Lighting, Light Fixtures, Ceiling and Exhaust Fans - LED Lighting Restricting wattage assistance desired

I am converting to LED's and I LOVE IT I love the light I love the massive insane power savings.

One problem. They do not last long (at full brightness) they dim inside of a month of 2 and go bluish like the earlier units did. I assume he is using a phosphor to get the whiter white and that he is either overdriving or they are not being cooled well enough (112F inside the tube)

He is going to look into trying to reduce the current load to prevent this from happening. I want to try something on MY end.

they consume 11-14 watts (198-300 led's)

Is there something I CAN DO on my end to physical cut off the wattage to ONLY give the lamps say 6-8 watts each and simply do not let them have anymore than this. this will insure they are not overdriven and reduce the heat (over 85'F begins to damage LED's)

These are finished products with 110V plugs on the end. I want something I can plug these into or insert inline by cutting the power cord to restrict the wattage

Suggestions ??

Chris Taylor
*************

Perhaps an inline dimmer.

Hmmm I know a normal dimmer will not work since it reduces voltage (led's must remain in a specific voltage range) BUT my pop just told me he thinks FAN dimmers are different because you can burn out a fan motor with the wrong voltage. the only other thing to modify is to either pulse the voltage on/off (doubtful) or to reduce current

So I am going to go buy a fan dimmer and meter it and see what it does. THIS would be very ideal since I can then connect ALL of them to one dimmer and simply use a meter to drop them to the desired wattage.

Any other ideas ? its VERY important that I DO NOT mess with the voltage. I need to lower the current/watts they draw around 11-14watts I want to lower that to 7-8 watts and see what it does to the brightness and lifespan

You are NOT allowed to use a dimmer to control a receptacle.

Well First I do not tend to permit people to dictate to me what I can and can not do and second I am not trying to control a receptacle I am trying to control a Light Fixture which is what a dimmer was designed to control

From Outlet or power source to Dimmer and then to Lights.

Also telling me I am not allowed to do something with NO explanation or reasoning as to why is not the most effective way to endear me to accept such "orders"

I am "allowed" to do anything I want, bunch of free will kind of stuff. I think what you meant to say is that it is unsafe for some reason to control an outlet with a dimmer switch. Which makes sense, altering the output of a normal receptacle that anything can otherwise be plugged into crosses me as unsafe ie how will X device respond if plugged into such a thing or what if it is a high draw item how will the dimmer itself respond. Yeah that crosses me as not safe.

You are NOT allowed to use a dimmer to control a receptacle.

Bob,didn't we go through this before and come to the conclusion you could ONLY IF the recep involved was not a general purpose recep and was somehow designated as such. It seems we were speaking of a recep above some cabinets or something similar.I don;t remember the thread enough to locate it. Maybe you can.


nerys, current is not typically controlled since it is a function of the voltage and impedance of the circuit. A circuit will allow what it will allow concerning current based upon those 2 items.

To lower the current, you must increase the impedance or lower the voltage.

a watt is a measurement of power and is determined by calculating the impedance of the circuit and the voltage supplied. It is not something you control without changing the parameters I mentioned and as you stated, you do need to maintain a particular voltage.

You idea of pulsed voltage would do what you want but may cause a noticable flickering. What you would be doing is reducing the watts consumed in a given time period reducing the heat build-up but what you are doing is turning the thing on and off continuously. That would reduce the average current but not the instantaneous current when the LED is actually fluorescing.

Chris,

In your very first post you state that you have "finished products with 110V plugs on the end."

That to me states you wish to control a receptacle., which is against code. The NEC does not generally state WHY they prohibit certain things, but the reason you came up with seems like a good one.

nap.

Yes, we had this discussion before. I remember that we came up with that you can't dim a receptacle, regardless of where it is, but that we agreed that a receptacle above the cabinets was not general purpose. But regardless of what we came up, Chris did NOT state where the recepacle he has will be located.

nono I don't want to pulse it (I was suggesting that thats how a fan dimmer might function)

How does a fan dimmer differ from a light dimmer ?

