Replacement light sensor cover in situ

[TheHypnoToad]

Just a thought on this given the material used by MG is highly coloured and reminds me of the Lens/front of an old infrared remote.
Could the sensor actually be measuring IR light rather than visible light given that IR light will be abundant during the day and completely absent at night?
Will have to think of a way to test this.

You could use a screw on IR filter that is intended for camera lenses, some of them even have docs on what wavelengths are filtered.

Failing that, you could try pointing your TV remote at the sensor and see if the lights turn off

 

[Daniel Brown]

If the sensor is basically uncovered altering the sensitivity from being to sensitive to shadows won't this now have the opposite effect and every time you go under a street light the headlights will turn off

 

[Rolfe]

Nobody who has done this has reported this happening. I've only done a couple of drives after lighting-up time since I made the change, but I certainly didn't notice the lights going off in built-up areas with street lights.

The lights still come on in early dusk. The first time I was out after the change was made I got to where my car was parked about ten. Sunset, and hence lighting-up time was about 21.50 that evening. There was still a lot of light in the sky and I could have read out of doors without too much difficulty. I would say there was a lot more light from the sky than from the street lights. The lights came on and stayed on.

A couple of days later I went out to the car to drive home round about lighting-up time. The lights came on when I started the car. I was parked in a shady corner, but they stayed on when I drove out into the open road under the open sky. They still stayed on when I got into the village and the (pathetic LED) street lights, and they still stayed on when I got into my own street, which still has brighter incandescent street lights. Even when I drove right under the street light at the end of my own drive.

 

[Daniel Brown]

Nobody who has done this has reported this happening. I've only done a couple of drives after lighting-up time since I made the change, but I certainly didn't notice the lights going off in built-up areas with street lights.

The lights still come on in early dusk. The first time I was out after the change was made I got to where my car was parked about ten. Sunset, and hence lighting-up time was about 21.50 that evening. There was still a lot of light in the sky and I could have read out of doors without too much difficulty. I would say there was a lot more light from the sky than from the street lights. The lights came on and stayed on.

A couple of days later I went out to the car to drive home round about lighting-up time. The lights came on when I started the car. I was parked in a shady corner, but they stayed on when I drove out into the open road under the open sky. They still stayed on when I got into the village and the (pathetic LED) street lights, and they still stayed on when I got into my own street, which still has brighter incandescent street lights. Even when I drove right under the street light at the end of my own drive.

Thank you for letting me know this.

 

[Rolfe]

I've seen others speculate that the sensitivity of the sensor has been set right for the uncovered instrument, and then they put the cover on it... It certainly activates when dusk is just beginning to gather, when there's a lot more light around than is the case during the night with only artificial lights.

 

[Rolfe]

Well this is annoying. After believing it was all sorted, I realise it's not, not really.

Around 3.30 I took the car up a very tree-lined road to visit the bottle bank with some empties. On the way back, at the point where the trees were thickest, the bloody lights came on again.

I didn't have time to go right back, but I returned about two hours later, minus the new sensor cover. All fine, lights not coming on. So I thought, time of day, need another run with the cover on. I stopped under the trees to put the cover back, and immediately I did that, the lights came on. OK, not good, maybe it's the material I used to cover the holes, perhaps that's not as transparent as it should be. I'd done a better job glueing it on than I realised, but I got it off. Same thing. Put the cover as printed on, lights come on. Take it off, lights stay off. This is a bright sunny afternoon, and nobody else's lights are coming on under these trees. It's embarrassing.

I took the cover off again and did a complete run up and down the avenue like that, and it was perfect. Lights stayed off.

To my mind this really needs a completely transparent cover. Not quite sure what I'm going to do in the interim. I can't really drive around with the sensor completely naked, but that seems to be the only way it's working correctly.

How about you, Siteguru?

 

[siteguru]

As I posted elsewhere, mine's OK so far.

What you could try is to cut/file away the pattern and just have an "open" cover. Then fit a transparent cover. That way the sensor is exposed without any housing material blocking it. (That's what I'll do if mine exhibits the same issue, although the pattern on mine makes it easier to do).

 

[Rolfe]

There are other possibilities in the 3D printing file too.

 

[tsedge]

I appreciate all of these efforts to solve this issue, there's lots of ingenuity here. And it is quite true that MG need to fix this.

However, I just flick the lights off when I start every drive and control them manually and that works for me, takes but a moment. There is the problem that I sometimes leave them on but I am training myself to look at the dash for that warning.

In general I don't like automation in cars unless it is completely autonomous as I think good driving requires attention and focus and deciding when to put the lights on is not something I want the car to do for me. I worry that we are creating bad distracted drivers because the car is trying to do things for them but sufficiently imperfectly that this creates it own problems.

All my respect to everyone here who is trying to find a proper fix, the DIY route can be a great one (eg Macadoodle's oil leak fix).

 

[Rolfe]

One more supposedly "automatic" thing to turn off manually before we start, because it doesn't work. Nope, not if there's a fix.

I've had a car with automatic lights for nearly 14 years. I am completely unused, now, to deciding when to turn them on manually. This is mature, working technology, and something else MG has screwed up for no good reason.

 

[Torryhun]

I must be really lucky with my auto lights, haven't encountered a problem with inappropriate lighting up, work just as good as my old Insignia. Also loved how the auto full beam worked on the drive down to Edinburgh Airport at 2 am last weekend.

 

[Rolfe]

You must be almost unique! Do you have your central screen on auto light/dark?

 

[Torryhun]

You must be almost unique! Do you have your central screen on auto light/dark?

Dark

 

[Rolfe]

So you'd have to notice the actual lights indicator coming on to know the lights had come on. But that's pretty obvious, I think you'd have seen it.

 

[Torryhun]

So you'd have to notice the actual lights indicator coming on to know the lights had come on. But that's pretty obvious, I think you'd have seen it.

Definitely would have.

 

[Kithmo]

My old ZS auto lights were perfect, it had a round sensor, light coming equally from all directions on to the sensor. I wonder if that's anything to do with it, the 4 one being this trapezoid shape.

 

[Torryhun]

My old ZS auto lights were perfect, it had a round sensor, light coming equally from all directions on to the sensor. I wonder if that's anything to do with it, the 4 one being this trapezoid shape.

Could be on to something, the one in the Insignia was rounded. ?‍

 

[Rolfe]

I think it's simply that the sensor has been set up to work perfectly with the ambient light unobstructed, and then someone added a semi-opaque cover.

For the record, I just did a test. Time for the lights to go on after chucking a magazine over the sensor, two seconds. Time for the lights to go back off after removing said magazine, ten seconds.

You'd think that was reasonable, but it can still manage to flash on and off repeatedly when driving along a road with intermittent light and shade, for example widely spaced trees.

 

[Kithmo]

I think they've got that programmed the wrong way round then, I'd rather have a longer delay before they come on so they don't switch on when I go under a bridge.

 

[Rolfe]

I don't really agree. I don't think you want to be driving without lights for ten seconds if you enter a tunnel. Also, it seems reasonable to wait a bit to see if it's OK to turn the lights off, because that's less critical. It's just that it doesn't seem to work, whereas the same facility on other cars just works without any fuss.