Sunset Function on SENPILL and Light Level of 77 LUX goes 0 when Occupancy Enabled

I am trying to get the same functionality from a SENPILL (firmware 1.9) that I am getting from a SENPIRSS (firmware 1.2.67)

I have the SENPIRSS set as as follows. (one.jpg and two.jpg) I have a group address that tells me when the room is dark, and I have a group address that tells me when there is movement in the room (in last 5 minutes). These group addresses are then used in the pac. This all works very well.


I have the SENPILL set as follows. (three.jpg, four.jpg and five.jpg)
The SENPILL is mounted in the ceiling. The room is 5m x 5m there is 180Watts of light in the room. The room is bright when light is on. the SENPILL registers 77 LUX. Now if I shine a torch at the SENPILL at a 30 Degree angle from the roof It registers 100 LUX. If point the torch at 90 degrees from roof at the SENPILL then I get 1900 LUX.

As these units are mounted on the roof I think they will only work well as a LUX sensor if you have shiny floors !! As the only seem to be sensitive to light aimed straight at them from below (90 degrees from ceiling)


My first issue is if I enable the occupancy sensor the Blue LED goes on and the LUX level goes from 77 to 0 LUX. Does the LUX calculation get affected by the coloured LEDS ??? (can this be disabled)

The Occupancy feature works OK and works like the SENPIRSS

I can not get the Sunset Feature of the SENPILL to work like the SENPIRSS
I have set POT A to be light level adjustment and POT B the PIR sensitivity
no setting of POT A or light level will get the Sunset Group address to change.

How does the sunset function of the SENPILL work? Does it use the LUX measurements ?? (if this is so then enabling the Occupancy sensor and turning on the blue LED is going to make this a problem if 77LUX is full illumination)

Does the sunset function in the SENPILL have a delay ?? i.e. need to see room dark for 1 minute before it activates ? if so can this be disabled.

Have I just missed some thing ????

Charles

 

Here's some technical info, from a Lighting perspective. It's a long post but will help you understand what's happening...

The SENPILL device IS DEFNITELY operating correctly as a Lux Sensor, and the information relating to it reading higher (1900) when shining straight at it and reducing as you move away from the vertical is CORRECT... this correspondes to the standard physics and formulas for measuring lux levels.

Lux is "Lumens per Square Metre", and when light strikes something at an angle (ie: your 30 degrees from ceiling experiment) the light fans out over a larger area - therefore at the same distance and same candelas (light intensity) will produces far less lux at 60 degrees off vertical. The actual factor is the COS of the angle (60) Cubed... so Cos60 = 0.5, and that cubed equals 0.125... OK, you found 1900 and 100 at 60 degrees, which is a factor of 0.05, but please keep in mind that a C-Bus sensor is NOT an accurate Luxmeter (a half decent lux meter costs over $1000) !

So the problem is that Lux is the amount of light FALLING ONTO a surface. Since you are mounting a Lux Sensor in a ceiling pointing at the floor, it can only measure Reflected Light. In this case, it's actually operating more like a Luminance Meter (measuring the light intensity coming back FROM a surface) than a Illuminance Meter (measuring the light STRIKING a surface).

As you have observed, the difference is that Luminance (amount of light that's hitting the sensor in the ceiling) is directly proportional to the reflectance of the surface (floor or table) directly underneath it. Surfaces at angles away from straight down will reduce their contribution to the light onto the ceiling by that same formula (Cos of the angle cubed).
- And darkish wood is about 0.2 (20%), pure white paint maybe 0.8, cream colour maybe 0.6 reflectance...

So, the best you can do is not worry about the LUX and instead just set it to the threshold that works for your particular application. And note that the surface below MUST STAY CONSTANT... a dark table will reflect very little, but when you put white paper on it the sensor will "see" a higher light level and therefore might switch the lights OFF ! Same thing with humans walking below a sensor - think of white shirt vs black shirt !

The only way of making ANY light sensor work as a "luxmeter" is to mount it facing the light source(s) you want to measure, and not straight down (unless you really DO want to measure the light level hitting the ceiling).

By the way - this type of problem is PRECISELY the reason that perimeter dimming (sensing daylight and dimming lights near the windows to save power) never really works properly. The sensors are inevitably mounted in the ceiling facing downwards, so they do not get an accurate levels of daylight contribution, and thus the lights get dimmed in a "silly" way (ie: they don't respond how you'd want them to).

As I mentioned before, the only solution is to mount the sensor FACING the light source. If you want to measure the light striking a floor, then you need to mount the sensor in the floor ! Alternatively, mount it elsewhere and use a mirror or something to direct the "correct" light into it. You'd need to allow for the reduction in intensity (due to the loss from the reflection) but at least you'd be sensing what you need to know. I've seen successful daytlight harvesting jobs that use a small mirror to "scoop" the light coming from a window up into a sensor... so the sensor gets to detect a fairly accurate level of daylight contribution into the space.

Oh - and "shiny" doesn't mean "reflective" in this situation... shiny black doesn't reflect as much light as matt grey. The light reflected up to a ceiling is much more dependant on the colour than the shininess of the surface below it.


Sorry, that rave probably doesn't solve Charlie's problem, but hope it helps someone out there to understand how the sensors work

Cheers, John

 

Yes, the LL readings are adjusted to take out the effect of the LEDs. This cannot be disabled. This works accurately once the LL readings are over a couple hundred lux. Unfortunately, as you observe, when you're operating down towards the threshold of the unit's ability to measure LL it doesn't work very well because it's having a big impact on the measured LL reading.

Recently released to production was version 2 of the Multisensor. This update includes hardware changes to give the unit greater sensitivity at low light levels, finer-grained measurements (10 lux increments), the brightness of the LEDs can now be controlled and the issue of the LEDs affecting the LL reading has been dealt with.

