Low light level switching

[mcockerell]

It's nearly the longest day, and once again I find myself having to adjust control of our hall lighting - so what's my problem?

Our hall is in the centre of the house with no direct natural lighting; when it is light outside it gets plenty of light from the surrounding rooms and also a skylight onto the landing above. During the winter months, when it may be relatively dark outside, we may need to have the wall lights come on by mid-afternoon - I have that working reasonably well. But at this time of year, when it remains relatively light outside until well into the evening, we hardly need the hall lights to come on at all - there is enough residual light to see clearly, but the lights still come on in the early evening.

My difficulty seems to come from the sensitivity of the light sensors at low-light levels - there is a significant difference between switching at 1% and 2% light level, even with zero hysteresis. And using a combination of internal and external sensors doesn't seem to work well for this particular application.

So my question is this: Is it somehow possible to have more granularity across a light sensor range of say 0% to 4%.
Or is there an alternative way of addressing this? We have a motion sensor in the hall, but when it's dark we really need the lights on to enter the hall.

On a separate, but related topic: Is there a clever way to set a light so it will not come on if it's (say) within 30 minutes before the end of its operating period?

Any help much appreciated.

Martin

[chris_j_hunter]

have you tried General Logic - not sure I sufficiently understand the ins & outs your problem, but it could provide a way, almost certainly ...

[Karam]

Aside from any other way, might be able to provide firmware to stretch the lower end a bit at the expense of some of the curve above it. Let me know which module is doing the sensing (send me a node profile for it).

Not sure I understand the second part - if you don't want it to come on 30 minutes before the end of its operating period why not move the operating period back 30 minutes? Or is it the case that you want the operating period to still act as a light turn off at present setting and don't want the light to be turned off if already on before this time?

[M.Harrison]

I have exactly the same issue with my landing lights - insufficient range at the bottom of the scale . 1% too low and 2% too high....
Was wondering about scaling (if possible) but not looked yet. The module is a BRS. Karam, if you do a firmware change then you already have a node profile for it from me for the pulse count issue.

[chris_j_hunter]

General Logic - analogue gates (eg: Multiplexer, Summation) include settable Gain, feed to a Threshold gate, and any other logic you might want, and you have what you need ... assuming the basic resolution is there, in the sensor output ...

[M.Harrison]

Many thanks Chris - Will certainly have a look at that

[M.Harrison]

Chris - Tried the multiplexer gate and it doesn't actually help... if the light level is 1% and you for example multiply by 10 on the multiplexer you get 10% as the output - when the level changes to 2% the output become 20% - there's no improvement in the granularity of the output which is what is needed here or, as I think Karam was meaning, a change to the shape of the sensitivity curve of the sensor.

[mcockerell]

Hello Karam, stretching the lower end of the light level curve is a better way of describing what I called 'granularity' - I've attached the node profile.

I'll have a think and try to define my other 'requirement' more clearly - we have outside lights that come on late in the evening, just before they will go off.

Thanks,
Martin

[Karam]

Hi Martin, Ok I see it is an LPS module despite the name of PTS . I'll get back to you when I've had a chance to try it. Also for the BRS as mentioned by M.Harrison

[Karam]

Hi Martin, firmware to stretch the lower 10% has been posted to you. Let us know if this helps at all.