Craig Moates

Or putting the WB back in the DIY I guess.

Is it time for this? Some people think it is. A wideband controller with no-frills.

There'll be a balance between rip-off and reinventing the wheel, no getting around it, so I'd like to have key contributors involved from the get-go. This should be a joint project. If the originators of the original DIY-WB dont want this to happen, then I won't pursue it.

I don't want another TechEdge scenario. Rather, I'd like to see a DIY-WB2004 that is BY the DIY community, FOR the DIY community.

The following is proposed:

Development and deployment of a wideband controller box designed for the Bosch LSU4 sensor. A simplified design which could be a kit or an assembled unit. Just the box, which will have the following I/O:
1) Power & Ground (ign) {2.1mm Plug}
2) Two-wire Serial Data (115.2k RS232 for firmware updates and datalogging) {RJ11/45 or DB9}
3) ECM-bound 0-1v narrowband simulated signal {Screw Terminal}
4) 0-5v Linearized voltage output {Screw Terminal}
5) Raw (buffered) Vout from WBO2 sensor {Screw Terminal}
6) Market-compatible connector for WBO2 sensor {DIN-based}

So that's not too extreme. This gives 1 power connection, 1 serial connection, 3 analog outputs, and 1 sensor connection.

Feature creep is a disease. Let's leave it at home for now.

I'd like to see the unit AVR-based, with a Mega8535 or Mega162 at the heart. If you really want, we could put some sort of header inside for 'future' I/O functionality.

So there it is. I'm throwing down the gauntlet.

Let's hear from the DIY-WB originators and those that would like to contribute at a technical level. If someone needs to be brought into this discussion, please round them up. If I'm missing something, or this is already overkill, please let it be heard.

Cobra289

Hi Craig,
You got my blessing, it is 100% what we need at the TGO.
This project should have a GO.
You did the correct thing, asking for permission and support.
TIA
Regards,
Cobra289

TRAXION

I, for one, think it is a great idea. However, I believe the I will be in the minority (for those that are known in the community). Fact is - if it is DIY then you will be biatched at for selling an assembled unit (for profit). Thus, where does that leave you? Nowhere. However, it would most definitely be an asset to the community if you could produce and sell a cheap controller box that is both reliable and accurate. Many many many people cannot build their own box ... or don't have the time to do it ... or don't have the patience to do it ... or, most importantly, don't have the knowledge to troubleshoot it when they have problems with their newly built box that doesn't work. IMHO, the more widebands that get out there ... the better the PROM community will be. If you build it - they will come. However, the question remains if you will be blacklisted and hated by the DIY community. No matter what - you would need to seriously undercut both TechEdge and Innovate since both are available relatively cheap.

Tim

RBob

iTrader: (1)

I can't see any blacklisting coming from me. Craig is in the open about what he wants to accomplish and offer. As long as everyone that contributes understands the direction Craig is going, how can there be any biatching.

RBob.

nape

iTrader: (1)

You laid it out well and I agree that we should have the consent of the DIY-WB innovaters.

I would want one that is a complete unit, soldering and I don't get along well.

I am all for it, Craig. I read the other thread and I really hope this project goes through.

RBob

iTrader: (1)

Craig, one design criteria would be to have the grounds between the heater control section and the UEGO control section isolated. As the heater will be PWM controlled and the currents high it can wreak havoc with data logging the output signal.

RBob.

Cobra289

You are my man RBob that is progress.
We have two contributors.
Craig mentions “This should be a joint project” So Bill, Ken, Grumpy, Mark, others you are invited.

Many people want to have one of this baby’s in particular if they get the typical answer to his question “But you only can achieve that with a WBO2”
There are several problems with the DIY WB one of them is that not everybody is capable to understand the codes of the resistors or soldering but when they read that you need to calibrate and stuff like that make folks scare and they drop the idea.

Craig you mention “I'd like to see the unit AVR-based” I don’t know what it is but sounds like a “Chip” stuff. I like it new type of design because looking to the original DIY WB is a little out fashion. Don’t get me wrong we are progressing, right?

Regards,
Cobra289

AlexJH

AVR is the Atmel brand of microcontroller. They're getting popular now that Atmel has made them available to hobbyist types.

www.avrfreaks.net is a good place to poke around to get started.

Craig, I'd be interested in doing some coding. I've got 50 ATTiny11's on the way in a few weeks. $0.25 USD for a 6MHz microcontroller is just too good of a deal to pass up.

