EFI: PowerTuner Maps vs. A/F ratio


25 replies to this topic
  • WRF-Rowdy

Posted June 25, 2015 - 08:01 AM

#1

setup:

I installed the euro header and muffler to play with the O2 sensor on the header.

Turns out the OD of the header is ~1 mm less than the OD of racing header_with_bomb

so I had to use the muffler that came with it.

Said muffler looks like a carbon copy of the aluminum racing muffler, except for being made of steel and chrome. 

The internal baffles and baffle penetrating pipes look identical.

 

I hooked up a digital votage meter to the O2 sensor and

ducktaped it onto the bars so I can read the digits while riding.

 

assumptions:

- header diameter and pea shooter won't affect air fuel ratio of the charge,

especially around low power demands, hence the measurements should be

not too far off the OEM racing exhaust w/o the pea shouter

 

findings so far:

- my current idle CO value of -4 still has the bike idle rich

- everytime you close the throttle you'll run rich (probably as zero throttle idel mixture cuts in)

- a fuel mapping with all fields set to -7 except for low throttle/low rpm (-6) still runs rich everywhere except for

STEADY 1/8 throttle and lower rpms.  edit: WRONG: checkout subsequent posting #21 http://www.thumperta...2#entry12497115

Exactly the short-shift mode yu'd use when you are afraid that some cop might bust you because of your exhaust:

very little throttle changes + shifting early, then you'll actually run lean.

- as soon as you increase throttle it'll jump to rich

- as soon as rpms climb into mid regions (w/ throttle unchanged at 1/8th) it'll turn rich

- every kind of short full throttle manoever in last gear will run rich

- only when you go WFO + step on the rear brake to keep rpms medium you have a chance of seeing lean combustion together with knocking.

 

So the ECU must have quite some acceleration enrichment (messing up any measurement).

At this scary "all  minus-sevens" map you still are mostly running rich.

But with -6 in the lower left corner you'd be cruising at 50 mph running lean (only at 1/8th throttle, mind you)

 

Conclusions:

- further experiments necessary

- maybe the header + muffler backpressue make the bike run rich? (sounds unlikely to me)

 

Except for applications where one might WFO through deep, gooey mud (keeping rpms medium)

one can only PowerTune the constant speed cruising mode to not run be rich

All maps that are floating around are rich, everywhere.

I think YAMAHA did that to protexct the engine.

 

plans:

- I'll search for the richest CO value that'll gives a non rich idle mixture, to not mess up readings so often.

- I definitely need to do some freeway testing.  I need a climb where I can go WFO w/o accelerating.

 

question:

Anyone got an information source describing how 1/8, 3/8 and 6/8 throttle are to be interpreted?

are they measured ate the throttle grip? 

Or calculated from the throttle opening value of the CO adjustment toll's monitor mode?


Edited by WRF-Rowdy, June 26, 2015 - 06:11 AM.


  • grayracer513

Posted June 25, 2015 - 08:52 AM

#2


 

question:

Anyone got an information source describing how 1/8, 3/8 and 6/8 throttle are to be interpreted?

are they measured ate the throttle grip? 

Or calculated from the throttle opening value of the CO adjustment toll's monitor mode?

 

The displayed throttle position is based on values from the TPS.



  • WRF-Rowdy

Posted June 25, 2015 - 09:08 AM

#3

The displayed throttle position is based on values from the TPS.

Hmm.  How does TPS position differ from what the CO-adjustment tool ouputs as throttle position?

How would I "calibrate" my throttle grip (with a white maker) to indicate

the throttle positions corresponding to PowerTuner's three throttle amounts?


Edited by WRF-Rowdy, June 25, 2015 - 09:12 AM.