Plus pulsing would be like turning it on and off that many times. While LED's are not as suceptable to cycle fatigue I imagine tens of thousands of times a day would do some damage :-)

So how do I lower the current ? LED's are current hungry. IE they nuke themselves. If you plug in an unlimited current source they will draw an unlimited current at least till they nuke themselves. As they get hotter they draw more current which makes them hotter so they draw more ... you get the idea :-)

So you have to restrict the current to an LED to prevent this runaway effect.

I do not think he is restricting the current enough. I think they are running "too hot" which is called overdriving. Fine for things like flashlights which are only powered for short durations but not so good for light fixtures that have to run for 8-18 hours a day :-) It will burn them out in very short order.

While he is looking into greater current restrictions I want to try and save the $400 in lights I already have :-) plus dimmer would equal longer lifespan. I would rather install a few extra tubes at higher initial cost to get the same brightness and net a longer lifespan. Especially if It will preserve the the white coloring that I desire.

So I want to externally lower the available "current" going to the lights so they will be not be overdriven. The problem is I can not lower the voltage. Right now they consume 110v at 14watts I want to reduce that to 110v and 8 watts or so.

Suggestions?

Also the receptacle is irrelevant. Here is how they are wired

Receptacle in my bedroom going to UPS circuitry to a battery array and an inverter to a power strip and then the Lights plug into the strip. When the power goes out sticking with "normal" usage I can go for days without power and illuminate the room normally. Cut out 2 fixtures and go conservative and I can go possibly 10 days. :-) Eventually I want to experiment with removing the need for an outlet at all and convert to solar for the lights. Since they only consume around 45-50 watts this should be feasible.

I would modify this array to include whatever gadget we come up with so no "receptacle" will be modified in any way.

So how do I lower the current ? LED's are current hungry. IE they nuke themselves. If you plug in an unlimited current source they will draw an unlimited current at least till they nuke themselves. As they get hotter they draw more current which makes them hotter so they draw more ... you get the idea :-)



The LED's need what they need.

Do you have aby idea how the lighting system is built now? What type of LED's? What resistance is added to each LED? Are the LED's wired parallel or series or groups of both?

The wattage of a circuit is not determined by what power is available to it. It is determined by the impedance/resistance of the circuit. Unless you change that, you aren;t going to reduce the power consumption.

The only way I can see to alter the power consumption is to alter the resistance within the circuit that is simply used to reduce voltage and provides no other benefit. Depending on how the circuits are constructed controls any gain that migh be possible.

It looks like he is using a single resistor and therfore running the whole thing in series ?? I see 1 resistor and 4 capacitors (square pillow style caps it seems)

which causes another problem if an LED fails there is now more current available to the rest of the LED's causing a cascade failure.

Since HE is controlling the current I KNOW I can also control the current I just do not know how to.

Can I wire a large resistor right into the 110v line ? how would I calculate what resistor to use ?

Racraft I believe the OP is talking about installing the dimmer in his appliance not on the wall. The NEC ends at the structure wiring. Even if this gadet has a recepticle on the end of it, it would still not be covered under the NEC. Getting UL listing is another issue, but does not come into play till the design is done and he wants to sell his product.

Neyrs, can you go over the wiring again. You said that led needs a specific voltage, but then you said that the original design has them wired in series with a resistor. It seems to me that you could increase the size of that resistor, or use a varying resistor like a dimmer. Can you explain why you cant?

I just did a little looking into LEDs and I think that a fan speed control will turn on and off the power within the flicker range of human site.

In addition a fan speed control will not reduce the wattage, it will just change it's relationship to time. So you will have less KWH at the lamp, but the same KW when at the moments it is in operation.

LEDs only work on DC so there must be a diode in the circuit and the caps are so that the DC sine wave is pure DC not half wave.

I found that the driving force for LEDs is current, but unlike resistive and inductive loads, the relationship between voltage and current is not linear with diods it is exponential. A small change in voltage will cause a huge change in current.

Adding resistance in series with the source will reduce the wattage and extend the life of the diodes. They also have newer diodes on the market that last longer. The problem is that you want to reduce the voltage just a tiny bit. So you must add very small resistance.

My source suggests that you need the mfg spces on the diodes to get accuate info regarding the voltage to wattage relationship for that specific diode. But I do have one other possible way to come up with an answer.