SBNova

I replied to the other thread, and just saw this one. I would like a WB that has a display, or one as an option as I have carbed cars too, and this kind of investment needs to work on all my toys.The display could be optional, maybe a plug in. I would need an input for RPM to correlate my readings, datalogging too, but the laptop could do that with some software (for non-EFI apps). In kit form- I would consider it, in assembled form I would love it.

I am not a contributorof this board, just a newbie learning all I can. I do believe that Craig has given plenty to the DIY community already with his innovations and low cost/high quality products. With his track record I dont see how he could be blacklisted when he openly asks for opinions. Keeping the price low would ensure his positive status. Charging a reasonable amount for assembly is fine with me. I dont expect him to give his time away for free. Im sure he will have alot of time invested in just development anyway. Go Craig !!

vernw

iTrader: (2)

D I T T O ! ! !

About the only contribution I could make right now is as a tester, which I would be more than happy to do using $8D and a 7730.



- Vern

Cobra289

A cheap digital voltmeter will do the job, just dial it at 5 volt and you get your display and this will be the part of the game (remember it will be some DIY).
There are more options that you can find.
Regards,
Cobra289

JP86SS

iTrader: (1)

Excellent idea, this will no doubt give greater range to all the DIY PROMmers out there. Craig and RBobs concepts for the unit already sound good to me, even if the cost is comparable to the others.
IMHO I don't know if trying to do the "low ball" approach to the whole thing is really needed. It may just cause re-design to implement new "basic" functions that WILL be considered as necessity as the program progresses (and surely there will be something!) Using the controller and doing firmware would be great progression from earlier design and allow for limited feature creep now but expandable as needed. Adding ports now will same design later.
I'm behind ya, All I can offer is to be a guinea pig for anything you need, but I'm here.

Ken73

This is what we call "feature creep" .. i.e. "Oh yea, and I want one with this.. and this.. and that.." We're trying to keep it simple.

Simple = Cheap.

You want more features, it gets more expensive and will take longer to develop, AND be a bigger pain to support. DIY support is usually a pain in the butt, as everyone does it on their free time. So the simpler we can keep it, the less of a burden it is on it's creators. An RPM input will complicate it beyond the original intention.

I mentioned previously (in another post I think) that the DIY AF meter Bill and I came up with can be used as a display for this, and at $40 for a complete kit (including PCB, enclosure, all parts) it's not a bad display at all. We've been tossing the idea around of making a commercial version that's a 2" round gauge, also.

Considering you most likely already have a laptop, that can be used as your display as the unit will output RS232 (serial) comms to a PC. (At least, that's the current plan with Craig, I believe.)

If you want the features you mentioned above NOW in an assembled unit, look at the Zeitronix stuff. I might be going that route myself, but I'll still build one of Craig's units once it's finished.

Note to Craig: Don't use a 2.1mm jack/plug set for power, please! Those things tend to come loose pretty easy. Something a little more sturdy that won't fall out as easy would be a good idea.

MTPFI-MAF

craig this would be a great thing. I can solder but not at it's greatest on small IC and such ,depending on price of the Parts kit I would give it a shot,and Put it togeather my self I just won't spend a few hundred dollars on something that I have possiabilty to SCREW UP a few hundred dollars is a lot to me at this moment (2 child on the way and all )a WB02 With no frills and a good price would be great.

Thank you for starting this thread Craig

Synapsis

So what kind of timeline would we be looking at for this?

JPrevost

Good luck! Supporting the Bosch sensor would be a good idea for lowering overall costs. Something that has dictated the direction of the original DIY WB. Hot and cold, prices are were unstable and sensors far and few between. Getting the boards made was the easy part but finding a sensor was a battle. I remember having to get mine off of ebay for $160.
Features; narrow band output, serial or USB connection for datalogging and a build in LCD display of the current AFR would seem resonable. Is there anything else this think would need to do? This is out in left field but having a pigtail adaptor that bypass's the narrow band sensor wire and feeds in the DIY WB would make for a great temp install. I just felt like throughing that out because I know some people hate cutting into the original wiring.

MTPFI-MAF

from a couple other web sites I found out the Bosch LSU4 cross references to a VW#021906262B which www.parts.com has for $30.00. and there are a couple other sites I found them going from $30-$50 that had them in stock. as long as the cross reference info is correct.

So maybe the assembled form I think that the buyer should have to buy the sensor seperate. That way it shows the actuall cost of the WB02 Controller it's self and the cost won't seem so High for the unit, or maybe this is just me Having a brain fart.

