OK, thanks, Eddy. I guess I need to find a manual on the lockup torque converter transmission to understand it a little better.

Charlie will chuckle at this bonehead move..... I've been getting some ragged feelings from the engine at times - particularly on decelleration - and finally remembered that when the S/C ate its rotor bearings and trashed the other engine, I was in the process of running AutoTune - which I had closed and saved with dramatically altered VE table values - due to aluminum powder hitting the o2 sensor, I suppose.

Since this engine has been installed, fuel mileage has been rather bad at around 15mpg. I was beginning to think that the head shim and lack of quench was causing poor fuel burn and that I needed to change back to a hotter spark plug.

Today, I suddenly remembered the VE table, grabbed the laptop and went for a ride to the machine shop with AutoTune running. Along the way, I saw controller adjustments that were hitting the upper and lower correction authority stops at up to +/- 40% while adjusting to AFR table values. <img src="/forums/images/graemlins/scared.gif" alt="" /> From what was a terrible VE table that looked like the Swiss Alps, I now nearly have a good-looking table again. A couple more trips should do it. <img src="/forums/images/graemlins/kewl.gif" alt="" />

Frank

Great, now I have to come drive it again to make sure it's still unbreakable.... <img src="/forums/images/graemlins/rolleyes.gif" alt="" />













<img src="/forums/images/graemlins/smile.gif" alt="" />

I'm curious about autotune. Does it have different AFR targets at different TPS angles/MAP readings? I'd assume your target is stoichiometric at cruise metrics, but you need 12-12.5:1 afr for power metrics and high boost levels, and could even go to 16-17:1 at low load cruise for better economy with some engines.

Everybody better go get their "drive the hot car" turns before I do...

Eddy, you'd do better reading about it on the MS site, 'cause dang if I can explain it. AutoTune doesn't set the AFR and Ignition tables - you have to enter your own target numbers for AFR and Ignition in their respective tables first, but it adjusts the VE (volumetric efficiency) table so the EGO control circuit(which reads the wideband o2 sensor) doesn't have to hunt so far and wide to hit the AFR targets.

A simple way of explaining it would be that when the VE parameters are "in the ball park", the o2 feedback can be more accurately adjusted through EGO correction control.

The table uses the same X, Y rpm and manifold pressure bins as the AFR and Ignition tables, but is expressed as the efficiency (I think) of the engine at the data points.

Does any of that make any sense?

Frank

Sort of.

As I understand what you are saying, the MS ecu gets an rpm signal, and from that and the VE table, determines the actual air amount pumped in, and based on the afr target table, injects the correct fuel volume to meet that afr target. At 100% VE, the air quantity would be the swept volume of the cylinder, so I would assume your VE table has values ranging from maybe .6 to about 1.2 (the latter for a forced induction engine)? Are you running a mass air sensor, or are you strictly speed/density with a 2bar map sensor? Using a VE table tells me speed/density...

I'm running pure speed density, although MS has the ability to run MAF or a blended input of MAP/MAF. Speed density is certainly the preferred choice and I didn't want to risk leaving the MAF sensor in the air filter housing for fear of a potential air flow restriction.

The standard MS build uses a single 2.5Bar MAP sensor. I've modified my board to include 2 MAP sensors - both 4Bar - one sensor is used for initial startup base settings.... the second sensor is used for on-the-fly calculations:

MapDaddy



That's about right as I understand it. I'll have to wait until the VE table tuning is finished before I know the upper/lower limits. VE isn't always what it would appear, though. As an example, in forced induction engines, the upper limit probably never gets above 130% because of exhaust gas in the cylinder filling.

Every engine will have a different VE table. The differences might be slight on similar engines, but I should have realized that when I advanced the cams, the VE tables would need to be modified.

Frank

Man, that is some deep stuff you guys are talking about. How funny that it is being done on 20 yeare old platforms. Don't stop. Thanks.

Bet the VE doesn't top 1.3...

What kind of resolution do you get with 4bar of MAP range? Why not just 2bar?, unless you are planning on going real bugnuts crazy with the boost. Am I missing something?

The reason for 4Bar was only because that's what was available in a component configuration that would fit the circuit board in a dual installation. I checked the readout for accuracy against a 2.5Bar sensor and found it was plenty fast, very stable and gave the same reading as the 2.5Bar.

Even if the reading wasn't quite as accurate - with dialed in VE tables the wideband readout should allow the EGO correction to adjust the fuel to hit the AFR table value.... as long as I set the parameters to give enough controller authority to the EGO circuit. The idea with VE table tuning is to start with about 80% authority, let it make table corrections centering within the range and gradually work the authority down to around 10% once everything is dialed in.

All of the sensor readings are relative anyway - and repeatability around a given vertex is more important than the actual number - because once you get on the dyno and adjust the AFR table numbers for max power to give the engine what it wants, it is what it is - no matter what number value is assigned. Also, the wideband o2 sensor and EGO circuit has the final say to bring the fuel trim back to the AFR target. Since the o2 reading is independant of any input data except the exhaust gas, it is used by the EGO circuit to modify the injector pulse width to bring the mixture back to the AFR target that was established on the dyno as ideal..... no matter what the VE table says.

I think there's a trade-off between a tight controller for accuracy and a loose controller for safety. It seems to me that if you dial it in tight and have a fueling or sensor issue, the controller may not be able to react to the needed extent to save the engine.

I don't know what the upper limit of the VE table will be, but I would guess it's not going to be more than 120%. It really depends a lot on the VE of the base engine. What would you say - 80%? If so, and that represents 143HP and I think I now have ~210HP designed in - then that represents a 50% increase in efficiency - or about 120%. The actual math may not be that linear, but I would be surprised to see a peak VE much over 100%.

There's a whole lot to learn with MS and I'm still a rookie. The more I learn, the more questions I have. Maybe one day it'll come easy, but it's not easy getting to easy.

Frank

I was taught that a NA engine VE is usually on the close order of 70% for a very efficient engine, one with high compression and helmholz tuning. I'd make your base non-boosted engine about a 60-65% VE. I'd make the stock 3.0 about 65-70% with the tuned intake.

My concern with the MAP was thus. You have a fixed voltage range for the sensor outputs. If you spread 4bar of readings over that 5 volts of range, you get half the resolution of a 2bar sensor over the same 5volts. As you say, the ego is an autocorrector, and this may not matter at all, unless you get a false map of MAP values as a result of the low sensor resolution.

I agree the sensor readings are all relative. The system reads the system voltage and does a correction to give repeatability as the battery voltage varies over time.