2002 Ford Taurus solar radiation sensor repair help

[poot36]

I have a light sensor from a 2002 Ford Taurus that is not working at all.  I have determined that it is not drawing any current.  I am having trouble finding a replacement sensor and am wondering if there is a modern replacement I could use.  I have attached pictures of the inside of the sensor to this post.  Pin G is ground, pin T is 5V and pin S is the output (is pulled up to 5V by the car so I assume that the sensor pulls it to ground to signal that it is light out but I could be wrong).  I think it is sensitive to green as the cover over the sensor looks opaque to me (but under strong light lets a green color through) and the circuit board is mounted in a black plastic enclosure with a frosted black looking (but semitransparent cover) over it.

[poot36]

More pics.

[KMoffett]

Photo of reverse side?

[coromonadalix]

have you tried to find them in kenny u pull scrap yards

[poot36]

I don't really want to go to a junk yard if I can help it (who knows if the unit you pulled works).  Here are pictures of the back and the case the PCB sits in.

[poot36]

More case pics.  Note the last photo shows it partially assembled.  Ignore the fact that I used the PCB to support the plastic case to get a better picture.

[picburner]

The new code of the sunlight sensor is: YF1T-14A597-AB
You can find it new or used on eBay.

[poot36]

I have that code on my sensor.  I found a used one on eBay for $50 CAD + what ever the import and duty fees would be.  I would expect the sensor to be at most a $10 part if I could find a compatible replacement.

[poot36]

Well I powered it up on the bench again and it started working!  I took some measurements and found that the current draw increases with the amount of light the sensor sees.  Also the output pin will swing from high when it is dark to low when it is light (this is with a 10K pull up resistor on the output pin connected to the 5V line).  It never reached zero volts or the 5 volts of the power supply though.  Lowest I saw was approx 1.8v and highest was 4. something volts (most likely 4.4V).  The output voltage did change with different pull up resistors but the amount of change did not change very much unless I went to high with the resistor.  I did confirm that the bond wires in under the clear blob cover are the issue so that is not fixable by me.  It would stop working when I pressed on the side of the clear blob.  I took a look at the die under the microscope and could not see any manufacture marking or model number but that could be because me scope is one of the cheap toy scopes that reports to have up to 900x zoom.  I used it on the 100x setting and that was enough but trying to light the die from above was not easy due to the short working length of the scope.

Does anyone have any idea what type of sensor I could replace it with that would get me into the ball park for correct sensitivity and triggering?

[amyk]

What is it for?

[poot36]

It is a sensor that turns on the exterior running lights when it is dark enough and dims the interior cabin lighting as well as adjusts the way the AC operates in response to the amount of sun that would heat up the cabin.

[james_s]

That sensor looks like a photodiode with an IC on the same die. The first thing to do is figure out what the output is supposed to be, it's possible that it's a linear DC voltage, or it could be a PWM signal or a variable frequency or even some kind of digital bus.

[poot36]

I had it working intermittently and was able to confirm that the output was just a high or low signal as far as I could tell that changed when enough light was detected.  The other interesting thing was that the current draw of the unit changed with the light brightness, the brighter the light the higher the current draw.  It went from approx 1 to 2 ma when I covered the light sensor to 12 to 15 ma when I used a bright  led flashlight on it.  This was with no green cover or frosted defuser in front of the sensor just the clear coated sensor on its own.

[floobydust]

It's kind of a dog's breakfast sensor? it looks like it originally started out as a TO-5 package phototransistor then something more sophisticated was bonded to an adapter board.
I would say it's a custom part for Ford. The sensor might be working fine, these things have a very long delay before they trigger. But they seem to be unreliable on Taurus forums.

The old part number is YF1T-14A597-AB, new part number is YF1Z-13A018-AA sunload sensor. You might have better luck with the new part number.

[amyk]

Have you actually scoped the output to confirm it's not PWM but a simple high/low binary output? A DMM is going to read the average voltage so a high-frequency PWM won't be distinguishable.

In any case, the function and interface suggests you should be able to easily create a compatible replacement.

[james_s]

If it really is just a high/low output then designing a compatible replacement should be fairly trivial. A photodiode and an op-amp configured as a comparator with a bit of hysteresis ought to do it. You'd have to play around to get the thresholds right and maybe add some delay. Or if you're more comfortable in the digital realm there are inexpensive digital light sensors that are easy to work with, combined with a 8 pin microcontroller you can code it to operate any way you like.

[poot36]

@floobydust I am aware of the two different part numbers and have found equally expensive replacement parts with both.  @amyk I have scoped the signal and it is just a high or low signal, no PMW or anything fancy.  I was thinking of using a circuit like this one: https://www.maximintegrated.com/en/design/technical-documents/app-notes/1/1909.html but I am not sure that is the correct thing to do.  Not sure if the LDR would be sensitive to IR or UV, I guess that is what the green cover and frosted cover should filter out.

[james_s]

The datasheet for any sensor you are considering will show the sensitivity curve across the spectrum, this has some generic information. https://eepower.com/resistor-guide/resistor-types/photo-resistor/

I would guess it is looking primarily at IR which is plentiful in sunlight but not in sources of artificial light. You don't want your headlights turning off whenever you drive under a streetlight. The color of visible light that passes through the filter doesn't necessarily tell you much about the IR or UV characteristics.

[poot36]

Well I ended up going the junk yard route and got a used sensor for under $10 and it works so far.  It also has a 30 day warranty I think.  Thanks for all the help though.

[poot36]

Well the junk yard sensor failed in under a week and then I got really busy so was unable to return it.  I have since replaced it with a out of production chip that gives a logic level output when the light gets bright enough but I do not think it is properly working (the HVAC system test passes though, I think it also needs a change in current draw as well proportional to the amount of light it gets).  I have finally got the microscope in so I have attached pictures of the die in the one that I disassembled to install the replacement chip.  Sorry about the cracks in the clear gel like covering, I was trying to see if it was the bond wires that failed by pressing on them with my fingernail but ended up damaging the gel instead.  I can see a die marking of ETIC on it and have found these two chips http://www.bfdisplay.fr/fiches-techniques/opto-etic/ETIC%20-%206501ds%20rev%20A2.pdf and http://www.bfdisplay.fr/fiches-techniques/opto-etic/ETIC%20-%206505ds%20rev%20A2.pdf that are similar but do not provide a logic level output as far as I can tell.

[dzseki]

Texas Instruments have the OPT101, which is also a photodiode+opamp with linear output, but it is not cheap either...

[Rasz]

You already had a working one in your possession and didnt bother measuring its electrical response, so it doesnt seem like something high on your agenda. Get another one from salvage yard and get on with your life.

[poot36]

I found a light sensor from some old electronics that put out a logic signal and managed to get it installed in the space of the original sensor.  Now I do not get the error on the HVAC system but the auto light system does not function.  I am suspecting that along with the logic signal it is looking at the current draw of the sensor and I have no way of controlling that.  Thanks for the help but I am going to leave it as it is as the car is nearing the end of its life so is not worth putting much more time into.

[james_s]

I really doubt it has anything to do with the current draw, the output of your new sensor isn't correct to trigger the lighting.

[poot36]

The new sensor just has a high/low output and it does make the test routine in the HVAC system happy but I guess that it could not be getting enough light or is getting to much.  Not sure.  I do know the HVAC system has a way of detecting a shorted sensor or a open sensor so I guess it has to measure the current through the circuit for that to work.