Hi Paskal,

Thanks for sharing your journey.

I think building on stilts is another idea given the reasons below:
a. coastline - wind or storm surge will bring in the water esp if there's a river nearby.
b. Ventilation - easily allows for wind to come in.

The car park area could be either heightened with an elevated bay or their structure augmented with storm doors if build at ground level.

Building materials like AAC help slow down heat build up on from the exterior. Natural lighting via clerestory windows as shown in some of your sketches and as some mentioned taking advantage of prevailing winds to naturally ventilate the house.

This post has been edited by halcyon27: Jun 10 2014, 09:03 PM

it is, but i still don't have any good method of measuring the light output. haven't researched how manufacturers carry out their measurement to come up with their lumens per watt rating.

i do have access to a lux meter, but need to build a makeshift reflector to channel the light and avoid diffraction. so for now it's just relative comparison when the room lights are turned off and the sample lamps turned upwards to light the ceiling.

nice to see another engineer. the led lights are freakin simple to construct. can even build ourselves DIY style.

Hope this info will speed up some of your research:
http://en.wikipedia.org/wiki/Luminous_efficacy#Examples_2

Have thought of return of investment? Price of LED with power saving(TNB Bill) vs CFL with much lower capital(Half as much if im not wrong [With Fixtures, etc]) while both having the same-ish LPW (Lumen per watt).

Personally, i have did a rough calculation on this topic, it turns out to be CFL favour...ROI was just too long and the warranty are usual 1 year or 2.

But i'm eager to read your thoughts and findings on these.

BTW, have you thought of have a 2nd floor or a ventilated roof system to isolate your first floor from direct sunlight heat ray, to saves A/C bills =P . Another topic im interested in.

It is good to know the actual running watt. Sometime really difficult to believe what they claim. Specially for the LED.

Is this the actual watt for the led light 18w? Or their state spec? What is the lumens ? How is the Amp measure before and after the power supply?

Agree, verify their "claims" first. We can crunch the numbers later =P

i've read that them Autoclaved aerated concrete-AAC have some sort of health issue? got a thread before this discussing the aac building material.
searched the net and found there's a sort of lawsuit regarding aac. no idea if the same issue is apparent in malaysia since i've seen a couple of building in my area built using something that resembles an aac block.

there's just the master bedroom wall that's gonna heat up from the evening sun. thinking of adding a sunshading wall later if the heat buildup is too much. quite sceptical of adopting aac after reading about the issues.

it serves as a rough guideline, but the actual light efficacy have to be measured since not all LED have the same efficacy.
manufacturing process, material used, the housing, etc etc all have a direct effect to the light output. i got a friend that's from the industry, worked till senior engineer post at a japanese company manufacturing led solutions before resigning. detailed explanation from him till get earbleed.

anyway the ROI is just too long. kindda not worth to implement the entire house with LED. at some parts of the house it's more logical and cheaper to use cfl.

roof insulation is integrated. i'm gonna use monier's bubble wrap roof insulation system just after the roof tile, with an air gap of 1-2 inch. that should provide the best possible insulation for the roof.

i've came across a study on roof insulation and bubble wrap with an air gap is indeed the best possible way to get the highest heat attenuation. monier have another double layer bubble wrap called super r or something like that. no idea if it's available locally.
the current quotation is for the monier roof with bubble wrap insulation and air gap.

must consider roof insulation since electricity costs isn't coming down. after all, the difference isn't by much. i'm even considering of adding another bubble wrap layer (or a few more layer) laid on top of the plaster ceiling to further attenuate heat transfer from above. only problem is the FIRE HAZARD haha.

to improve ventilation inside the roof structure, i'm gonna use these ceiling panels for the outside:



i actually have this:

so i visit friend's house, other peoples house, house in construction and measured their ceiling temperature. haha
some using high roof, low roof, metal deck, metal deck with insulation, asbestos ceiling, plaster ceiling, low pitch, high pitch, no ceiling, roof tile, etc etc.
with this little device it's all scientific and require no guesstimation

it's quite late. till next time.

In Saudi Arab, they have this kind of aerated clay brick. Very effective.
Even at mid noon can sleep in hotel room on high floor with just fan on.

But not sure if can get locally. Or if local sifus know how to build with this. Much bigger size than our standard brick.

