12 Volt lighting

[aidan.f]

I have voluntered to design and install battery powered lighting in a log cabin located near Moffat in Southern Scotland
The building is 10M x 7 and principally used as a meeting room for forest skool activities.

I'm keen to support the organisation who own it and won't be all give, they are O.K. about camping,

There is a plan for a stove install... cosy..maybe some interest from forum members?

I will specify and cost options then report back for review and approval.

Adrian (my customer rep.) would like both task and mood lighting.

Once I know how many Watts and  planned use I  can then determine battery/charger specification.

For task lights 120 LED lights such as these:

https://tinyurl.com/52d7cdyh would fit under the roof trusses, unless anyone can suggest  better.

For the 'mood' lighting I'm thinking about red(er) lighting to maintain night vision in a dark-sky area.

Getting something similar to above is spendy.

https://www.12voltplanet.co.uk/12v-led-interior-strip-light-with-red-leds-150mm.html

- I could do with some  ideas  - perhaps with a bit of interior style such as pendant fittings or lanterns?

Just to expain more - my outline plan is light control using IP65 switches in a wall mount box.

Power - solar has been requested  - with trees to South panels would have to be remote. 

I have just realised, writing this, that planning consent maybe required for ground mount panel(s)

IMO portable power would be lower cost, but more user bother for the predicted 1-2 hours use a week.

Below is a view of the site; look foward to some YACF knowledge sharing.

 

[quixoticgeek]

2 hours if use a week makes for very easy solar quantities.

How much do the trees shield the roof? If you our a panel on the top of the roof near the ridge would they be shaded by the trees? Do you need year round use, or just spring through autumn?

If you do need winter use, spec the panels based on the winter sun levels, and then accept that you're massively overkil in the summer.

If you can give me a lat and long (to one decimal place is enough), I can do the maths for you when I get back to my laptop tomorrow evening.

J

[fruitcake]

Very interested to see this thread. 12V makes a lot of sense off-grid, due to availability of car batteries / leisure batteries. When I considered something similar, I realised that thought needs to be given to the cable rating, since lower voltage means more current, and thicker cables, in order to reduce fire risk. A few more thoughts in no particular order:

I think we may see homes lit via battery circuits, because those lights run when the mains fails. So the industry may catch up with you!

PIR actuated lights may make sense in areas where you don't need lights on for more than a few minutes - which can do away with the need for lightswitches, which would simplify installation there.

Lighting tech is getting to the stage where lighting won't have to be fully installed, I'd suggest. A powerbank can power an LED (desk) lamp for a number of hours, which may suit some of the task lighting. A few powerbanks on charge from a solar panel could be enough to start with. The kind of self-contained PIR battery lamps that are sold as nightlights are good for corridors, vestibules and stairs. 800 lumens can be enough to light a room and is achievable with current LED lamps. The lamps sold as security lights with connections for batteries exceed this.

[Canardly]

Is emergency lighting an issue to be considered?

[quixoticgeek]

PIR actuated lights may make sense in areas where you don't need lights on for more than a few minutes - which can do away with the need for lightswitches, which would simplify installation there.

The load of the PIR device for 166/168 hours a week will be more than the lights for the other 2/168.

J

[perpetual dan]

What’s the interior like? Are there white walls / ceilings to bounce light off? Tables in the middle or everything movable? Anything that behaves like a spotlight needs to not point through someone’s head on it’s way to the work to be illuminated. Reflecting off walls can make a bigger, softer lighting if that’s possible.

When’s that photo taken? It looks like the roof gets some sun, so the right panel / battery combination might be easier than a panel in the field (which I’d guess they want to use for outdoorsy things) and burying the wire.

It might be worth having a timer to ensure lights go off. Especially if the person locking the door as they leave can’t see at a glance if everything is off.

For mood light, I’d be looking at a long string of warm white (or red, good idea) fairy lights. But maybe that’s just me

[lissotriton]

Worth having a switch for the whole circuit, so can make sure it is all off when you leave?

Can get multicolour LED strips, so can set them for whatever mood you like.

[chrisbainbridge]

would it be worth lurking in some camper van forums?  This is exactly what I would be speccing for a van if I got permission.  Solar panel with modern 12V Batteries and LED. 

Could you consider wind power as well.  I don't know enough but I understand that you can get some small turbines which will provide at least some charge.

[quixoticgeek]

would it be worth lurking in some camper van forums?  This is exactly what I would be speccing for a van if I got permission.  Solar panel with modern 12V Batteries and LED. 

