possibly pursuing a wideband o2 controller

[zlyricist]

I've been pondering taking up a wideband o2 project. something that can be used for almost any application, something that is more affordable than anything on the market, including the JAW unit

My old man is a software and electrical engineer. He can help me set things up and get going for a cheap price. He is already making good money engineering aftermarket boards for paintball markers, as well as selling configurable electronics for model rocketry. I know he can make reliable electronics that work, and for cheap.

My job is to find out what all is wanted is needed to make it work, and what the DIY/standalone efi crowd wants the most -

Does it need to be compact with minimal wiring?
Do you need/prefer a numeric gauge w your wideband?
Is output compatibility the most important thing?
What features do you want incorporated with your datalogging?
Basically, what is the most important features to you in a wideband controller?

All comments and suggestions welcome

[Fred]

Hi,

Welcome aboard :-)

A cheap accurate DIY wideband would be an excellent project to have around for sure!

Provided the code is GPL, and preferably that there are some open schematics accessible (preferably KiCAD project), I'll be happy to give it exposure here. If a board design is available for the few people that desire to do so, all the better! You will find though, that if its sufficiently cheap that not too many will do their own boards anyway, and that you will sell plenty.

Heres what I would want in a wideband and why I chose the Innovate LC1 for my truck :

Accurate (this is most important, wideband control isn't as straight forward as many think)
Compact (innovate squash an aweful lot into their tiny case, it will probably be bigger than that, but the smaller the better)
Reliable
Durable
Configurable
Simple to build

Simple wiring is a given because you only need the connector to the bosch/nsu sensor and the following :

1 x power feed to the unit
3 x grounds (one dirty for controlling the heater with a reliable FET, one clean for the controller inside, one for the output signals)
2 x configurable analog outputs (one for the EMS and the other for a gauge of some sort)
serial config wire(s) (this can control the calibration process and feed responses and data back to the pc)
1 x digital output, maybe a std digital output could be developed that would result in the EMS getting a perfect reading from the controller rathing than sampling it via analog means.

I personally would like to watch it through a digital led voltmeter or similar 7 segment display that is bright and easy to read at night. Some don't watch at all, it's just a datalogging tool for them.

http://www.moates.net do a nice unit called the ometer for 70USD

However, a cheaper option is a small case and 2volt digital volt meter with the decimal shifted across and the output of the unit set to 1-2v. These are around 10USD.

In any case, I think the display device should be separate to the controller and use one of the two analog or one digital outputs.

By no means have I put any serious thought into this. There could be other ways to achieve a good system and I may have missed requirements out. Hopefully however that serves as a good starting point for discussion.

Admin.

[26chevute]

I think Admin pretty much covered it.

Should be Compact, simple and Cheap.
Accurate.
Optional display (colored led bar graph or digital dispay)
Configurable analogue outputs.
serial connection.
Etc.

[zlyricist]

thanks for the input. i still have a bit of homework to do, and will know for sure within a few weeks to a month how serious i am. i will update this thread randomly

[zlyricist]

what kind of output to a laptop would be preferrable for data logging?

i think usb is the most common these days. im not sure if serial or usb is cheaper, but serial definitely seems to be less common anymore. any input is appreciated.

[slacker.cam]

there's two ways to look at this.

you're right in saying that more laptops have usb (pretty much none have serial these days) but if you want to make this cheaper than JAW then you need to minimise your part count as much as possible.

If you go USB I imagine that you will use a uC with a uart onbaord and then have a Serial-USB chip to give you the USB output. This is a pretty bad way to go about it in my opinion as why not allow the user to use his/her external usb-serial adaptor and save some parts costs? That way the user is free to choose if they want to use serial or USB.

Another way to go about it is to write code to make the uC interface directly as a USB device. This is the best way to go about it in my opinion. There is obviously more code to write on both ends (uC and PC) but its a far cleaner and 'proper' solution. You could even have an unpopulated RS232 output for the oldschool people if they want it.

Im sure others opinions vary...

[Fred]

He has some solid points there :-)

You could go one step further and provide a socket for a max232 and a set of pads/pins for a uart to usb board to connect to. That would keep it even more flexible.

The advantage of USB to uart is that these serial adapters (no matter where you put them) are fairly unreliable beasts at the rs232 level. eliminating that physical layer and using the same serial protocol is a very good thing.

However, if you configure it to be something the user/builder chooses, all the better :-)

Good thinking Cam.

Admin.

[Fred]

Here is that article from the magazine showing that only AEM and Innovate are accurate enough :

http://www.emspowered.com/vb/showthread ... t=wideband

Admin.

[toalan]

Selecting the right microcontroller is very critical. 90% of the work is writing code for the micro. The right micro will not magically make everything perfect, but using the wrong micro will ensure that your project never gets off the ground. When I mean right and wrong, 50% has to do with the specs of the micro and 50% has to do with ease of development, price of the dev tools, user base, documentation, etc...

[zlyricist]

well ive gotten started on a few things. as far as sensor inputs to go along with the wideband controller, so far im thinking of the following
rpm
tps
iat
map
egt
pulse width

what other inputs would people want, if any? thanks guys!