wideband smec available? or stand alone

[86-89 Lancer]

I can't tell from the websites if any of our vendors offer a SMEC for wideband O2 /Lamda scale. I guess it's just not available. Since my car is nearing completion, lately I've trying learning as much as I can about narrowband Vs wideband and IMHO It's not worth investing anymoney in a narrowband calibration. Here's why I feel this way and correct me if I'm wrong with these figures.

narrowband accuaracy = 14:1 - 15:1 AFR on voltage scale
Wideband accuracy = 10:1 - 17:1 AFR on Lamda scale

At WOT under boost the ball park AFR needs to be about 12:1. if these are true then it's impossible for the narrowband to work.

but it could be helped with an seperate wideband added to monitor real AFR, However, if the wideband indicates something is off with the calibration you'd have to send it off for adjustment or invest in EPROM tools. (more money in a narrowband system?)

- With Stand alone just plug in the laptop and adjust way- right?
- Can the stand alone be mated to the SMEC connectors (60 & 10) or do you have to change sensors and harnesses because sensor values are different?
- can the stand alone work all the factory inputs and outputs - even charcol canister?

It seems like a ton of work to go stand alone but why bother with narrowband? Why won't someone program a SMEC for Lamda? I'm sort of mad at this situation too because I just spent months rewiring from the dash to headlights converting from 86 lodgic module to 89 SMEC

Jeff

[GLHNSLHT2]

Quote:

Originally Posted by 86-89 Lancer I can't tell from the websites if any of our vendors offer a SMEC for wideband O2 /Lamda scale. I guess it's just not available. Since my car is nearing completion, lately I've trying learning as much as I can about narrowband Vs wideband and IMHO It's not worth investing anymoney in a narrowband calibration. Here's why I feel this way and correct me if I'm wrong with these figures. narrowband accuaracy = 14:1 - 15:1 AFR on voltage scale Wideband accuracy = 10:1 - 17:1 AFR on Lamda scale At WOT under boost the ball park AFR needs to be about 12:1. if these are true then it's impossible for the narrowband to work.

That's why above 5-7psi a stock ECU and more than likely all the vendors calibrations ignore the O2 sensor completely. But don't think this is just a narrowband thing. Even Vdubs and other manufacturer's like GM who have widebands from the factory ignore the O2 above a certain amount. Why? Because things are happening too fast in the engine for the ECU to adjust fuel based on what the wideband is putting out. So they run the car on the safe side of rich as well.

So just because your ECU is setup for wideband or narrowband doesn't really matter.

Quote:

but it could be helped with an seperate wideband added to monitor real AFR,

Now you're onto something. What you want is a wideband that has a simulated narrowband output like the LC-1 from Innovate. You feed the narrowband to the ECU to keep it happy and monitor the true AFR off either a laptop or a gauge.

Quote:

However, if the wideband indicates something is off with the calibration you'd have to send it off for adjustment or invest in EPROM tools. (more money in a narrowband system?)

True. But since you have a smec you have an easy route to take. For about $130 you can have the smec converted to be flashable and get an interface cable to plug into your laptop and with Dcal (free) you can upload any calibration you want. This is my preferred choice because it's cheaper than the vendors and if I want to change something it only takes a couple minutes.

[86-89 Lancer]

that makes sense Now I need to get Chris / Paul to sell me a cal and eprom tools. That could take awhile huh?

[TurboBuggy]

socket your own computer, its easy as hell. Download D-cal and Chem2 and start playing with stock calibrations to get the hang of things. Buy an EEPROM and a willem burner for cheap and start flashing chips.

Then if you want on the fly changes the easiest thing to do is to get an emulator like the ostrich from moates.net

Basically this stuff has been developed enough to where you can change any part of your stock code and even rewrite how it works if you wish

check out moparchem.com

[90TurboVan]

The SMEC is so easy to flash now, there's no reason to go to a standalone. Don't go the EPROM/Programmer route. D-Cal can flash your SMEC with the simple modifications detailed in a word doc on the Yahoo group and the serial interface also described on the group. You'll never go back to burning/swapping chips if you start flashing.