Informative and interesting, thanks guys ...

Maybe you have to create a vent hole up for the hot air to escape. The bubble wrap seem like wrap the whole top roof without giving any hot air to vent out. The ceiling panel is good for cool air in.

For the ceiling, it is best to use thick rockwool or Cellulose. This are fire retardant material. Any thin material doesn't help.

How about this ? Its a bit ugly tho.

this is what i measured for the oxyone 18w led.
this is the product:

sorry i thought it's rated for 18w power, but the spec says it's 16w.


252.2 VAC = 178.36 Vrms * 0.1107 = 19.74W

actual DC power sent to the LEDs after conversion:

26V * 0.4890 A = 12.71W
efficiency of the DC converter = 12.71/19.74 = 64.38%
meaning it's freaking low. nearly 20W ac power sent, only 12.7w dc are received by the LED.

comparing the light output versus the 18w CFL from phillips and panasonic, i say these LEDs are on average ~15% brighter compared to the CFL.
is it worth it? probably not, considering it's gonna be a hassle to maintain them if should the LED driver malfunction after a few years. as with any electronics, it's not gonna last forever.

the oxyone leds says on the box that they're rated for 120lm/w efficacy. verry hard to believe since on average the actual led chips have 100lm/w efficacy at best. on DC power.

say you get an 18W led. that 18w is rated on AC.
18w * 65% efficient dc conversion = 11.7w of dc power. so at best you're gonna have 1170 lumens from it.

remember this small led lights?
supposed to be 9w (3 led) and 12w (4 led), epistar chip.


actual measurements.
9w:

12w:

which makes the efficiency of the 9w at 42.5% and the 12w efficiency at 46.5%
not the led efficacy mind you, but the DC conversion efficiency.

it's freaking low. waste of electricity.
even if the epistar is capable of 100lm/w, the dc conversion efficiency pulled it down. say you get 300 lumen from the 6.56w ac input, the lumen per watt is only 45.73 lm/w.
for the 4 led chip, say you get 400 lumen from 8.64w ac input, the actual output would only be 46.3 lumen/w

lower than even the cheap 11w phillips CFL that i have which is rated for 54 lumen/watt.

install LED to save energy huh?

you know how to calculate aircon consumption?

Thanks for your testing. Your testing do have similar patent to what I have test 2yrs ago.

I run a simple test on the famous CREE ecosmart led. https://forum.lowyat.net/topic/1481183/+560

Than lux check last yrs. https://forum.lowyat.net/topic/2872552/+20

I give 2yrs of testing for reliable. Time have pass fast. The LED still run without any problem. But haven't check the Lux reading got any output reduce. Will conduct the test when have the time. I have more confident on this product now. And probably looking at how to import more and modify it to be efficient running.

you could measure the current draw using a clamp meter or directly tap into the power line.
or use a wattmeter, which is in a way better because it provides the power factor reading.

power factor reading is needed to calculate the actual power usage, but some older tnb meters doesn't consider the power factor in calculating the power usage.
as far as i know, the newer tnb digital meter charges you in actual power usage by calculating the power factor while the older analog meter only calculate VA rating.

to improve the efficiency, we could bypass the led driver like what you did and run them purely on DC using a high efficiency SMPS power supply. that should pump the conversion efficiency to over 80%. downside is there's no current limit and we're limiting the light output that we could extract from the led using this method.

or replace the led driver with better drivers from mean well. they have some pretty high quality drivers which is rated for upwards of 80% efficiency. from experience working with their SMPS, their products are pretty high quality rivalling those expensive unit. and their rated spec is pretty much spot on without any bloated artificial number.
i've opened quite a number of (expensive) equipment and not surprised to see the power supply from mean well in them. downside is that the choices is pretty limited. there's no peculiar current limit like 600mA, 125mA, 480mA and the such.

From the testing and the actual reading, pretty much that the LED can save energy. Provided the power supply(driver) is efficient. I run this test everyday for 8hr+ for 2yrs. And with a good driver, it can sum out,

1) 8hr x 2yrs = 6000hr average to date. That is already over the CFL rating at 8000hr with 4hr running everyday.
2) With the 6W consume, it only consume 1.5kwh/mth. For 2yrs, it only consume 35kwh or RM8 bill.