Could you consider wind power as well.  I don't know enough but I understand that you can get some small turbines which will provide at least some charge.

Wind isn't cost effective at this scale. For the same money you can just throw another couple of panels of solar at it...

J

[grams]

Campervannists get most of their electrical power by running the engine or using external hook ups. They spec 100 Ah batteries to charge their phone and run the sink pump for 5 seconds. Solar panels are mostly to stop it going completely dead when unused. A lot of the professional kit providers are as bad as the amateur bodgers.

There’s probably a pocket of off-grinders who’ve done real maths and measurements but you’re going to wade through a lot of bullshit and fairy stories to find them.

[aidan.f]

Lots of good replies, I knew you would be interested!

2 hours if use a week makes for very easy solar quantities.

I have a suspicion that may increase.. nice cosy hut with wood burner..

If you can give me a lat and long (to one decimal place is enough), I can do the maths for you when I get back to my laptop tomorrow evening.

I'm of to do a detailed site survey tomorrow  - I'm going to plot horizon angle from various locations then use this
https://re.jrc.ec.europa.eu/pvg_tools/en/tools.html

What’s the interior like? Are there white walls / ceilings to bounce light off? Reflecting off walls can make a bigger, softer lighting if that’s possible.

I'm thinking along that way as well - the roof is boarded - not highly reflective but uplighting on the trusses would  look nice and conceal the van luminares which are a bit 'harsh'  I could always suggest some white panels to bounce back light.

It might be worth having a timer/pir to ensure lights go off. Especially if the person locking the door as they leave can’t see at a glance if everything is off.

The load of the PIR device for 166/168 hours a week will be more than the lights for the other 2/168.

There are battery PIR lights about, so microamp load, but I cannot find a 12V in - load out PIR switch.
I did modify some cheap AC PIRS for  12 volts but as QG sez the LM324 and bits inside take about 6 mA of quiescent..

Next  - I am aquiring a two Tilley lanterns - clean them up convert to LED and hang them as feature lights.

[nikki]

Sounds like a nice project!

I've been lurking on van conversion internet for a year or two now.
From that perspective, but with limited practical experience at this point, I'd echo

Anything that behaves like a spotlight needs to not point through someone’s head on its way to the work to be illuminated.

and suggest led strings over puck lights for task lighting. I'm not sure where the ones you linked sit on this spectrum.

Could you use RGB LED strings for both white, task, lighting and redder, dimmer, ambient/night lighting? I think the latter is going to be very subjective depending on person and context, so all else being equal would an adjustable approach be good? You'd have to bolt any remote control to the wall!


Questions:
Do dimmed LED strips allow for low power drain when the battery is getting low?
Does the install include battery monitor so folks can know when to ration their usage if needed?
Does the install benefit from an MPPT module to optimise battery charging?


If you've not found it already, 12voltplanet has a voltage drop calculator in case that's useful when thinking about cable length and thickness: https://www.12voltplanet.co.uk/voltage-drop-calculator.html

[quixoticgeek]

Could you use RGB LED strings for both white, task, lighting and redder, dimmer, ambient/night lighting? I think the latter is going to be very subjective depending on person and context, so all else being equal would an adjustable approach be good? You'd have to bolt any remote control to the wall!

No. While an RGB LED can give the impression of white, as a work light, it doesn't. You can get RGBW Led strips, where the RGB is augmented with an additional white led. This means you can have actual white for working, but can then add the colours for when you want mood lighting.

The problem is the CRI of the white light on the RGBW LED's is... poor. Having said CRI, Kim will be along shortly to explain it all way better than I can, and recommend light fittings.

I have a LED work light that I made from a 330mm aluminium cake tin, some 97+ CRI White LEDs, a 20khz pwm controller, and a 12v brick. It gives a really good quality light for working by. All parts from Ali's finest Chinese express emporium.

J

[SoreTween]

I have just realised, writing this, that planning consent maybe required for ground mount panel(s)

In England non-domestic permitted development allow for one ground mount installation of up to 9m².  That effectively gives you 4 panels and should deliver roughly 1600W peak.  However, you may struggle to fit all that within 3m x 3m and less than 4m high and if you did 3 portrait with one landscape above it would be a complex frame to install.  You might be better to plan on 3 panels (1200W) which would be a pretty straight forward installation.

I can't find a simple equivalent explanation for Scotland but did find there was a consultation on expanding the rights including for non-domestic back in 2015.  The baseline must be buried in The Town and Country Planning (General Permitted Development) (Scotland) Order 1992 to which there are updates, I'll dig some more when I have time.

