question

[Mouse!]

Hello. This seems like a stupid question, but I had no luck searching for the answer.

My LC1 wideband has been giving me trouble, and I have switched back to a narrow band o2 sensor. What do I need to change in software to make this work?

I have a v3 MSII running ms2extra_beta20080210, megatune 2.25p1

Under ego control it is set to narrow band with a target voltage of 0.5069. I calibrated the AFR table for "Narrow Band".

My logs show a good looking o2 voltage, ranging from close to 0v up to close to 1v. Gego never goes below 100. When the o2 sensor is above .5v Gego will rise, enriching it further(!). In steady state with o2 below .5v Gego stays at 100.

I don't have a stim here to play with, so testing is a bit interesting. I am hoping I just missed a setting somewhere.

Thanks for reading.
-mouse

[pigga]

Hi. I drove around with NB-O2 for a while as well.
I had calibrated a NB-AFR Table as you did and choose "single wideband" instead of narrowband, then it worked!?
Thomas

[Mouse!]

I played with it a bit more today and managed to track it down. Someone should check me, but I think it's an ini bug, and it looks like it's been there all along. Am I really the first person to try a narrow band sensor?

Anyway, I replaced this line in megasquirt-ii.ini

#if NARROW_BAND_EGO
egoTarget = scalar, U08, 598, "v", -0.004883, -204.800, 0.00, 2.00, 4 ; * ( 1 byte)

with this one from the b&g ini

egoTarget = scalar, U08, 598, "v", 0.01000, 0.00000, 0.00, 2.00, 4 ; * ( 1 byte)

and it seems to be working how I expected it to.

Edit: fixed a stupid typo...

[Mouse!]

Huh, I never thought to try telling it it was a wideband...

[offtrack]

isn't the narrow band useless for tuning?

it is really a digital device,
where the wide band shows magnetude and direction, a vector.

the narrow bad says, not this way, the other way , you guess how much.

useless , except in a fully tuned engine.

that is my take.

hope it helps someone.

now you help me,? oh, please.
can I mount my LC1 WB, in a vertical down header pipe?

20 inches long thin wall steel CRS tubing, mount it 10" from header flange. I know too hot is bad.

my oem oxygen sensor, is right at the collector just 10" from ports.
some say too close.

Mounting the wide band is the only hard part.
Placement.
I know to keep it hot at all times or it dies. and water collection issue.


great product, great site , Bless ALL !

[natesully]

Set your megatune config to narrowband to get a decent gauge, then select "single narrow band" in the EGO config dialog. IIRC, that should do it, the AFR map will disable and MSII will try to make your car stoich during closed-loop.

[Mouse!]

I have a forum question: Since I think I have found a bug, but I already have a thread discussing it should I start a new thread to report it or assume that a coder will stumble across it here?

Set your megatune config to narrowband to get a decent gauge, then select "single narrow band" in the EGO config dialog. IIRC, that should do it, the AFR map will disable and MSII will try to make your car stoich during closed-loop.

That's what I did. The bug in the megasquirt-ii.ini file was trashing the egotarget sent to the MS, and it the because of that the correction would not work correctly.

isn't the narrow band useless for tuning?

...

useless , except in a fully tuned engine.

Ummm, no.

I got a pretty good tune before I got my wideband. Compared to looking at plugs and such, a datalog showing a narrowband sensor is a godsend. Don't get me wrong, a wideband is way better than that, and cheap, these days, to boot.

That said, this car is a driver. I would not call it fully tuned (is any car ever fully tuned?) but it should be fine with a narrow band sensor. I just want to drive it.

can I mount mine in a vertical down head pipe?

I know too hot is bad.

I know to keep it hot at all times or it dies.

I'm no expert, but I think it's like this.
The wideband sensors have a heater in them to keep them hot enough, but no way to cool themselves off.
IIRC the LC1 sensor should not be mounted in something that gets hotter than ~500F.
I think the heater is generally big enough to measure cold gases. So while too hot is bad, (I've had that problem) don't worry much about too cold.
Running the engine with the sensor cold (not powered) can foul the sensor.
Downstream of an exhaust leak is bad.

I would think the thinwall tubing should cool down pretty well, but there are a lot of factors. Your sensor should come with some guidelines, and I would trust them more than my memory...

[ChevelleFan]

isn't the narrow band useless for tuning?

...

useless , except in a fully tuned engine.

Ummm, no.

I got a pretty good tune before I got my wideband. Compared to looking at plugs and such, a datalog showing a narrowband sensor is a godsend. Don't get me wrong, a wideband is way better than that, and cheap, these days, to boot.

