Has anybody tried this:
http://www.streetrays.com/catalog/pr...830c718f83ce01

Looks like it will take the 0-5vdc output from the Wideband controller for the display and convert it to a 0-1vdc output that we could input to the ECM at the stock NB O2 sensor terminal.

Will this work or has anybody checked in to this.
Kinda sounds like it will work.

 

Speedreader is the same thing it appears. Bill and Ken? are working on it. do a search, should net u something.

Were testing prototype on my car at the last pow wow

later
Jeremy

 

**Kinda**

Problem is that it's says it's for there M200 controller, so you need to know if you WB has the same output as there's. Then you need to know of the voltages as they skew from 14.7 / 2.5v are the same. Could it work, probably/ sorta. No bets on how well, it's one of those 50/50 things.

 

Does your car have a cat, if not, open loop and save money. Open loop = no converter, no narrow band sensor replacement, no cat to go bad, better gas milage, stronger engine.... wow, closed loop stinks doesn't it .
If you've got a cat like me then closed loop is required to keep the cat working, but every 2 years is when they check.... so if it looks like a cat, has a wideband sensor bung welded to the side of it, it might still be a cat, or is it .

 

Does it meow? If so, then it's definately a cat......

Back on topic:
How is this anything like the speedreader?

I really doubt it would work to replace the stock NB O2. I think it will just drive the AM display. I would think the bias voltage the ecm puts out would make it all jumbled up.

 

Wonderful,
just wonderful,
now Doc's out, checking all the cars in the neighborhood to see if they have cats. Yes, he's just walking up to each one and going *kitty, kitty, kitty* and waiting to see if one emerges.
Oh well, at least with him doing that, there's no one to be driving the truck for *almost* skiing purposes.

 

Jeremy, we had a narrowband AF meter that was non-linear that we tried out on your car at the pow-wow, but the SpeedReader turns an old TBI ('7747/'8746) ECM into a more powerful ECM with fast datalogging (hence the name) and converts it to a ROMless ECM so you can rewrite the code.

Not quite the same thing.

I think you can *make* your own WB->NB converter for aftermarket gauges quite easily with a very simple resistor network. Cost about $.50 or so.

 

The question here though is not to make the WB drive a guage. I believe it is whether using this converter to take the 0-5v WB output and converting it to a 0-1v output that could be hooked into the ECM harness's NB input line. In effect, trying to drive the ECM with the more accurate (albeit a lot more expensive) WB sensor.....

 

Seems like just mapping 0-5 V down to 0-1 V would not work very well. The ECM wants to see those cross counts coming from the sharp transition around 14.7 on the NB. The wideband would be very smooth and linear and I bet the ECM would not behave as expected, at least not without retuning or coding to compensate.

I could very well be wrong though. And maybe these converters aren't just voltage dividers, I don't really know what they do exactly.

To me it would make more sense to use a wideband in conjunction with a different algorithm altogether; one that didn't mess with cross counts and just closed the loop with the goal of keeping steady on the desired AFR. Assuming no cat that is.

 

From my understanding, using my drivability training, on a stock wideband application, the O2 should still swing above and below 14.7:1. That's how the instructor explained it, but I haven't looked at one with a scope to see.

It is interesting to watch an O2 sensor with a scope to see how much it actually does swing across 450 mv! Supposed to swing between 200and 800 mv, 5 times a sec at engine speeds above idle (around 2000 is a good place to test.) It also has to go above 900 within like 100ms when you stab the throttle, and drop below 100 when decel-ing. This is all for stock driveability testing.

 

I sure can't think of any good reason for the AFR to swing wildly if not either to count cross-counts or to somehow satisfy a cat converter. The guys who run open loop seem happy with the constant AFRs they get. If it was important, wouldn't the ECM oscillate the fuel delivery in open loop too?

 

My cats stay warm. They're sleeved

 

Swinging above and below 450 mv IS to satisfy the cat, period. Even on a stock wideband application (Honda Civic VX for example) it swings above and below the voltage for stoich. While it swings slightly lean, the cat stores oxygen to help it burn what is extra when it goes back the other way to slightly rich.

IMO, Cross counts as seen on a scan tool aren't really much use, other than the fact that there are some, and not near zero. From what I've seen, cross count numbers don't actually match how many times the O2 sensor voltage crosses 450mv.

Anyway, I've gotten off topic here. I think as long as the wideband converter made the ECM see what it's expecting, at the rate it's expecting to see it, things would be good.

What would be even better would be to hack the source code to use a 0-5 volt input for the O2 instead of the NB input. I'm just barely touching cource code right now, and nobody that I know of has even worked on it very far due to the complexity of what it would take to do so.

Another option would be to take the 0-5 volt input, grab a value from a table that you had setup and spit it out as a 0-1 volt value that the ECM could use in place of the NB.

I'm just thinking aloud here on these ideas, feel free to shoot them down!

 

Pin in the wideband o2 voltage into an unused AI pin, source code a desired AFR table vs RPM vs MAP. Get an alogorithm that'll adjust the injector pulsewidth to get the desired AFR . I'd also like to get code that adjusts the desired AFR by cts. Then get your correction terms, tables, etc. and turn closed loop off . I wish it were that simple but AE really isn't "easy" with wetflow, not to mention this rom on board junk.
If I were port injected that would be my major focus... a closed loop wideband operation for getting close.

 

I think that such a system would be almost "self tuning", in that you could be way off and the ECM would know (through the algorithm) how far to adjust to get back to the desired AFR. The only thing is that the current BLM setup (4x4 cells) would be the holdup, with the lack of resuolution and all. I'm seeing that the algorithm is the key to such a thing working.

 

This is going off on a tengent a little, but if you were doing it right, would you go with IAT instead of CTS? It seems to me that the correlation between IAT and CTS cannot be relied on and IAT is what you really want for fuel calculations.

Has anyone modified their 7747 to use IAT? What application would be good to take the actual sensor from?

 

The CTS is actually more important, especially for the warmup cycle. I don't know exactly how to explain it correctly, so I won't go any further into it. The IAT is also important, but not as much so as the CTS. Maybe someone can elaborate on this for us?

 

Check out this thread: Speedreader

When this is finished, the TBI guys will have it all!!!

 

CTS tells you heat soak and oil temps which change the AFR just as much if not more than just IAT on a n/a setup. Now on a turbo car the IAT is very important because it can vary by such a large ammount.
IAT, CTS, all that stuff could be done pretty easily but the ultimate code would have a full scale table for every IAT range and CTS range, that's a LOT of table memory . So you'd have to datalog everything and let your laptop do some calculations for the IAT and CTS scalers... doesn't take long. Besides, with a cold engine (low CTS) the engine generally wants a richer AFR so you can tune this with a scaler pretty easy. Example; desired AFR (warmed engine) * multiplier from cts lookup. That way on the warmed engine the afr would be say 13.8 , a cold engine it would be multiplied by 0.9 and the resulted desired AFR would be 12.4 .
Also Judd, the speedreader wouldn't be limited to TBI's, it's good for the ccc and grand national guys as well! The only draw back to keeping the c3's is the processor speed, it's half that of the p4's. So there is still reason to go with the p4's, just not as much .

 

Thanks for the better explanation! I sometimes can't put my thoughts into words very well!

 

You can make a device very easily that will cross at 14.7 AFR. You need single supply op amps and an adjustable voltage regulator. Set up the op amps so they saturate at +1 and 0, just as the wb voltage swings thru stoich. This way, the circuit draws no current from the WB, and can provide more than enough current for the ECM. Using a resistor network is a bad Idea, unless to like impedance matched devices (circa 1950).

Just my $0.02