Purchase a 10K ohm potentiometer. I say POT cause I cant spell worth a darn. A 20 turn pot would work best for this, but make sure it can handle the voltage and wattage for your system.

The POT will have two ends and a wiper. Wire the wiper and one end in series with your source. When you turn the POT all the way in one direction you will have zero resistance, or what you have now. Gradually turn the POT till you get the desired light intesnity. Remove the POT from the circuit and measure the resistance. This will be the value of the resistor that you need to add in series to your lights.

Important note. I am talking about using the POT in series with the DC source to the lights, not the AC source from the wall outlet.

My source suggests that you need the mfg spces on the diodes to get accuate info regarding the voltage to wattage relationship for that specific diode. But I do have one other possible way to come up with an answer.
===============

that is key info. you cannot calculate anything unless you know what you have.

if the led;s are in series, when one burns out, the entire string would go.



I never said you cannot cause the current to change but you cannot change it without altering the voltage applied to the led's. a led needs, allows, (or causes if you wish) whatever current ohm's law decides based upon the voltage supplied. The led is a fixed resistance device and cannot be changed.

It also requires a specific minimum voltage to operate. That is what the added resisitance does. (alters the voltage applied to the led)

You can reduce the current that flows through the led but to do that you lower the voltage applied, which is very controlled with an led.

Also unless you change the circuit impedance, you aren;t going to change the wattage of the circuit. You can circuit leds differently and utilize resistors to limit current (by altering voltage applied) but each different variation has its own detractors and benefits.

by utilizing the resistance of the leds circuited differently to allow the removal of added resistance allows the circuit to be designed to reduce the wattage of the circuit and still provide correct operation of the leds. There is no magical device you can add that simply reduces the current to the leds but still allows proper operation.

I'm backing out of this thread. OP needs to read up on ohm;s law if he wants to understand the current situation and how to change the current flow and why his ideas to this point are not practical or effective.

I'm backing out of this thread.

That is a good idea

damn - I was afraid thats how Fan Dimmers worked :-(

So this means there is no simple thing I can plug in to reduce wattage without taking the unit apart ? - that sucks.

With there being ONE resistor in theory I can just replace that one resistor with a resistor with higher resistance right or will this also alter voltage ??

Have any good resources on resistors so I can tell what is in there and what I should replace it with? I will have to find my minibook on it I know the colored bands tell me what it is plus a tolerance band or something like that.

I was wondering how he went from AC to DC since LED's can only take DC - I figured you needed more components than I am seeing. unless he orders the LED's so they are 120vdc and then just flips to DC ???? Might take fewer components ?

If I do use a normal dimmer does this just lower voltage ? will that also at the same time lower the current? I am not very skilled at this but I do not see this working. IE while it might lower the current I feel that the available current will still be higher than what the lamp needs meaning the current "INSIDE" the lamp will not change because the external current is still higher than what it needs. I guess I can just try it and see what happens. If I frin one its not that big a deal. Plug a meter in and "see what happens"

Thanks for the info guys if anyone has anything else that might be useful please let me know

as for wiring. I ran an extension cord into my closet where I have a UPS I run output from the UPS to a strip in the ceiling. the 4 Light fixtures plug into this strip. I also have a wireless "switch" between the UPS and the Strip for the lights so I can turn it all on and off. Thats it.

Ok, now you're starting to see the light. (pun intended).

resistor color code:
http://engr.astate.edu/circuits/resistor-code1.jpg

some informational sites:

http://www.kpsec.freeuk.com/components/led.htm

http://www.physics.uoguelph.ca/tutorials/ohm/Q.ohm.intro.series.html

http://www.physics.uoguelph.ca/tutorials/ohm/Q.ohm.intro.parallel.html

a circuits resistance determines the current flow of that circuit. If you change the resistance, you will also change the voltage applied to individual componants of that circuit. You need to maintain a specific voltage across the leds lest they won;t work. When that voltage is applied to the leds, the resulting current is predertermined and is a function of the voltage applied and the internal resistance of the led. You can;t change that.

Now what you can do is: determine the optimum voltage for the leds. calculate the circuit as is. figure in additional resistance to lower the voltage of the calulated voltage is above the optimum voltage.