ETA:

Found it I think: Permitted Development Rights: Non-Domestic Solar Panels and Domestic Air Source Heat Pumps
It looks like you can go nuts up to 50kW with the restrictions mostly being about visual impact so long as you're not within 3km of an airfield boundary or in a national park or other sensitive location.  There are other caveats under Current Situation on this page.

Obviously you'll have to take proper advice but it looks hopeful you won't need planning permission.  Plonk some pre-cast concrete blocks down & build a frame off that.

[aidan.f]

More and more useful advice. I'm going to keep the lights simple, if red night vision lights are needed use separate luminares.  Also nice big toggle switches that can easily be operated whilst wearing gloves.

Does the install include battery monitor so folks can know when to ration their usage if needed?
Does the install benefit from an MPPT module to optimise battery charging?

Couple of good questions. I am leaning towards a power pack as a 'product' that comes with battery monitoring and may be removed for AC charging. It's also the cheapest way to buy Watt/hours. I have read that MPTT is not the be-all. At low light levels PWM works better and PWM controllers have a much lower burden current 5-6 rather than 16-20 mA. Cannot find a charger that incorporates a LiFePO4 battery level monitor, not surprised given flat discharge.

[nikki]

I am leaning towards a power pack as a 'product' that comes with battery monitoring and may be removed for AC charging.

 
I use an EcoFLow Delta 2, powering a range of 12v components via the cigarette charger output and a fuse panel if that's a useful reference point for questions/tests further down the line.

[sojournermike]

I’m about to do a solar, battery inverter install for my shed and office so this is interesting.

However, to contribute an additional thought, I’d shy away from 6000k for a task light and look for something like 4,000k. Also, as noted above, CRI matters. Ideally 90+ and certainly not below 85 if you want your ‘subjects’ to have a pleasant experience.

Mood lighting - dim and warm and indirect would be my call.

Kim is the expert on this and on flicker, although you may minimise that on 12v DC - depending on power control circuitry of course.

[Diver300]

Kim is the expert on this and on flicker, although you may minimise that on 12v DC - depending on power control circuitry of course.

Dimming using PWM (pulse width modulation) is common. It is often designed under the misunderstanding that human vision has an upper frequency cut off of around 50 Hz. In reality, whether flicker can be detected depends on a lot of things, such as contrast, depth of flicker, duty cycle of flicker, who is looking at it, how tired they are and what part of the field of view the light is in.

I've detected a flashing light at 5 kHz.

I'm so old that I can't hear high frequencies. I recently made a power supply that squeaked quite loudly at 11 kHz, and I had to go back and fix it. I didn't just say that I can't hear it so no-one can, but that seems to be what happens with lighting flicker.

[Kim]

Kim is the expert on this and on flicker, although you may minimise that on 12v DC - depending on power control circuitry of course.

Dimming using PWM (pulse width modulation) is common. It is often designed under the misunderstanding that human vision has an upper frequency cut off of around 50 Hz. In reality, whether flicker can be detected depends on a lot of things, such as contrast, depth of flicker, duty cycle of flicker, who is looking at it, how tired they are and what part of the field of view the light is in.

You forgot 'objects in motion', which can be the light source itself (eg. when it's mounted to a vehicle) or when it's used to illuminate stuff that moves.  Indeed, waving a pen or ruler or similar back and forth and looking for distinct images rather than the expected motion blur is a good test for flickering lighting if you can't see it directly.

It's not uncommon for machine tools to be illuminated by a small tungsten lamp, to avoid the danger of appearing stationary under flickering room lighting.

 

I've detected a flashing light at 5 kHz.

I'm so old that I can't hear high frequencies. I recently made a power supply that squeaked quite loudly at 11 kHz, and I had to go back and fix it. I didn't just say that I can't hear it so no-one can, but that seems to be what happens with lighting flicker.

...which is why I reckon we should be aiming for above auditory range for lighting PWM.  You need maybe 10kHz to eliminate flicker for everyone, so you might as well get it above 20kHz so there isn't audible whine from the electronics, either.  Or do current dimming, but that can reduce efficiency and muck up the CRI of LEDs at low output.


Another thing I've recently encountered, in the context of RGB(W)(WW) LED drivers, is a PWM strategy where the channels are activated in sequence, rather than in phase with each other, in order to reduce the peak current draw.  This seems like a recipe for VHS-style colour artefacts on moving objects.

