AC vs DC lighting efficiency

newl

Curiosity exercise here.

Situation:
12VDC LED strip lighting drawing 2.83A = 33.96W (measured from PWM controller on pass through)
240VAC lamp drawing 0.08A = 19.2W (calculated), 13W measured with inline watt meter (akin to kill-o-watt).

Knowing the inaccurate nature of the inline meter and losses due to the inverter and the possible calculated value of the PWM controller, how close are the two of these in terms of consumption? I'm wagering on the lamp.

BB.

Re: AC vs DC lighting efficiency

newl wrote: »

Curiosity exercise here.

Situation:
12VDC LED strip lighting drawing 2.83A = 33.96W (measured from PWM controller on pass through)
240VAC lamp drawing 0.08A = 19.2W (calculated), 13W measured with inline watt meter (akin to kill-o-watt).

Knowing the inaccurate nature of the inline meter and losses due to the inverter and the possible calculated value of the PWM controller, how close are the two of these in terms of consumption? I'm wagering on the lamp.

Take the AC lamp--Actually both numbers can be correct. With AC, we have "power factor" to take into account. PF can be calculated by measuring the phase angle between the AC voltage and AC current wave forms. Plug a induction motor (inductive load) on an AC main, the current sine wave will be delayed from the voltage wave form by some number of degrees. Power Factor = Cosine (degrees delayed).

When the current is delayed from the voltage, the farther the delay, the less power the current will provide.

Sort of like pulling on a rope tied to a car--If you stand in front of the car, 100% of your pulling force is used to move the car forward. If you stand 60 degrees to the side, then Cos (60) = 0.5 PF -- Or only 1/2 of your pulling force is moving the car forward.

And for other circuits (like electronic power supplies), many just take current at the peak of the sine wave voltage. The "current efficiency" can be represented as power factor too (different/more complex math).

So, for the 240 VAC lamp:

Power = V*A*PF
PF= Power / (V*A) = 13 watts measured / (240 VAC * 0.08 A) = 13 watts / 19 VA = 0.68 PF

And that 0.68 PF is not unusual for CFL or LED lamps... Note I used the term "VA" -- the wiring and AC inverter need to be designed to carry the current (VA), and the battery bank (DC Side) will be discharged at 13 Watts (=19VA*0.68PF).

AC power is actually quite complex math--And if you have more questions, please ask. Some of this is actually quite important when designing AC Off Grid systems.

Similar issues with DC power systems. Many DC "electronic" devices do not take DC as pure current... There are switching power supplies that can also cause "pulse train" DC wave forms too--And there are VA (PF) issues too.

For example, remember one of the power equations is P=I²*R -- If you double the current (due to "poor" power factor), then the "self heating" due to resistance goes up 4x (and other effects too).

Sorry for going all over the map here--I am not sure if I am answering your question, and how much experience you have with power and the math.

-Bill

bill von novak

Re: AC vs DC lighting efficiency

A few points.

1) It is important to measure both with the same tool, since devices like kill-a-watts are not super accurate (2% max error when new.) For understanding AC loads a good current probe with a good voltage probe, going to a good oscilloscope, is very instructive (but expensive.) For most applications a good true-RMS multimeter does a pretty good job. Two identical true-RMS multimeters let you measure current with one and voltage with the other.

2) AC generally has power factor issues. Magnetic ballasts cause lagging power factor and cheap electronic ballasts have problems with peak rectification. These will both cause errors in your measurements if you are just measuring voltage and current separately.

3) The lumens per watt number (often on the lamp's data sheet) is generally where you want to start and then look at power quality issues.

4) Low power DC LED's typically use dropping resistors which are pretty wasteful.

inetdog

Re: AC vs DC lighting efficiency

bill von novak wrote: »

A few points.

1) It is important to measure both with the same tool, since devices like kill-a-watts are not super accurate (2% max error when new.) For understanding AC loads a good current probe with a good voltage probe, going to a good oscilloscope, is very instructive (but expensive.) For most applications a good true-RMS multimeter does a pretty good job. Two identical true-RMS multimeters let you measure current with one and voltage with the other.

2) AC generally has power factor issues. Magnetic ballasts cause lagging power factor and cheap electronic ballasts have problems with peak rectification. These will both cause errors in your measurements if you are just measuring voltage and current separately.

The effect of power factor is enormous for motors and large enough for non-linear loads like LED and CFL drivers that I would suggest relying more heavily on the 2%-5% accuracy of a wattmeter with good handling of distorted waveforms compared to a 1% true RMS meter and a power factor that is 10% or more different from 1.