  • grayracer513

Posted June 25, 2015 - 09:47 AM

#4

The TPS is a potentiometer, plain and simple.  The system uses high voltage to indicate closed or idle, and low voltage to indicate full throttle.  The intermediate voltage stages are determined by the CDI by reading the signal.  The throttle angle displayed by the technician's scan tool, as opposed to the Tuner, should be accurate enough to use as references, but bear in mind that the map "switches" as it crosses from one zone to the next, rather than "flowing" along a progressive curve.  The map is modified to a different algorithm within each of the 4 zones, idle, above 1/8, above 3/8, and above 6/8.  If you use the indicated boundary as a reference point, be sure you take any readings from well within the appropriate throttle zone so that you don't get crossed up by a disagreement between the scan tool and the CDI as to where the switch-over point is between one zone and another.



  • chu

Posted June 25, 2015 - 10:10 AM

#5

By "rich" do you mean richer than 14.7:1 or do you mean like pig rich where it's reducing power?



  • WRF-Rowdy

Posted June 25, 2015 - 10:32 AM

#6

By "rich" do you mean richer than 14.7:1 or do you mean like pig rich where it's reducing power?

Yepp, rich relative to perfect A/Fratio of 14.7.

 

The O2 sensor basically jumps to 0.1 V or below above 14.7 and to 0.7 V or higher below 14.7.

The Voltage to A/F ratio curve is extremely non linear. The simple O2 sensor are only good

to control a periodically varating fuel amout to maximize the "number of crossings of 0.45 V" per second.

 

The sweet spots of max. power and max efficiency at quite at bit off the 14.7

http://www.mummbroth...ages/Stoich.gif


Edited by WRF-Rowdy, June 25, 2015 - 10:53 AM.


  • WRF-Rowdy

Posted June 25, 2015 - 10:52 AM

#7

The TPS is a potentiometer, plain and simple.  The system uses high voltage to indicate closed or idle, and low voltage to indicate full throttle.  The intermediate voltage stages are determined by the CDI by reading the signal.  The throttle angle displayed by the technician's scan tool, as opposed to the Tuner, should be accurate enough to use as references, but bear in mind that the map "switches" as it crosses from one zone to the next, rather than "flowing" along a progressive curve.  The map is modified to a different algorithm within each of the 4 zones, idle, above 1/8, above 3/8, and above 6/8.  If you use the indicated boundary as a reference point, be sure you take any readings from well within the appropriate throttle zone so that you don't get crossed up by a disagreement between the scan tool and the CDI as to where the switch-over point is between one zone and another.

 

Personally I don't think they switch.  That way you could reintroduce the old CDI's

surging if you'd programm a decent ignition advance jump using the PowerTuner.

 

The maps each have 9 points to control their shape. I'd say it works like this:

If the system needs an injection timing (fuel amount) it checks current TPS and rpm.

As those two values seldomly will be one of the nine points where the fuel (adjustment)

value is readily available, it has to interpolate (linearly maybe) between the four points

closet to current location on the map.

That interpolation would avoid jumps of fuel amount (less an issue) or ignition advance

(very noticeably) when parameters change only gradually.

 

Hence I'd assume the map to be interpreted as such (for an assumed TPS of say exactly 3/8th):

anything rpm below and up to the first rpm point (4000) uses the low value

for rpms between first and second (4000/6000) it interpolates between low and mid value

for rpms between second and thrid(6000/9000) it interpolates between mid and high value

rpms above the thrid rpm (9000) produce the high value

 

That way every possible rpm value would result in an easily calculated result,

w/o introducing noticable jumps.   And yamaha can be sure that their offered

adjustment values are never exceeded.

 

But this is all guesswork, the internal workings are regrettably not documented, AFAIK.



  • chu

Posted June 25, 2015 - 11:06 AM

#8

 

I didn't watch the video but there's a graph around three minutes in.  Looks like it interpolates to me.



  • grayracer513

Posted June 25, 2015 - 11:34 AM

#9

There is not necessarily an immediate change in timing or fuel delivery as the CDI shifts from one zone to another.  What does change is the values and/or "weight" used by the mapping algorithm to account for changes in RPM and load within any one zone.  As the "recipe" changes, changes in situation from that point on are handled differently until the next zone is reached.  This is a more simplistic approach than is required for, for example, a street driven, emissions compliant automobile, or even a modern street bike, which has a much more varied and complicated set of circumstances to operate under.  