Just my .02

EDITI: just realized you said ECM-bound 0-1v narrowband simulated signal

dosent this mean that we can Put the LSU4 in the NB Bung that is already on our headers. Cool

JPrevost

No, that's the good about the Bosch sensor. The NTK is a better sensor but the Bosch is cheaper. From talking to a few people it seems like the Bosch sensor will be harder to control. Something to the effect of a calibration resistor being built into the unit where it can't be measured. Don't quote me on that but that's what I remember.

MTPFI-MAF

my info i see say that there are many auto manufactures that use the LSU4 but The differences are in the connector and the calibration values for that auto manufactures requirements, Hence the reason for different part # for the same sensor.

Ok So now I see you point JPrevost If this info is correct there will need to be a Standard Sensor coming form One source.One Specific Part #.

http://www.plxdevices.com/onlinestore_domestic.htm sells a Bosch LSU4.2 sensor with part #0 258 007 057/058 for $60.00

JP84Z430HP

Realistically, I think all the features should be left out for now. With the ability to put the WB into the ALDL datastream, all I would want is the WB Volts (lin, and raw), and the Sim NB volts.

Although, I do understand the need for the people that run C3's that aren't Romless, for some of the features....

funstick

im not one of the originators of the DIY WB but ive wathced things come and go and this and that arguement. If somebody wanted to black list craig for designing his own WB etc then i think its time to reasses the state of the upper class of the diy community. I think everybody should be given fair credit for what they do but if craig wants to offer a comparable product of his own scratch design i dont see where it would possiably infringe on anybody.


Onto design ideas.
In them OEM form the bosch lsu4 is standad calibrated.

every ECM pcm be it volkswagen,cadilac,sabb etc they all have the same calibration. At the bosch sensor factory they calibrate every sensor to the same stanrd. so thee is an agreed upon output value for a given heater contorl strategy. if you simply copy the factory heater control strategy or ill try to contact bosch since the US head quaters is minutes from my house with proper data sheets etc it should be little to no trouble to build this unit.

the sensors are pre calibrated. so make the box standardidzed calibrated,

Craig Moates

I was talking with my wife Bonnie, and she suggests that a display might be appropriate, even though it represents feature creep. It could be integrated more easily at the AVR level, prior to D/A conversion. So we can think about that as an add. It'll make things a bit more complex (no pun intended), but like I say, better at that level. Probably three LED segment displays. Could turn it off with a jumper, whatever.

7Point4

Not for sure how i could help, but willing to try.

Some electronics, Basic STAMP and BasicMicro programming experience.

For my 2 cents worth, the more basic the better. Display (simple LED would be cheapest), 0-1v output, 0-5v output, and serial output. and actually the serial would be nice, but not absolutely needed.

MrBill

The way I see it is simple: if there's something the people want and it don't exist...then make it. If the DIY-WB has gone beyond the capabilities of the typical DIY-er, then that leaves them without a good DIY solution. I was not personally involved with the DIY-WB project, but I don't see anything wrong with the proposal.

Craig, you know my capabilities, and that you can call me anytime...though time is stretched pretty thin I'll do what I can. I can probably best assist in the component-level design work, and PCB layout/fab (though I won't use Eagle for it! ). Definately contact me before you have any new boards made.

MTPFI-MAF

Two-wire serial Data My vote would be {RJ11/45} connectors and wire, Crimpers Ect. readly available at local rat shack.

For Power & ground I would say go with a 2 pin Molex or the ones radio shack carries {plastic locking non reversiable}or a 2 pin Weather pack so people can go to a junk yard and get the connectors to make their power cable.

stuckatcuse

Craig- count me in. I pretty experienced with the motorola HC08 automotive line of u controllers, but don't know much about the AVR. I have been looking for technical document on the LSU4 sensor so I could start doing the same thing. There has to be a "suggested design" out there somewhere for the oxygen pump device. If you have any techinical papers, send them my way and I'll get to work.

-Sean

MTPFI-MAF

I e-mailed Bosch asking How I would get a technical data sheet on the LSU4 and this is what I got back

Bosch Research & development (248) 553-9000

MTPFI-MAF

I contacted Bosch R&D and spok with Jim Iwonkvitsch {not sure on spelling of last name, i will find out so I can give credit where credit is due, well anyways jim tells me that the LSU4.2 sensor wether it is on a Prosche or a volkswagon the sensors end result is the same. The connectors are different for different car manufactures but on final assembley they use a trim resistor right before the connector to make sure that all sensors are the same.

