natural ventilation is indeed considered. together with sun orientation.
sun movement and natural lighting for the proposed design is even simulated in google earth and sketchup to make sure i don't get too much sunlight and heat up the house.

any good idea on how to improve the ventilation? i've posted the wind direction and sun movement before.

it is quite small. the HT room is the single biggest room in the house.
for the current funding situation, must make do. can move the master bedroom to another part of the house if should i proceed to extend the house. refer extension plan below.


something like this. not in any way final.
but i could add another ~1000 sqft to the house for another RM80-100k. additional 4 more bedrooms and an indoor garden.
in 10 years time, maybe. if necessary.

bought some LED lights and your normal CFL lights to check for their efficiency and light output.
i don't trust manufacturer specs, especially those from shady brands like these oxyone lights. so bought the ones that my wife really like to do an actual measurement.
the oxyone leds are rated for 18w, compared against phillips 18w, phillips 11w and panasonic 18w.

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.

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?

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.

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.

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.

you mean this right?

i've bought some for sampling. another project for another day. a viable source of energy pack for embedded systems.
yes i agree they're worth more than what xiaomi is asking for them. so good that i've bought more than i need and given as gifts to family members.

have considered solar charging. but the costs are not justifiable. as least for me.
got a friend using them tho, with a 1kw panel powering his led downlights.

he's using your normal lead acid battery. a car battery to be exact. and i'm with him on this. lithium ion cells are expensive and require extensive protection circuit. even charging circuit is quite complicated. the advantage is only for size and energy density where lithium cells stores more energy for the size.

it's better to use lead acid in this case as we have no size and weight constraints. it's reliable, have no problem supplying huge amount of current, robust, no problem with heat, less risk, doesn't require complex monitoring circuit, able to work with unstable charging voltage, and it's cheap. and it doesn't have memory effect.
having previously worked with NiCad & NiMH cell, LiOn and lead acid i really think lead acid is the best for this situation.

the reason why i won't implement it is because of the adoption costs.
it should cost about RM10k for a 1kW panel.
and if my memory serves me right, that 1kw panel should give you around 3kW per day average. 4kw at most. i'm realistic so we're gonna go with 3kw.

N70 battery at RM250 a piece. 12v x 75A = 900W per battery. 3 years working life.
say you're using all 3kW to power all the lights inside the house. realistic as it's gonna cost RM0.90 per night, ~RM30/month to power just the lights.

adoption cost at RM10k (panel) + RM750 for the battery + ~RM250 for the charging circuits, wiring, etc etc.
total RM11k, for a saving of RM30 a month for the lights.
that RM11k / RM30 per month saving = 366 month just to break even.

that's 30 years. just to break even.
not worth it IMO.

Haha, as to date, I brought 10 units of Xiaomi power bank

Haha, add another 2 more units, just bought today

I referring battery inside,, is 3.6v or was it 3.7v exactly good for led light.

Yeah, and is cheap if buy in bulk. Saw in Alibaba can go down to less than 1USD each. Imagine buy 200 units

The killing is the panel and inverter.

Lets use Xiaomi Powerbank as a base to calculate, is selling at RM 36 with circuit + 4 battery 2600mah inside. If i take everything as zero cost except the battery, that make it RM 9 per battery. I don't think Xiaomi selling at lost right? Think they maybe getting it at RM 3-4 per unit.

Think can get lower, need to buy more lo. Think we are oot already

to make battery to last, need battery optimizer like hybrid (nvr go below 30% & charge up to 70% nvr full) & automated process top-up dehydrated lead acid

with 10000mAh, usable is 3000-4000mAh

base on NiMH can last abt 18months- charge full n drain 10% (battery power flashlight)

maybe consider this Iron Edison

[attachmentid=4028629]

nice to see so much technical discussion going on. sorry i can't add much since class is just starting.
i don't mind you guys OOT and continue your discussion.

wife provides the architectural insight, and i provide the engineering standpoint.
i can understand why architect and engineer fight all the time.

been designing things and building them for a few years now, made that much easier with the advent of sketchup.
it's much easier to use compared to autocad (for 3d design/editing in 3d space), much faster result, less hassle, while retaining much of the precision that you get from a cad software.

some of my past projects designed and fully built with the help of sketchup:

all built with less than 1mm error. the precision is there, provided you could accurately follow the design.
project is documented over in lowyat's home entertainment section and hifi4sale's DIY section.

i freakin love sketchup. because i freckin hate autocad.

it's not space partition, but internal bracing for the outside walls. it's a subwoofer, a high power at that.
the internal pressure at high output causes the outside walls to flex. so internal bracing is designed to provide more support than is needed, to avoid the flexing issue.

the internal space can't be partitioned as it'll reduce the effective internal area. large opening for the bracing is needed to avoid wind noise when the sub is pressurized.
thus the maze look.

the house layout is really straight forward. there's not even a pathway or walkway inside the house.
there's no change to the floorplan when we erected the 3D.

we purposely went with a straightforward design to maximize space.

no idea if it's gonna be dark. but there's gonna be a large glass opening at the kitchen wall to aid lighting. wife did say that the kitchen area looks like it won't receive much natural lighting as it is. the dry kitchen needs a lot of lighting for cooking.

for family area and living hall i think we're keeping natural light off of those. the house area usually gets too much sun during midday.