  • WRF-Rowdy

Posted June 25, 2015 - 11:37 AM

#10

Cool vid :thumbsup: 

Definitely show the map to be linearly interpolated. See from 3:10 till 3:40



Visit the ThumperTalk Store for the lowest prices on motorcycle / ATV parts and accessories - Guaranteed
  • vlxjim

Posted June 25, 2015 - 04:02 PM

#11

Don't know were to start. But the ECU comes with its base map if you will. This map is a living map as it is adjusted by the CPU in the ECU using TPS, APS, Temp and RPM. At the same time overlaying the users map. The user map as well as the base all crossover between the values at each point. It is very straight forward as far as open loop FI goes. 

 

You can't just use a DVM with an O2 sensor to make out A/F ratios. You need a CPU that can add buffers and firmware to interpret the fast changing values. 

 

I don't get were you come to the conclusion that they are running so rich when they are actually running lean. The stock idle CO setting is so low (for EPA). I don't know your setup but I would not be running those -7's. Unless your having a problem leaky injector or something you should not be running rich. If your bike is backfiring on decel your running lean. If you header is glowing at night your to lean. 



  • Chaconne

Posted June 25, 2015 - 04:24 PM

#12

Don't know were to start. But the ECU comes with its base map if you will. This map is a living map as it is adjusted by the CPU in the ECU using TPS, APS, Temp and RPM. At the same time overlaying the users map. The user map as well as the base all crossover between the values at each point. It is very straight forward as far as open loop FI goes. 

 

You can't just use a DVM with an O2 sensor to make out A/F ratios. You need a CPU that can add buffers and firmware to interpret the fast changing values. 

 

I don't get were you come to the conclusion that they are running so rich when they are actually running lean. The stock idle CO setting is so low (for EPA). I don't know your setup but I would not be running those -7's. Unless your having a problem leaky injector or something you should not be running rich. If your bike is backfiring on decel your running lean. If you header is glowing at night your to lean. 

What do you mean a "cpu that can add buffers" ?



  • chu

Posted June 25, 2015 - 04:25 PM

#13

It should be good for setting the idle though.  I get a lot of pops when engine braking since I got a straight through muffler.  I guess when the throttle plate closes and it's under vacuum the exhaust gasses recirculate.  I tried adding fuel in that cell but it doesn't help all that much.  I figure I might as well save the fuel if it's gonna pop annyway.


Edited by chu, June 25, 2015 - 04:35 PM.


  • vlxjim

Posted June 25, 2015 - 05:54 PM

#14

What do you mean a "cpu that can add buffers" ?

A micro processor that can collect smooth an average all of the data interpret the data and provide a usable calibrated number.



  • vlxjim

Posted June 25, 2015 - 05:56 PM

#15

It should be good for setting the idle though.  I get a lot of pops when engine braking since I got a straight through muffler.  I guess when the throttle plate closes and it's under vacuum the exhaust gasses recirculate.  I tried adding fuel in that cell but it doesn't help all that much.  I figure I might as well save the fuel if it's gonna pop annyway.

Whats your CO set to?



  • chu

Posted June 25, 2015 - 07:34 PM

#16

I used to have it set to +8 before I knew that you could adjust CO with the motor running.  I now have it at -7 which seems to be lean best idle.  Might be a little better off idle throttle response if I set it to rich best idle but MPG would probably suffer.



  • WRF-Rowdy

Posted June 25, 2015 - 09:57 PM

#17

I used to have it set to +8 before I knew that you could adjust CO with the motor running.  I now have it at -7 which seems to be lean best idle.  Might be a little better off idle throttle response if I set it to rich best idle but MPG would probably suffer.

I, too, found that these +6 +8 values effectively kill e-startability in gear.