There is no delay in the operation of the Sunset function and it should work the same way as in the 5753L. I suspect your problem is that the light level is not getting below the target that is set in the unit. Given that when the LL target/threshold is hooked up to a pot it can only reduce the value from that set in the UI, twiddling the pot can only make it even less likely for the sunset feature to work because the threshold is being lowered even further.

 

John,

Just to let you know - I never needed glasses until you starting posting on this forum as I never had to read more than 4 lines at any given time...

 

Is 11 lines too many, PSC ?

Thanks
for
your
comments,
Peter.
Pity
I
can't
explain
things
better !

 

Thanks John for taking the time. I understand LUX much better now..

 

So are you saying the Sunset feature uses the LUX readings ?? (ie not a seperate LDR)

If this is the case and having the 77 LUX going to 0 when the occupancy sensor is on. This means the LUX reading is 0 with lights on and off. Does this mean the sunset feature will never work.
I swapped the SENPILL for a SENPIRSS and there is more that enough light for the SENPIRSS to function as a Sunset switch at same location

Does this firmware update also give you the ability to disable the BLUE NIGHT LIGHT !

Is there now support for NEC remote codes like the new neo firmware ??

Are you talking about the LUX slider on the Light Level TAB ??

So the POT sets the level between the LUX slide level and 0





I removed the Sunet light level from pot A

I set the LUX target to 18

With torch on light level 116 LUX
with torch off light level is 0 LUX

the sunset feature still does not work

I noticed the "PIR Night move" works but the "PIR Day move" does not (even with 1000LUX shining at it !)

It is as if the sensor thinks it is dark all the time


how do you reset a SENPILL to factory defaults ?

Charles

 

Correct, no separate LDR. If the lux reading is 0 when the sensor is both On and Off then the sunset feature will never work. If the LL reading is always below the threshold set on the Light Level tab then the Sunset feature will never work. The sunset feature is activated when the measured light level goes from above to below the target set on the LL tab.

On the Global tab you can assign the LED brightness for all LEDs to be fixed at the current level, follow the block 1 level or follow the PE Cell level.

No, NEC IR remotes are not supported.

Yes.

Yes.

You probably need to try 37 lux or 55 lux as the target. This setting is influenced by the margin parameter as well and, in 1.9 firmware, there is +37 and -37 lux of hysteresis around the target to minimise the risk of switching oscillations around the operating point (even when the margin is set to 0). This has been reduced to only +10 and -10 lux in multisensor 2.

On the Occupancy tab there is a Reset button that sets all the occupancy functions back to their defaults. This also re-assigns the default key-block mapping for the first 3 blocks.

 

This is all hopless if you enable the PIR as even at 70 lux the reading is 0


with PIR disabled, Target set at 300 LUX (NO pot assigned), Margin 0
Light changing from 0 LUX to 1000 lux (torch light) still does not work
I know you are very busy but if you can get this working on the bench on your version 1.9 SENPILL can you please post screen shots of the config.



I have 4 of these sensors in each bed rooms. I tried one of the others and it does not work either. The other rooms are worse only registering 18 LUX when all lights are on.

I know fixing this will not solve my problem but I would like to see it work


At 18 LUX (version 1.9) NO PIR. Is version 2.0 going to work at this level with PIR??


a SENPIRSS at this same location works fine even when light dimmed to 30%!

Is version 2.0 smart enough if you dim the LEDS that is acounts for this in the LUX calculation.

Thanks Newman

Charles

 

I don't have a 1.9 unit handy but attached is the setup for a 2.0 unit which works just fine. Make sure your bank switching and Environment tabs are all clear as this might be mucking things up.

Yes, the lux reading should be independent of the LED settings. It's a significant improvement over 1.9.

 

Part 2...

 

Newman,
Thanks for taking the time to do that....


I setup my SENPILL EXACTLY the same as yours. Hey Presto all works....

Hummmm Confused me.....

I then changed the setup to what I hand before (ONLY 3 screens Three.jpg, Four.jpg and Five.jpg as per the first post in the thread)....... Hey Presto it worked also....

Hummm.... I thaught ... whats going on here !


Then it came to me... .

I have had the PIR disabled (because of the LUX LED interaction issue)

You had it enabled and I only changed 3 screens but did not disable the Occupancy sensor.

I then disabled the ocupancy sensor... Hey presto the sunset function stops working....

I then changed the config back to exactly as your example...
Sunset worked Fine.

I then changed on Occupancy from "Enables" to "Disables"
Sunset nolonger works..... !

Now this may only be a bug in version 1.9 you may want to test in 2.0



You missed one question on my last post

Sounds like it will, now that the sensitivity is UP, LEDS can be turned off, hysteresis now +-10

Charles

 

There's 2 things going on here. 1.9 and 2.0 are the same in this regard:

1. When the PIR is enabled/disabled then the Sunset feature gets enabled/disabled at the same time. I don't know if this was intentional or not but it's the way they work.
2. Normally when you select the 'enables' radio buttons on the Occupancy tab this means that the PIR is enabled when the group is on, disabled when the group is off, vice versa when you select the 'disables' radio button. When the Group is set to unused the radio button selection just enables/disables the PIR outright (including Sunset).

Can't say absolutely but I'd be reasonably confident. I'm really keen to know how it goes so keep us posted.

 

Well intentional or not I think this is a bug ! better get that fixed in verion 2.1

It is a PIR enable/disable not Sunset Enable/Disable

Does the SENPIRSS sunset function disable when you disable the PIR. No it does not !! You need uniform functionality between products.


I can think of a few times when you would want the Sunset function and not the PIR.


The 4 x SENPILL's 1.9 in my bed rooms have now become expensive 5031NIRL's with a blue night nights