[sojournermike]

Kim is the expert on this and on flicker, although you may minimise that on 12v DC - depending on power control circuitry of course.

Dimming using PWM (pulse width modulation) is common. It is often designed under the misunderstanding that human vision has an upper frequency cut off of around 50 Hz. In reality, whether flicker can be detected depends on a lot of things, such as contrast, depth of flicker, duty cycle of flicker, who is looking at it, how tired they are and what part of the field of view the light is in.

You forgot 'objects in motion', which can be the light source itself (eg. when it's mounted to a vehicle) or when it's used to illuminate stuff that moves.  Indeed, waving a pen or ruler or similar back and forth and looking for distinct images rather than the expected motion blur is a good test for flickering lighting if you can't see it directly.

It's not uncommon for machine tools to be illuminated by a small tungsten lamp, to avoid the danger of appearing stationary under flickering room lighting.

 

I've detected a flashing light at 5 kHz.

I'm so old that I can't hear high frequencies. I recently made a power supply that squeaked quite loudly at 11 kHz, and I had to go back and fix it. I didn't just say that I can't hear it so no-one can, but that seems to be what happens with lighting flicker.

...which is why I reckon we should be aiming for above auditory range for lighting PWM.  You need maybe 10kHz to eliminate flicker for everyone, so you might as well get it above 20kHz so there isn't audible whine from the electronics, either.  Or do current dimming, but that can reduce efficiency and muck up the CRI of LEDs at low output.


Another thing I've recently encountered, in the context of RGB(W)(WW) LED drivers, is a PWM strategy where the channels are activated in sequence, rather than in phase with each other, in order to reduce the peak current draw.  This seems like a recipe for VHS-style colour artefacts on moving objects.

A great example of addressing one problem and creating another. Whackamole
…

[sojournermike]

Kim is the expert on this and on flicker, although you may minimise that on 12v DC - depending on power control circuitry of course.

Dimming using PWM (pulse width modulation) is common. It is often designed under the misunderstanding that human vision has an upper frequency cut off of around 50 Hz. In reality, whether flicker can be detected depends on a lot of things, such as contrast, depth of flicker, duty cycle of flicker, who is looking at it, how tired they are and what part of the field of view the light is in.

I've detected a flashing light at 5 kHz.

I'm so old that I can't hear high frequencies. I recently made a power supply that squeaked quite loudly at 11 kHz, and I had to go back and fix it. I didn't just say that I can't hear it so no-one can, but that seems to be what happens with lighting flicker.

I had the slightly humbling and very enlightening experience of switching on an audio amplifier I’d built and, as I was thinking that something wasn’t quite right but couldn’t put my finger on it, my (musician) daughter ran in clutching her ears and asking me to turn it off. Don’t know where that landed on the spectrum but a very clear demonstration of ageing:(

Otoh, I hear things Sue doesn’t, so shouldn’t complain.

[Diver300]

Or do current dimming, but that can reduce efficiency and muck up the CRI of LEDs at low output.

Most LEDs are slightly more efficient at lower currents, although that may not hold for tiny current less than 5% or so of the rated current. Obviously, the driver circuits aren't the same efficiency at all power so that will come into it as well.
https://assets.cree-led.com/a/ds/j/JSeries-3030-HE.pdf graphs on page 6
The same data sheet shows only very tiny changes in chromaticity with current.

I think that the idea that efficiency and chromaticity change a lot with supply current comes from the lack of details in data sheets. Cree are rare in quoting typical light outputs and CRI at various currents. If the efficiency and chromaticity are not specified, it can be argued that next year's batch could be completely different at low currents, so the designers play safe by running at the rated current whatever, giving worse flicker and worse life.

[Kim]

That makes sense.

IMHO, CRI's pretty unimportant when you're dimming to such low values anyway.  At that point you tend to be mostly interested in having enough light to not trip over things while keeping things nice and dim before bedtime, or conserving battery life, rather than accurate rendition of colour.  Plus high colour-temperature sources look weird when dimmed anyway.

[aidan.f]

Well I now have a modified Tilley lantern...



Unlike most electric conversions top down, could be reversed as there are no holes drilled in the pressure tank

I'm looking for a mantle shaped diffuser - currently 35 mm film container  - maybe my first 3D printing job?

LED assy : https://bedazzledledlighting.co.uk/product/g4-tower-18-led-small/

[fruitcake]

That's a cute lantern conversion. I can imagine there being a market for LED-retrofit paraffin lamps. An Etsy shop perhaps.