He is going to E-mail any info and technical info to me later today.

He also recomended Chief Enterprizes for a source for the LSU4.2 sensors. Chief Enterprizes Ph# 1-800-831-7294

I will post info that Jim Icosich sends me

Just my .02

Ken73

If everyone insists on having a display (which is fine, I was trying to avoid feature creep) then I highly reccomend using two 10-LED bar graph devices rather than numeric output. It will be difficult (if not impossible) to show rapid transitions with a 3-digit numeric display i.e. a quick lean condition when you stomp on the pedal! A laptop can be used for precise A/F outputs.

Maybe Mark had the right idea - make it modular because not everyone is going to want all that. I know I don't - it just makes the cost creep up and up and up.

However, if Craig thinks he can do it without the cost going too unsightly, then I'm all for it.

Grumpy

DIY, means Do It Yourself, IMO.
The original one was for the average guy to be able to build. It used cheap easy to find parts, analog circuitry, and good clean simplistic design. From an engineering stand point, even after 5 or so years, I don't see anyone making any improvements on the core design. That in it's self is solid testimony to how well it was done. There's been a total of 2 people, that I received boards from, that couldn't master soldering, and I replaced there problem boards with functional ones. Part of the process of being a true DIY, means learning new skills and mastering new software from time to time. *We* avoided going digitial so that the average guy didn't even need a programmer, or had to buy special chips to be able to build it.

After a while, I didn't keep track of the letters from guys that learned how to solder, and learn basic trouble shooting from having built one.

I can't think of a thing we could have done better.

MTPFI-MAF

Here is the Pdf that jim e-mailed me. Not quite as technical as I hoped it would be but maybe it will help.

EDIT ok try this: http://www.f-body383.100megsfree5.com/lsu4pdf.htm

Cobra289

I get the next:
A Hotlinking Error Has Occured!
Regards,
Cobra289

Synapsis

I mirrored it here as well. Just in case.

http://www.blackcow.org/lsu4.pdf

Mangus

Waaaaiiiit a minute. You mean you don't have GOBS of spare time like Craig and I?

Modularity is key. Readouts are good, but there's zero reason why an external, modular readout can't be made to work off of the existing outputs (either raw, 0-1, or 0-5 linearized).

I think the controller itself should be as simple and cheap as possible. This sentiment gets repeated in this thread I believe because its so important!

Oh, and you can naturally assume full native tunerpro support for any data acquisition for this beast.

funstick

ill see what i can do in person down at bosch for some actual electronics drive stuff. it would be good to know the actual current requirments etc etc etc of the sensor as well as its electrical response curves.

ill see what i can do.

RBob

iTrader: (1)

Full specifications for the Bosch LSU4 series of sensors are all over the net. Many questions will be answered by finding these.

As for feature creep, yep, already too much. A basic box is all that is needed.

Please skip the NB simulated output. It won't do any good as the response time does not and can not match a true NB sensor. Just tune it in open loop, wack in the MAF/MAP tables, and be done with it.

RBob.

Mangus

Not only that, but once again if this is desired, there's no reason why yet another simple inline module couldn't create this type of output from the existing rawor 0-5v outputs.

jamesbob02

I am totally against a visual display as a mandatory standard. I don't see why one couldn't just hook up their preferred voltmeter and read the output? This seems like the very definition of feature creep for me, because I have a feeling most of us would be happy having it in the ALDL stream and just datalog it. That's what I intend to do, anyways.

As for the original DIY, I think the only problem is that it only works with the more expensive sensor, which is reason #1 that we're trying to make the simplest, cheapest one we can.

If you guys think you can put together a really simple DIY design that I think I can put together, that's cool, but I was hoping this would be a Craig or Bill thing exclusively so they could build them themselves and sell them with a premium. That's what I pictured, but I'm cool with it becoming a community thing, if it's simple enough.

For the record, what I'd like to see is simply WB data in the ALDL stream so that it can be directly compared to RPM and other readings. I'm not sure what all would need to happen because I lack the skills, but I'd like to be able to hook this thing up, and see the readout directly in the ALDL logging of TunerPro when that's completely finished, interpret the results and burn my chip. I don't need anything further than that, and I don't see why there wouldn't be a lot of people who feel the same way.

What do you guys think?

