Don't know the 95 STi wiring, but if the WRX RA is the same there was a post recently on it and Cosie Convert posted a picture. It was in Drivetrain forum.

 

STI 95 is probably the same as all pre97.

It's the 4th wire but is actually the 6th loaction, as 1 and 5 are unused.

 

- Being trying to get the AFR working last night but to no avail.I have used Scott board design (Much appreciated btw) When checking pin 4 & pin 6 & trying to adjust them via the pots - I can't seem to get voltage on either pin. When I check pin 7 & 8 I get usually the same voltage of around 8.44 when using a 9V battery.

I don't know enough to tell if something is not working right & may need to be replaced. Can this be built in stages & tested in stages rather than plugging all in together & testing then.


- Spent a bit more time testing the output signal from the lambda sensor again. It seems to be reading low compared to some figures on the thread for the AFR. It reads mostly below 600mV, which would mean that if the AFR were set to light up between 700 & 900 it wold hardly light up at all.

I did test the heater connectors for resistance measurements but got nothing. So I am now beginning to take a view that my Lambda sensor has failed & needs replacing. Am I right with this assumption - just don't want to have to fork out if it's not gone.


Regards

Johny.

 

It does not sound like your lambda sensor is working correctly if there is no offset. They have a reasonable change of being duff on a 6-7 year old car.

All I can say on the AFR is check everything again. If you have power and ground connected, then to get the references going all you should need to do is connect the potentiometers and the reference connections - do not need to connect any of the LED outputs. I can't think what else to suggest. Not really much else to go wrong if the chip is healthy.

 

Cheers John!

Will try again tonight - not great to be trying to concetrate on such small areas when you're tired tho.

Regards

Johny.

 

When warmed the Lambda should pulse approx every 2 seconds with a reading of 700mV, although a DVM will only measure about 600mV as it is averaging the signal.

Increase the revs and the reading should still pulse but may increase to about 900mV.

You probably won't be able to get it to go closed loop until you go for a run and the ECU see's some boost.

 

Sorry if I'm re-hashing the same stuff. I'm still trying to figure out the chip & pot 1. When I hook up a PP3 9V battery to the chip & test pin 7 & 8 they roughly show near 9V with a bit of difference between the two.

Then when I connect the pot, T1, as shown with middle leg to pin6 & the two outer legs to pins 7 & 8 & monitor the output of the middle leg or pin 6 I still get around 9 V.

I have tried the pot seperately & it works fine.

What should be the output voltage on pin 7 & 8 when power is supplied to pins 2 & 3?

Is pin 6 supposed to get a percentage (depending on the pot setting) of voltage from?? pin 7 or 8? How come they both have near 9V?

Any info appreciated as my heads cookin!

Regards

Johny

 

Johny,

I'm not sure if I'm following you correctly but you should be measuring pins 4 and 6 (for 700mv and 900mV). By adjusting the two pots, you should set these values.

Only then would I both applying power to the input to test it....

Hope this helps,

Matt

 

Sorry Matt - bit lost here! I haven't applied power to pin 5 at all. Just + to pin 3 & - to pin 2 & monitor pin 7 & 8.

They roughly still give near 9 V! When I use the pot (T1) - on side leg to pin 8 & the other to pin 7. Then middle leg to pin 6. I monitor the middle leg of the pot it still gives the same voltage near 9 no matter how I adjust the pot.

So I'm just using the chip, one pot & power. This is just to test as i go along in stages as I built it twice already & cant seem to adjust the voltages at pin 4 & 6 when I built it.

Hope the above isnt too confusing to follow.

Regards

Johny

 

T2 needs to be connected so that pin 8 output can flow through a resistor to ground, so I wouldn't expect it to work as you have it.

 

i think i may have a problem, i've tracked it out on a pcb, wired it up on the bench and can only get led10 to light up.
no matter what i do with the pots.

i assume its a duff ic.

i will try again in the morning.


Kevin

 

Just for reference a scrap piece of veroboard or similar makes a really good template for drilling the box.

i.e. drill the veroboard with how you want the LED's spaced, then use that as the template on the box.

 

you noticed that the first led was a bit off

 

It makes me very happy to see all these pictures Glad the circuit has been of use!

 

yes, well done mate
i've still got to fit it yet.