Today I'll go and find the CO value where the O2 tilts to lean and report here.

 

reg. off idle response: I don't think you'll see an effect. I'd say the CO idle mixture stops to have any effect at very tiny throttle values above fully closed.

In my test setup I could watch A/F ratio jump from rich (idle) mixture to my crazy -6 as soon as I moved the throttle the tiniest bit.

Ther is no loading up wiht rich mixture and need to clean out like on two smokes, the mixture change is instant, as soon as you leave fully closed.



  • WRF-Rowdy

Posted June 25, 2015 - 10:09 PM

#18

You can't just use a DVM with an O2 sensor to make out A/F ratios. You need a CPU that can add buffers and firmware to interpret the fast changing values. 

That's only partially true:

Of course I can check for "A/F leaner than 14.7" vs. "A/F richer than 14.7", no need for any CPU to do that.

 

But one can't read any other values w/o using a modern 6 wire A/F sensor and a controller.

The cheapo std. O2 sensor (as in WR450F euro header) only produces two usable statements

"above 0,45 V" (rich)  and "below 0.45 V" (lean)

Hence first generation catalytic converter ECUs kept jiggling the fuel amount between two very close values and

were counting how many times the O2 sensor's output voltage would jump from below 0.45V to above 0.45V per second.

If the output voltage wouldn't jump they changed both fuel amounts until jumping would set in.



  • WRF-Rowdy

Posted June 25, 2015 - 10:23 PM

#19

I don't get were you come to the conclusion that they are running so rich when they are actually running lean.

 

I know if it's running lean or rich, I'm measuring it.

 

 

 

The stock idle CO setting is so low (for EPA). I don't know your setup but I would not be running those -7's. Unless your having a problem leaky injector or something you should not be running rich.

 

Certainly not.  CO set to 0 is rich, +8 is so extremely rich that e-starting in gear is not possible.

CO -4 is truly and definitley still rich!

(with "rich" I mean compared to perfectA/F ratio of 14.7, mind you. other people might mean somethuing else by "rich")

 

 

 

If your bike is backfiring on decel your running lean. If you header is glowing at night your to lean.

Not true.  It needs to backfire a little.  It does back fire a little with CO set to -4. It is running rich while slightly backfiring.

 

If ones header is glowing one is running the weird WR exhaust cam timing, which open 22° crank degrees earlier than sensible, thereby blowing out less expanded read much hotter exhaust gas.

That can be proven by asking any non-EFI WR owner how going to YZ timing affected his header glow.

 

The O2 sensor only is reliable if one's exhaust gas path is air tight (until to where the sensor sits, at least)

Should someone have an airleak say at the header - cylinder head junction one would introduce air (O2) into the exhaust gas and the sensor's voltage would immediately drop to near zero, indicating "lean".

Hence "lean" readings might true or they could be the result of air leaks, broken wires or dead sensor.

But "rich" readings (voltage above 0.45V) are proof of a leak free exhaust, and therefore are always trustworthy.



  • chu

Posted June 26, 2015 - 03:57 AM

#20

reg. off idle response: I don't think you'll see an effect. I'd say the CO idle mixture stops to have any effect at very tiny throttle values above fully closed.

In my test setup I could watch A/F ratio jump from rich (idle) mixture to my crazy -6 as soon as I moved the throttle the tiniest bit.

Ther is no loading up wiht rich mixture and need to clean out like on two smokes, the mixture change is instant, as soon as you leave fully closed.

 

I had a carbed bike that had way better throttle response with a fat pilot but it would suck down the gas.  Maybe because it was a simple carb with no accelerator pump.






 
x

Join Our Community!

Even if you don't want to post, registered members get access to tools that make finding & following the good stuff easier.

If you enjoyed reading about "" here in the ThumperTalk archive, you'll LOVE our community. Come join ThumperTalk today!

The views and opinions expressed on this page are strictly those of the author, and have not been reviewed or approved by ThumperTalk.