MTPFI-MAF

AMEN. I would rather have the Read out directly in TunerPro Via The New ALDL dash That Mark is woring on among other things,

I personally think that the DIY approach is what is best for me because it will hone my soldering skills, while helping me learn some new info by being hands on, which in turn will lead too a new understanding of things and help me to Proceed better on my own. I think it would help a lot of people who are afraid to Build things their self.

Just my .02

jamesbob02

I'm cool with that. It would definitely keep the cost down. I dont know much about electrical engineering, but I can solder without much trouble.

MTPFI-MAF

ok I found another source of info it explains a bit more than the last.

craig I think this might be helpful

http://www.f-body383.100megsfree5.com/lsu4pdf.htm

Click on the one in red.

JP86SS

iTrader: (1)

Good job finding the data !! :yourock:
That doesn't hurt the effort.
I think that the whole NB end of things was so that the WB could be installed in place of the NB and run that way after tuning.
That's all.
If it causes problems that can't be overcome due to speed issues or are deemed "creep" than forget it, or like Mark suggested just get the signal from the WB 0-5V output.
It was just to avoid additional drilling and welding and a set it and forget it mentality.

JP84Z430HP

If I'm not mistaken, it seems that it's really easy to implement the sim NB stuff at the level Craig's talking about.

I didn't know that th eNB had a response time that would be slow enoug to cause a problem. Anyone have any spec of the average response time on a new WB? I'll look at the stuff that's been linked in the post to see if I can find anything.

Synapsis

http://www.blackcow.org/LSU4.2.2.pdf

He asked me to mirror this one too, but I was watching a movie at the time.

Craig Moates

I think that second file will be most useful.

OK, no display then. I might go ahead and include provisions for a digital display interface though if it is easy enough. It should be mostly firmware really. The reason I'm saying have provisions for it is because that would keep it at the digital level. We can still have it modular, such that an external display module could be added, but at the digital level we preserve some accuracy rather than going through another daisychain of D/A->A/D conversion.

Anyways, I'll say this:

"HOPE NO ONE's IN A BIG HURRY!"

Let's kick all the ideas around a good bit. I'd like to draft a baseline schematic within a week or so. Then we can start the real design process ;^).

As far as kit -vs- assembled units, the opinion seems to be that a kit form would be appreciated. I'm planning to set it up with a little bit of surface mount stuff to try and keep the costs down. If everyone's OK with that, then that's fine. We can do kits. Troubleshooting will be a bear though, and then there's the initial firmware. Hmm, maybe not surface mount? That'll drive the cost up. Anyways, kick the idea around some and decide how you want to go. It's a ways off to be worrying too much about that, but still...

Again, about a week at least before a first-pass schematic. Then we can start kicking it around and get some more input to try and make sure the design is robust and reasonable.

vernw

iTrader: (2)

I for one applaud the idea of a NB simulator so that we can install and use the WB sensor all the time! Like most folks, I only have the one O2 sensor bung. Hope this ends up being possible to include w/o too much trouble or additional cost...

Again, I'll also volunteer to be a "dummy/tester" on assembling a kit, trying out the interface and/or installation w/o being intimately knowledgeable with the product to begin with, etc. If it matters, I'd be using a 7730 ECM running a speed density prom on my 383 SuperRam setup with a manual transmission.

Hope I can help out in some way.....



- Vern

MrBill

heheheh, let's just say I can relate. I like it that way though, I HATE having nothing to do!

jamesbob02

I'm supportive of the NB sim idea only if it's reasonably simple to incorporate, that wouldn't drive up the cost or complexity of the product.

Personally, I have no problem cutting a hole and welding in a bung for the WB sensor if that's what it comes to, but if I can just throw it in the stock location, all the better.

Craig, I'll definitely lend my support for testing if you need it, but I don't mean that in the "yay I get free stuff" sense. Since this is a community effort, I'm more than willing to stomach the cost, take a schematic of yours and put together a board from Radio Shack pieces and give my results, even if I don't possess an intimate knowledge of the circuitry. Whatever is needed, I just want to help and speed along progress.

Ken73

I'm not sure I understand this, Bob. Is the NB slower, or faster?

You might have to help those of us who don't know how to tune in open loop....

JoBy

This is a simple 'moving dot' display I made that works well with the DIY-WB. It is very fast and easy to read.

http://jobyteknik.homeip.net/corvett...D_circuitboard





And a movie:

http://jobyteknik.homeip.net/corvett...d/MOV09033.MPG