What doing this will lower the current through the leds but it isn;t what you are choosing to control. You are controlling the applied voltage which results in the resultant curent flow.

If you simply reduce the voltage below an operational level, the current will go down but more importantly, the things won;t light up either.

concerning the ac/dc thing. he may have a rectifier in there. You did mention some other componants. Because of the low wattage of the fixture, a rectifier could be quite small and possibly mistaken for other componants.

the one thing you need to understand is you do not control the current. the circuit controls it. Having more than enough current available does not cause the current to be greater. Think of this. You probably have a 100 or 200 amp service in your house. just because you have that much current available, it doesn;t mean your appliances use that much current. They use what they use based upon the impedance ( a calculation of the effective resistance of a circuit including reactive resistance sources) of the circuit. Voltage is what you apply, current is the result.

An LED is a polarity-sensitive device but it most assuredly WILL light on AC. It will only light on the proper half-cycle of the AC sine wave.

There is no such thing as a DC sine wave. If you have a sine wave component of your DC supply it is because of superimposed AC or less-than-perfect rectification of the original AC.

It is generally recommended that LEDs have individual current-limiting resistors and be wired in parallel rather than all wired in series with a single current-limiting resistor. With the series arrangement if any component (LED or resistor) fails the entire string of LEDs will fail to light. Also, by using a single resistor you cannot allow for the inconsistencies of the individual LEDs.

Nap has stated the absolute rule: A circuit will take whatever amount of power is necessary for its operation. The ONLY thing that may be done external to the circuit is to change the VOLTAGE and that change will effect the change in current and subsequently the wattage.

All simple lamp dimmers and fan controls work by pulsing the output voltage which in turn reduces the AVERAGE voltage but not the peak voltage. Since a LED works from the DC component of the applied voltage it is necessary to reduce the PEAK voltage and this cannot be done by a simple dimmer control.

The easiest way to control the existing LED fixture would be to install a variable transformer between the supply (from the UPS) and the LED fixture. With this arrangement one could "dial in" the exact amount of light they required. A second way would be to install the variable resistance potentiometer as suggested by jwhite. I don't know that 10k ohms would be the proper value and one would need to be sure that the power-handling capability of the pot was greater than the power drawn by the LED fixture. A fixed resistance installed in the power leads of the fixture (of the proper value) could also work.

Ok most of that makes sense. but it logically fails on several aspects which means either we are missing something or some OTHER aspects of my knowledge are flawed. I am going to assume the latter.

From what I understand LED's are basically resistors (diodes??) when they get hotter they produce greater resistance drawing more current.

This causes a runaway meltdown effect.

To prevent this they must be current limited.

Yet you just said we can not control current so this creates a logical conundrum for me. What am I missing?

Also a charger for a car battery has different AMP settings for its charge rate. the battery is not DRAWING a current I am PUSHING a current. ALSO when I jack up the amps the voltage remains unchanged??

This again creates a logic problem with us not being able to control current.

I assume I am just not understanding something here - can anyone clarify?

Also I could have a failed memory (it was a rather long time ago) but the last time I ran a resistor off a 9v battery I still got 9v on the other end ??

The logic "fails" because your knowledge of the subject is limited. (Sorry, that is about the nicest way I can express it.)

LEDs are NOT "basically resistors", they are basically rectifiers. An LED is a special type of semiconductor that when energized with the proper amount of electricity will glow, or "emit", light. They do not increase in resistance when they get hot and draw more current. The temperature increase is a function of the electron flow through the semiconductor material. More electrons (higher current) means more heat generated.

To keep the heat that is internally generated in the semiconductor from destroying that device it is necessary to "limit" the current flow and that limited current flow is what limits the internally generated heat.

I did not state that current flow cannot be limited, what I stated is that current flow cannot be limited external to the circuit under study. Ohm's law dictates how voltage, current and resistance are interrelated. In its simplest expression Ohm's law states the voltage (E) is equal to current (I) times resistance(R) or E=IR. This statement can also be expressed as I=E/R or R=E/I. By knowing two of the three values you can always find the third value.

Power, which is the rate of doing work is expressed by the equation power (P) = voltage (E) times current (I) or P=EI. Power is measured in watts.