The scope is only going to help you for loads with a displacement (phase-shift) power factor and otherwise linear behavior.
Eyeballing non-linear current effects is just not possible, IMHO.

newl

Re: AC vs DC lighting efficiency

BB. wrote: »

I am not sure if I am answering your question, and how much experience you have with power and the math.

I believe that you did. The PF calcs indicate that the AC draw is actually less than the DC draw. Experience with the power and math...much of what I knew as an Aviation Electrician has fallen out of my head I reckon.

My questions were not focused at design, merely curiosity and trying to better understand the values that I was seeing and which one (with or without losses and/or PF) would actually be drawing more and which less.

BB.

Re: AC vs DC lighting efficiency

The sad thing is that efficiency (and power factor)--And even specifications--All the "variations" add up...

For example, the standard derating for charging from solar during the day and running AC loads at night:

0.81 solar panel derating from STC * 0.95 typical MPPT charger eff * 0.80 typical flooded cell charging eff * 0.85 typical AC inverter eff = 0.52 end to end system efficiency

So--we are talking about almost 50% losses between that shiny solar panel marketing spec., and those electrons being consumed by your AC loads.

Why? Hot panels lose ~20% of their output voltage (all panels get hot under the sun--especially on nice summer days). A typical MPPT controller will average around 95% efficiency (less if you oversize the controller).

For PWM controllers, they are "more efficient" but because the way a solar panel (current source) interacts with a battery (voltage source/sink), there is also ~0.77 losses because of "miss-matches" in Vmp-array vs Vbatt-charging (just the way the physics/electrical engineering works out).

Lead Acid Batteries discharge at ~12 volts and recharge at ~13.5-14.5 volts--That is an energy loss. There are ways to operate a flooded cell battery bank at >90% efficiency--But most of us don't do it, or cannot really do it most of the time. You can get AGM batteries which are >90% efficient and can be almost 98% efficient. However AGM batteries cost 2x as much and (usually) do not last as long as comparable flooded cell batteries.

AC inverters--You run them at 40-70% load or so, and they can be >90% efficient. You run them at 100% load and they can be 80-85% efficient. Run them at 10% load, and you may get 65% or less efficiency... It all depends.

Add PF--Which is not great for induction motors (~0.67 PF is common) and poor PF for our "fancy lighting" (many CFL's have around 0.50 PF--"waste" 1/2 the AC current when making light)... Much of what we end up doing with off grid solar is conservation and efficiency improvements (i.e., buying smaller/energy start rated equipment) and/or using Kill-a-Watt type meters to figure out what is really happening with our AC loads (Name Plate ratings are almost useless for system sizing).

Energy usage is a highly personal set of choices... For many people who have never tried conservation before--Reducing power usage by 1/2 is not impossible.

In general, the rules of thumbs allow us to estimate/design a system that will, most likely, meet your needs very nicely--We try to be a bit on the conservative side and not bother with the detailed analysis--It gets confusing, there are all sorts of shifting values/efficiencies--Which, in the big picture--Can usually be just tossed into the overall rule of thumb design/efficiency numbers for "close enough".

Add that the energy received from the sun can easily vary by 10% or more from one year to the next, and nobody can predict when those 1 week of storms knock your solar panels out of the picture (with no sun)--You still have to have that backup genset or simply stop using most of your electricity when solar is not available.

We get very used to turning on a switch with reliable utility power... And we byte the bullet when the bill comes. Off Grid solar power will cost (out of pocket) something like 5-10x the cost of utility power (long term 20 year system life). All of a sudden, those conservation plans and electrical power usage/efficiency numbers really attract your attention.

-Bill

BB.

Re: AC vs DC lighting efficiency

I should add--Check the amount of light produced by each light as measured at the counter/where you need the illumination... Our eyes "see light" with a logarithmic scale.

If we see a 3 foot kid standing next to a 6 foot adult--That is a huge difference.

Turns out, if we walk into a room lit at 300 lumens and walk into another room lit at 600 lumens--Our eyes will hardly know the difference. A large difference is 10x different illumination levels. However, that 2x (or 10x) increase in lumens is a 2x (or 10x) increase in wattage. A significant change in your power budget.

-Bill

arby

Re: AC vs DC lighting efficiency

My pal across the lake has a 12 volt system. We checked the Track LED lights through his trimetric on 120 volt, and then removed the transformers and tried them on 12 volt. The draw on the trimetric read the same. We concluded that running separate wiring was not worth it.