But ROI is long. Probably take about 5-6yrs equal to CFL ROI. Provided the LED doesn't blow up and TNB rate keep increase.

If the whole house or certain area is light with LED, what is the best setup? My initial plan is to have 1 driver for several led. Takeout the crap driver and run direct from the main driver. A 100W driver probably can drive 12pcs of led (6w). That should give,

1) max efficient.
2) reduce the break down,
3) less heat
4) cheaper cost.

I know this mean well brand but never use before. Selling a lot in local electrical shop. But mostly I m using Omron, TDK, idec or nemic lamda. Given Omron is my favour. Max average Efficiency is 86% while you can go for 3P which can drive till 91% efficiency.

read below your posts and finally understood when you typed 6w. you corroborate my findings. the AC-DC converter that the LED lights are using have pretty low efficiency. surprising really, because yours are from CREE themselves.
ROI using the default led driver is just too long for me. i never believe electronics will last more than a couple of years, especially switching circuits. 5-6 years is just too long.

tho LEDs have a few important advantages that CFL could never provide. most important for me is:
1) heat dissipation to the backside of the case, which is to the ceiling.
2) can be powered by DC power using backup battery in case power failure.

best setup IMO is to isolate the DC conversion and current limit (led driver). use a high efficiency SMPS to convert to DC, then use a separate current limiter to drive the LED. you could use multiple LED per driver but there's a few downside.
upside:
- should provide higher efficiency as the current limiter is rated for ~97% efficiency, SMPS rating above 85%.
- cheaper maintenance cost since the smps and current limiter is isolated. replace just the problematic item.
- easily retrofitted to run from battery power for backup lights.
downside:
- switches will be handling DC. arching could destroy the switches in the long term.
- multiple equipment to maintain.

another easier setup is as you explained. using multiple LEDs wired in series to a single driver.
upside:
- cheaper initial cost. cheaper to modify existing wiring.
- easier to maintain. driver could be placed in easy to access area and since the LED itself are hard to spoil it's shorter downtime.
downside:
- higher replacement cost. single high power driver will cost more compared to either smps or current limiter.
- most led driver (integrated driver with smps and current limiter) have higher efficiency if nearly fully loaded. so it's gonna be hard to maintain high efficiency throughout the entire house.

have you considered building your own LED downlights?
i can't seem to find any supplier in ebay or aliexpress selling just the LED downlight case/housing. epistar 1w and 0.5w SMD LEDs are selling quite cheaply on the net. it's gonna be easy to hit 100 lumen/DC watt (or more) if we build the LED ourselves.
pair it with some cheap, high efficiency driver and string a couple of them together, it's gonna hit 90 lumens/watt or more on AC.

anyway for my garden lights, wall lights and flood lights, i'm gonna feed them DC directly (with or without current limiter) and build my own LED string. i've already bought the cases for the garden and wall lights so it's do or die.

mind to explain what is 3P? 91% efficiency is quite high.

main inspiration for the outdoor deck:

integrated steps and sitting area. tiled.
integrated planter boxes, auto irrigation system.
plumbing works will be designed to support such system, gravity fed from elevated rainwater collection tank.

inspiration for the partly shaded deck:


inspiration for the outdoor courtyard, with outdoor projection:


why yes. the wiring and plumbing is gonna be complicated.
main reason why it has to be done in DIY fashion.

For the fun and mother nature is really good to use the method you mention. But for pocket and practicality is another story. I doubt it will generate any good ROI.

Have you consider Solar -> Battery and Let the battery power the LED in series. This way you have minimal wastage in conversion, and "free" energy from solar. Than again, you have to spend money investing on Battery. I been researching lithium ion battery after Xiaomi selling a good and cheap Powerbank. It actually don't cost much to buy the similar battery Xiaomi using LGABB41865 (universal 18650), maybe RM 5 / unit for low quantity order.
If you can get the same or higher mAh batter capacity like Xiaomi powerbank using (2600mAh), you get 9.6watt hour per battery. That enough to power one 9watt Led for 45mins (assuming 75% efficiency).

Long I have not work on the power supply, I maybe wrong but I think the 97% efficiency can't be achieve in real world environment with the heat, and component quality.