 

which wire on the ecu should i use for the signal inputkevin

 

depends on MY. have a look at the link manual and the thread in gen tech -Dawes AFR and connection thereof.

 

Kevin

I did the same myself. Then used the veroboard for a template and drilled the other side of the box much neater. No one would ever know....

 

Kevin, looks best in the "right hand" of the driver pic.

Nice to see a finished job, me and the Mole are going to fit ours soon...it is in a smaller box (I think) and has a "bar graph" led set up.

Pics soon

By the way Mega, Moles made yours...just got to find a box.

Cheers, MD.

 

I suspect either a short or a dodgy T1.

There is no way if there is a resistance between pin6 and pin7 and pin6 and pin8 that the voltage would not alter.

Suggest you disconnect the IC if you have socketed it and measure resistance between these pins and move the pot.

Try another pot and double check for shorts.

I am away for the rest of the afternoon, so you are on your own. I expect a glowing report later on!
If you are getting 700mV I doubt the chip is dodgy - they have a very low failure rate.

Perhaps I should start an 0898 49p/min support line

[Edited by john banks - 1/5/2002 1:39:53 PM]

 

Seems like JMCA is getting the same readings as me from his Lambda, is the output different on the Pre 97 cars? I noticed on the Link manual if says 0 - 0.7 volts, which is in line with what I'm getting.


[Edited by Ken Ross - 3/5/2002 11:19:34 AM]

 

I hope you don't mind me digging this up!

I'm in the process of making this and whilst doing a few calculations for LED current (Iled) noticed that according to the LM3914 datasheet Iled is programmable from 2mA to 30mA.

Iled = 12.5/T1

T1 is the pot in John's schematic. I'm trying to drive these superbright LEDs which typically run at 20mA so I'm going to try a 1K pot and see how bright they are.

So if you want to drive your LEDs harder, try reducing the value of T1.

Also, if pin9 is connected to +V it will be in dot mode, which I believe means only one LED will light up at any one time.

I'm going to try it in dot mode, with two 1k pots for T1 and T2, REFlo at 790mV, REFhi at 890mV and taking red, yellow, green and blue LEDS at 850, 860, 870 and 890 respectively.

Lets hope I'm right about setting Iled, I ordered 2 chips just in case

Justin

 

I worked out T1 and T2 by trial and error as the led current interacts with the setpoints for the bottom and top LEDs! There is a method described in the datasheet or application note using extra components where the current and setpoints can be set independently, but I never got around to doing it.

 

Having re-read the datasheet too many times, I think its current drawn from pin7 that determines Iled - a bit tricky to work out, time to play with the values

Justin

 

I put this on the breadboard last night. I was going to use an LM317 to provide a seperate 5v rail to run the LEDs off, then power the lot from 12v from the, the, where can I get 12v from? Radio, phone, maybe one of the ecu sensors with a 12v feed.

Paul

 

I took mine from the radio.

 

Pavlo,

Behind the Stereo/Ashtray there is a white/clear coloured connector with female spade terminals in it.
This will give you 12V & Gnd, but only when you switch your side lights on.

Thats what I use, just have to remember to switch the lights on if I want to check AFR.

Meter is also mounted in the ashtray, so open/close it when you want to check whats happening.

 

Just to update - it worked

Using 1k pots for T1 and T2 gave me 15mA for Iled, which is great for the superbright LEDs I chose - except the blue one, it was a different make and so bright I had to use a 15K resistor inline to get it at the same level as the others.

Here is a pic of my efforts


The cicuit is in the bottom right of the picture, my test one is top left on the breadboard and I found a nice place to put the LEDs in the pod

Components used (RS No. in brackets)

LM3914N (308-174)
Red LED (590-519)
Green LED (590-525)
Yellow LED (590-531)
Blue LED (320-9991)
LED holder (589-569)
1K pot (186-996)

In the end I set Vhigh at 880mV and Vlow at 780mV, taking red, yel, grn and blu at 850, 860, 870 and 880mV

Close up of Pod



The pod is <97 52mm A pillar from Power Engineering, the EGT gauge is from the Phantom range at Scoobymania. Its all going in this weekend - I'm a bit excited

Justin

 

Looks very good.

 

I was going to put the LEDs on the end of a small cable, and put it on the sun visor. That way they should be right in front of you, bu tyou can see they without having to take your focus from the road.

Paul