Grinnin

Re: AC vs DC lighting efficiency

Do you need to run the inverter all the time if you use 12V for the lights?

Most of my lights use about a half amp at 12VDC (about 6W per area). The inverter uses a half amp just to be on with no load. For reading, turning on the inverter doubles the current used.

I have used LED lights, stereo, water, vent, and whatnot for 12 years (OK, 11 years 11.5 months). I added an inverter about 9 or 10 years ago, a Morningstar Sure-sine 300W.

On low-voltage LEDs and resistors -- there are many constant-current ICs out there to allow manufacturers to make efficient 12V LEDs. But the market for LEDs has had low quality and high quality components for sale for years. The LED arrays I built use a 7810 or 7809 voltage regulator to match the voltage to the LED string more closely making the resistor much, much smaller. And I'm just an amateur.

(So this being my first post I probably need to explain. I started building in Summer, '01 and have lived off-grid since then. It's now time to build the "real" house. (What, me hurry?) I registered to learn about modern components and revisit all the decisions I made then. I ran 12AWG for AC and 10AWG for DC. That was too much -- all the DC circuits carry so few amps that 12AWG would have been fine. PVs are so cheap now that making the house all-AC might work.)

ILFE

Re: AC vs DC lighting efficiency

Welcome to the forum.

Grinnin wrote: »

Do you need to run the inverter all the time if you use 12V for the lights?

No. It is only necessary for AC powered lighting and appliances.

By the rest of your post, I feel you know this already. Or, am I missing something here? Am I living up to having been born with blonde hair?

Grinnin

Re: AC vs DC lighting efficiency

ILFE wrote: »

No. It is only necessary for AC powered lighting and appliances.

By the rest of your post, I feel you know this already. Or, am I missing something here? Am I living up to having been born with blonde hair?

Thanks for the welcome.

I don't want to make assumptions about other people's systems.

My Internet provider would prefer that I leave my inverter on all the time. I don't comply and they live with it.

If other equipment in a system requires the inverter all the time, the question is about only the efficiency of the lighting and the efficiency of some additional current through the inverter. If the inverter can be switched off -- as mine can -- then the calculations must include the overhead for having the inverter on just so the lights work.

westbranch

Re: AC vs DC lighting efficiency

this is the conundrum we face when going from a temporary , part time use to a full time residence. There are several mod-cons that need full time power, such as modems or ? for eg, satellite connections.
Some makes of equipment can be shut down for a long period of time while others have problems reconnecting.
There are a couple of threads on that issue...

So it boils down to 'living with' the power losses in the inverter or possibly having 2 inverters, on small one for those constant loads and a larger one for the intermittent loads eg fridge, ad have it in standby/sleep mode to reduce that tare losses.

ILFE

Re: AC vs DC lighting efficiency

Grinnin wrote: »

Thanks for the welcome.

I don't want to make assumptions about other people's systems.

My Internet provider would prefer that I leave my inverter on all the time. I don't comply and they live with it.

If other equipment in a system requires the inverter all the time, the question is about only the efficiency of the lighting and the efficiency of some additional current through the inverter. If the inverter can be switched off -- as mine can -- then the calculations must include the overhead for having the inverter on just so the lights work.

I use 12vdc lighting for my wife's house and farm. I have an inverter, but will not install / run it there until we transfer there full time. Currently, we are running CFLs. For some reason, maybe they are too sensitive to the DC voltage, or maybe just not very well made, but three of eight bulbs have lasted less than two months from install. (It isn't like you get the best quality equipment available in SE Asia.)

Either way, I finally located some low voltage LEDs that will screw directly into the standard bulb sockets I installed at the farm. So, I will be replacing all of them with the LEDs, soon. I already know they draw less power than the CFLs. I just hope they last as long.

After we are living there full time, there will be a decision to be made to determine if we will convert all lighting to 220vac, or to leave it at 12v. I'm still not sure what I will do. But, the inverter will be running full time at that point, powering my computers 24 hours per day. So, the .4 of an amp that it draws while idle, will be a non-issue.

Regarding internet, my router just happens to have a 12vdc input. So, there is no need for me to run my network off AC.

newl

Re: AC vs DC lighting efficiency

ILFE, thanks for bringing up the 12vDC router input. That reminded me to check the input specs on my new one. Thankfully that's 12vDC and the new femtocell that I received today is also 12vDC input.