That said, this car is a driver. I would not call it fully tuned (is any car ever fully tuned?) but it should be fine with a narrow band sensor. I just want to drive it.

Let me be the 2nd to say that narrowbands ARE USEFUL. True, a wideband is better, but a narrowband can be used to at least get your car drivable and running pretty well.

There's a thread on msefi where we were talking about NB sensors and using MegaLogViewer's VE Analysis. I just recently (this past week) used a NB on my high-compression 406 to get the car pretty drivable by running my logs through MLV.

-Dave

[offtrack]

few if any, pro tuners use NB O2 sensors! { not that ,that matters, )
"the only way to find the free power and efficiency in your engne is with a WB sensor. " quoting J.Hartman. there.

the NB has tunnel vision and will waste all your time trying to find
out where your engine is running. (at all speeds and loads)

14-15 AFR is way too narrow.
but...

edit: i see why pros use it ,its faster, and time is money to them.
for us, NB can do a good job. , and its not the cost its:
mounting problems, not too hot, not too cold, not to collect moisture.
on some cars ,this is hard to do ( with out sledge hammering underbody)
The $200 aint nuttn. ( and i will need one on 2 more cars)
{ welding a bung , now,just in case}

I am almost done idle tuning.... so....

the only confusion for me now is sorting:
Mstw3k
mslvv
vexme
and now , MLV.? mentioned above.

are they all needed or are they all competing.

very confusing.
if im missing some NB tools, please let me know.

I have a new idea, do it the NB way , and learn lots more ,faster.
Lots of the fun is the journey!



cheers to all and bless the makers of MS for such a great product.
A barrel of monkeys is no competetion.

[MegaScott]

few if any, pro tuners use NB O2 sensors!
the only way to find the free power and efficiency in your engne is with a WB sensor.

the NB has tunnel vision and will waste all your time trying to find
out where your engine is running. (at all speeds and loads)

14-15 AFR is way too narrow.

Try it , you will like it.

imho

A wideband sensor alone is not going to tell you when your making the best power...that's what the dyno is for...the wide band is there so the ECU can repeat the same AFR time after time under different conditions.

The narrowband sensor will get you in the ballpark to getting your engine running right, and sure for anything outside of it's measurement capability it's nearly useless, but as the manual states, you can estimate what VE you need for nearly all conditions once you find 14.7-1 on as many load points as possible. From this known starting point you can very nearly estimate the VE you need for any given AFR. Sure the pro's don't do it this way..they don't need to.

Not too long ago "the Pro's" didn't use anything but seat of the pants, plug checks, and dyno measurements to figure out how the engine was running, notice I said "was running" Lot's of so called "Pro's" still do it this way, with no sensors at all.

[offtrack]

I just finished Jeff Hartman's new book and liked his comment about
why save $200 on a Wide band, you will loose that in just moments on the dyno. He has a point huh? But after reading many threads with guys bragging "only spent $135" i can see why. grin.

My G16B:
I'm tuning mine now with LC1 and when happy I'm going to try my hand at using a G-tech (home made , chip board and a fluke 88) to set my advance on a NA engine. I'm tuning for economy. Don't need power have crotch rockets for that.

Just practicing for my real project , going from scratch on a old 389 GTO.
My step son has one and we will squirt it soon.

for economy again, because gas is $7+ a gallon where car , resides.

do you know of anyone with a write up on tuning spark advance using a G-tech accelerometer or the like? I can log my data using my setup.
0-1.5 g's . any ideas or is it a waste of time.
no power adders, just stock engines.

thanks for any clues.

Old carb guy. (specialty SU carbs and messed up quadrajets)

[offtrack]

if finally discovered how to use MSTweak 3000 ,blush. kind cool,and
will try it asap.

it says that it is grossly limited ,under load, or mostly.

says just to guess at loaded side. and provides a guess ,example.


I printed the manual and the chapter 11 is missing,"how to get a good data log".

Sounds like a fun project step.

thanks for pointing the way.

cheers

[ChevelleFan]

MSTweak is old, unsupported software. Plus you gotta muck up your datalog settings to make it work. Try MegaLogViewer instead. Much easier to use.

-Dave

[offtrack]

thank you so very much for that nfo.

i will try that now .

bless all , and good night.

[MegaScott]

I
My G16B:
I'm tuning mine now with LC1 and when happy I'm going to try my hand at using a G-tech (home made , chip board and a fluke 88) to set my advance on a NA engine. I'm tuning for economy. Don't need power have crotch rockets for that.