The SMEC changes fuel based on oxygen sensor input by modifying a table of ~6 values (definilty less than 10 values) which correspond to the amount of fuel to add or take away at different MAP ranges in order to get the O2 sensor to switch rich/lean (A/F 14.7). I.E. one "cell" of the table has a +/- value that is added to injector PW for -14.7 psi to -12 psi, another for -12 to -9, and so on.

This works fine for low boost or cruising MAP values, but the resolution isn't there for control in higher power ranges. Even if you were to put a wideband input to the ECU, you would need to do some serious modification to the source code to implement "self-tuning". Since it would have to modify its own fuel tables in real time to get the needed resolution, the processor probalby couldn't run the car at the same time. Megasquirt has that feature, but it's run off of a laptop, not on the ECU itself. What you can do is put a wideband input to the ECU 0-5volt 7-20AFR, change the setpoints in the cal for the rich/lean switch, and datalog that value. You get like stock O2 feedback along with an accurate A/F measure in datalogs, as well as the ability to change the A/F that the ECU tries to get to with the adaptive fuel. I would imagine those cars with widebands from the factory do adjust fuel at all MAP/RPM values, otherwise why not just go with a narrowband?

[86-89 Lancer]

Thanks for the help guys I may socket my own SMEC or buy one I'll decied which way to go soon but not yet, I need to learn more. I talked to Cindy and she said their program is proprietary and different from Dcal and the others - which is also something new to me, I didn't know there's different programs. Does that mean there's different tools and chips for each? Also, flashing means it's rewriteable - does burning mean it's a one time deal or can it be reburned? How many programms are out there and what are the highlights?

Delaware state inspection involves a tailpipe sniff test, check engine light check, check charcoal canister, and gas cap. if you pass those you can do whatever else you want, therefore I need to keep the canister even If I went stand alone.

The whole concept of: at 6,000 rpm things are happing so fast that the computer can't keep up with after the fact exhaust and by the time it compensates it's a day late, which could do more harm than good - that hit me as a wake up and makes darn good sense. thanks for setting me straight We need to run on look up tables and adjust them based on datalogging.

I wonder how many tables take speed sensor into account for wind resistance? A dnyo run has no wind resistance - but the drasgstrip does.

There goes my brain with all kinds of random thoughts now.

Jeff

[90TurboVan]

There are D-Cal and ChEM to modify cals. D-Cal will flash, as well as the earlier version of ChEM. I wrote my own program to datalog/flash/edit my SBEC, also should work with SMEC's. Flashable means you can plug your ECU into the comm port on a laptop and make changes without pulling the computer and swapping chips. Burning means you burn the cal on a chip with a programmer, swap the chips in the computer. If you want to change something, you have to pull the chip out erase it and burn a new cal.

Some cals use the speed sensor and a table of Boost target vs Speed, mostly for traction at lower speeds. The speed sensor is also used for some fan control tables, idle control, and cruise.

It's not so much response time of the ECU, its just very calculation intensive to change the fuel tables. The computer could tell how much fuel it wants to add or remove, but I don't think it would be able to apply those changes to the fuel tables without missing some cycles at higher RPM's. It is possible that it would work, but its very hard to make that big of a modification to the code.

[Kevin]

Quote:

Originally Posted by 90TurboVan The SMEC is so easy to flash now, there's no reason to go to a standalone. Megasquirt has that feature, but it's run off of a laptop, not on the ECU itself. What you can do is put a wideband input to the ECU 0-5volt 7-20AFR, change the setpoints in the cal for the rich/lean switch, and datalog that value. You get like stock O2 feedback along with an accurate A/F measure in datalogs, as well as the ability to change the A/F that the ECU tries to get to with the adaptive fuel. I would imagine those cars with widebands from the factory do adjust fuel at all MAP/RPM values, otherwise why not just go with a narrowband?

Megasquirt can be logged to a computer, and you need to use a computer to do the tuning/use the auto tune/ adjust tables, BUT megasquirt does not run off a laptop. It has internal memory and a processor, both of which are much better than the stock pieces in the factory SBEC/SMEC, that store the fuel and timing tables (for megasquirt II.)

The reason why newer VWs and the newest Lexus cars use a factory wideband is simply economy and emisions. They are both systems that are designed to have a cruising afr in the 16-17 range for economy. Conventional O2 sensors could not do the job. As far as I know the ecu still goes into an open loop mode at WOT.