My plan is to build a mini system to run the two full time as I have poor mobile coverage with my current provider and the femtocell acts as a mini base station using the internet where my provider of that is solid as for coverage (3G anyway..4G coverage not so good and am hoping an antenna can resolve that). I've done similar for the radio fencing for my dogs and that has worked out quite well. It won't, however, support any other gear as it was purposely sized built.

Regarding inverter usage, I have two, a bigger one for my fridge and a smaller one for other electronics (and now lighting), the latter gets switched off at the breaker before I leave every day so as to not be a vampire.

Grinnin

Re: AC vs DC lighting efficiency

ILFE wrote: »

Currently, we are running CFLs. For some reason, maybe they are too sensitive to the DC voltage, or maybe just not very well made, but three of eight bulbs have lasted less than two months from install. (It isn't like you get the best quality equipment available in SE Asia.)

I had a similar experience but 100% failure rate with fewer bulbs. When I moved into this building I bought 3 12V CFL bulbs. The first failed at about a month. By 3 months, all 3 had failed. I switched to undercabinet quartz-halogen "puck" lights mounted on my ceiling while I learned to build LED arrays myself. (The state of LED bulbs then was pretty dim too.) CFL technology was mature by 2001, but producing CFLs for 12V required finding a manufacturer that could make some odd ones for a small market.

Here in the U.States, products for "normal" 110VAC are often built to work better than products for 12VDC. Twelve volts is for cars and a tiny number of nuts. :cool: While the first LED bulbs were for 12VDC, there has been an explosion of lights for household 110VAC in the states in the past 5 years.

ILFE wrote: »

After we are living there full time, there will be a decision to be made to determine if we will convert all lighting to 220vac, or to leave it at 12v. I'm still not sure what I will do. But, the inverter will be running full time at that point, powering my computers 24 hours per day. So, the .4 of an amp that it draws while idle, will be a non-issue.

Technology is always changing and a "modern" house will have new ways to use power that we don't even think about today. Standardizing on the most common voltage in your area could make changes easier in 10 years. I'm sure you're considering how much work you have in the 12V lights and how much work it is to switch over. I suspect that your system will change and making change easier will probably pay.

ILFE wrote: »

Regarding internet, my router just happens to have a 12vdc input. So, there is no need for me to run my network off AC.

I can't tell if you're currently running your router off 12VDC or if you plan to do that in the future. While your system may nominally be 12V, all the circuits rise to 15V for normal charging and higher for the occasional equalization charge (if you're using flooded lead-acid batteries). Check to make sure your router can handle the higher voltage. Products made for cars can usually handle 14 or so volts, but products made to run off an AC power supply often require a more stable voltage. Not all the time, but often. You can build or buy a regulator to keep the DC voltage down, but there are typically 2 limits: the regulator will drop the voltage a bit and the total amperage will be limited. At that point, using your already well regulated AC circuit may make sense.

Just like the new guy to wade in and start lecturing and assuming you haven't already hashed all this out. Sorry.

jcheil

Re: AC vs DC lighting efficiency

ILFE wrote: »

Regarding internet, my router just happens to have a 12vdc input. So, there is no need for me to run my network off AC.

Just remember, as it has been discussed here before, your 12v electronics may not like the 10.5v to 15.5v "swing" that exists in your battery bank. Some will have a tolerance for the slightly higher or lower voltages, some let the blue smoke out.

zoneblue

Re: AC vs DC lighting efficiency

Grinnin wrote: »

On low-voltage LEDs and resistors -- there are many constant-current ICs out there to allow manufacturers to make efficient 12V LEDs. But the market for LEDs has had low quality and high quality components for sale for years. The LED arrays I built use a 7810 or 7809 voltage regulator to match the voltage to the LED string more closely making the resistor much, much smaller.

The best way to drive led arrays is with constant current. This drives them to their max spec without risk of thermal runaway. To do this effciently switch mode PSU regulatiors are employed with a current sense system instead of a voltage sense system. Many (most?) plugin type led bulbs now run these type of reguators now. They are tiny about size of a dime. The main issues with them is that the regulator is strapped right on top of the led array heatsink, pretty much guaranteeing short life of the regulator components, filter capacitors predominantly.

Using constant current, you really want as many leds in series as you can. 24vdc just happens to be rather ideal to power 6 typical high output leds. Regrettably 48v dc exceeds the Vmax of all the typcial regulator chips.

On build quality i hear you about being not that flash in general. There are some signs of improvements in the latest ones, external phosphors, potted circuitry etc. But if i was interested in playing around with these id be looking to mount the regulator seperate from the bulb, say inside the cieling rose.