Just practicing for my real project , going from scratch on a old 389 GTO.
My step son has one and we will squirt it soon.

for economy again, because gas is $7+ a gallon where car , resides.

do you know of anyone with a write up on tuning spark advance using a G-tech accelerometer or the like? I can log my data using my setup.
0-1.5 g's . any ideas or is it a waste of time.
no power adders, just stock engines.

thanks for any clues.

Old carb guy. (specialty SU carbs and messed up quadrajets)

Ok, never heard of doing it this way...i think on a 389 GTO your more likely to bump up against traction limits, this will make tuning spark with a Gmeter very difficult.

Currently most people use a knock sensor to adjust spark limits, on the good knock systems you can get a reading of knock before audible indications or damage.
The new systems of Ion sensing are the best way to go, and some DIY solutions are available, nothing on MegaSquirt YET.

[Mouse!]

To tune spark for economy shouldn't you just go for the lowest pulse width, just like tuning the fuel (and for that matter the cruise speed)?

[ChevelleFan]

To tune spark for economy shouldn't you just go for the lowest pulse width, just like tuning the fuel (and for that matter the cruise speed)?

There is no 'pulse width' to spark. You either have spark, or you don't. The tuning of spark is deciding at what point to fire the spark -- somewhere in the range of 10-40* BTDC, depending on RPM & MAP values.

You should probably start a new post, as this one has gotten away from narrow-band O2 sensors.

-Dave

[Mouse!]

I already started a new post...

What I meant was tune the spark advance at cruise for the lowest fuel pulse width. With the clutch (or torque converter) locked up the pulse width is telling you how much fuel was used to move the car a certain distance. Tuning for economy is a game of lowering the fuel pulse width while avoiding driveabilty problems.

So, get out on the highway (or whatever other steady state you want to tune) retard the spark a bit and watch the fuel pulse width for a few seconds. If it goes down do it again. If it goes up try advancing it.

(Tuning spark for power is usually not steady state unless you are on a dyno, so that's where measuring acceleration comes in. I think it is easiest to look at the engine speed on the datalog. If you made it from 2500 rpm to 3000 rpm under the same driving conditions faster than before then you must be making more power.)

If you have a knock sensor, by all means, pay attention to it. But remember that your goal is to make the spark happen at the right time, which is not necessarily right on the edge of detonation.

[offtrack]

We pulled the GTO dizzy timings easy with a dizzy tester.
it has 303 HP stock at flywheel. so thats another deal. far later.

but back to gutless rice burner: ( im economy tuner , out grew power head 40years ago)

the challenge is this.( dyno it is not , that Gtech, or equiv , i made my own using a 3 axis accelerometer chip and my analog logger device i have)
cost: $3 for chip. I have a fluke meter that can log analog too.

but hard to do?
best timing at idle ?
the best timing at WOT max rpm , no way to do that. ( ? cops)

but in between these points might work well with Gtech.
i could take the 1k to 4000 rpm range and tune that.
in what ever gear is the most safe. 2rd gear maybe?

same flat road, same start point. far from people.
and log max g force from 1000 -4000 with my helper.
this is called the PPP peak power point.

raise the whole advance line from 1k-4k with MT.
2 degress more, and run again.
find the point where Peak G, rises and then starts to fall. BINGO.
and call it good.


the peak rises and falls on most engines. I am told.

on a real dyno they just load each RPM point (delta 500) and log torque.
piece of cake. ( $20,000 tool too)

I guess i will find out soon.

first I will use MLV (just bougth it ) and do many runs with it first with generic conserv. timing base. then the Gtech guage .

if it doesnt work ,at least it will be loads of fun trying.

the MLV is more fun than a barrel full of monkey's.
too.

Try it !

[max78]

Hi,

Can I use a different (user defined) calibration table?
I have some problems with the NB-AFR table.

0,5V -> AFR 10,2
0,29V -> AFR 14,5

Should not 0,5V be 14,7AFR (NB)?

With WB the voltage shown is different.
1,78V -> 0,38V -> AFR 12,7
1V -> 0,5V -> AFR 10,3
0,5V -> 0,73V -> AFR 5,5

#if NARROW_BAND_EGO
egoTarget = scalar, U08, 598, "v", -0.004883, -204.800, 0.00, 2.00, 4 ; * ( 1 byte)

with this one from the b&g ini

egoTarget = scalar, U08, 598, "v", 0.01000, 0.00000, 0.00, 2.00, 4 ; * ( 1 byte)

i think this is a bug, works when I correct it.