As an example let's suppose that your LED fixture is rated 10 watts at 120 volts. This means that when connected to a supply of 120 volts the the fixture will "draw" 0.0833 amperes of current. The calculation is: 10 (watts) / 120 (volts) equals 0.0833 amperes. This is an absolute, it is not changeable without changing either the voltage or the current.

In other words, the current cannot be changed without changing either the voltage applied to the circuit or the resistance of the circuit.


Let me point out here that the above is true of all direct current (DC) circuits. Alternating current (AC) introduces a few more variables but for this discussion I will ignore most of them.


You stated: "Also a charger for a car battery has different AMP settings for its charge rate. the battery is not DRAWING a current I am PUSHING a current. ALSO when I jack up the amps the voltage remains unchanged??"

Sorry, you're wrong. The battery in the charging circuit is a load, it "draws" (for lack of a better word) energy from the charger. When you change the "AMP" setting you are really changing the voltage applied to the load. The battery has a fixed resistance (that changes with the state of charge) and to change the charging rate (the AMPS flow) it is necessary to change the voltage applied by the charger.

And no, there is no logic problem because you are changing the circuit parameters.


As for your experiment with the nine-volt battery and the resistor...you need a load for the circuit before the resistor has any appreciable effect. I assume that you used a meter to determine that the resistor had no effect upon the voltage. The meter simply was not a great enough load to make any difference in the circuit.

From what I understand LED's are basically resistors (diodes??) when they get hotter they produce greater resistance drawing more current.
==========
the more resistive something becomes, the less current will flowgiven the same voltage


Yet you just said we can not control current so this creates a logical conundrum for me. What am I missing?
=========================
not sure. have you researched any of the sites I posted?

Also a charger for a car battery has different AMP settings for its charge rate. the battery is not DRAWING a current I am PUSHING a current. ALSO when I jack up the amps the voltage remains unchanged??
======================
who says the voltage remains the same?



Also I could have a failed memory (it was a rather long time ago) but the last time I ran a resistor off a 9v battery I still got 9v on the other end ??
=========
without knowing where you took any readings, I can't comment on your observations. On the other end is not clear enough for me to tell you much.

that makes sense that no load would not kick in the resistor so to speak. I already know my knowledge in this area is VERY limited :-)

As for the charger I know the voltage remains the same because the voltage meter says so. Then again who knows what that meter on the charger is supposed to indicate. maybe thats target voltage ??

There has to be a way for me to control wattage. Example if I plug a 500ma solar panel into something that something can not take more than 500ma because that is the max that the panel can produce.

SO how do I build or buy a power supply that will put out a voltage and wattage that I specify. I am thinking along the lines of a bench power supply. Do any such beasts exist that let me define both voltage and wattage ? so I can tell is produce 110v and 35 watts and no more ? (4 fixtures)

If I can change voltage it stands to reason I can also change current somehow while keeping the same voltage. I have accepted it will not be easy and probably not cheap. but there has to be a way to do it ???

If I can not control the power going to the lights maybe I can build or buy a power supply that physically can not produce any more power than 110v 35watts so that this is how much the lights will receive simply because there is no more to be given ? Like using a solar panel. it limits the current by simply not being capable of producing any more than that. Can I build something that works similarly ?

Say I wanted to build a device that will convert 11ov to 110v at 35 watts and then run the lights off that. Now the max available is 35 watts so they can not take anymore than that.

?? What would it take to build something like that or can I buy something like that ?

"In other words, the current cannot be changed without changing either the voltage applied to the circuit or the resistance of the circuit."

So can I put a new circuit in place between the ups and lights to limit the power to 110v 35watts before going on to the lights so that only 35watts is available to them ?

remember the brown outs in California a couple years ago?

That is what happens when the supply does not meet the demand. Voltage drops.

You can;t get away from the physics of the game.

the wattage of a circuit is not figured by what amount of power is supplied, it is determined by how much power it uses. That is fixed at a given resistance.

head over to the thread about a water heater and voltage and wattage. That ,ight help you understand.

http://forum.doityourself.com/showthread.php?t=301490

So there is no way to build a regulated power supply to limit the current ?

Read the water heater thread just shows the relationship between volts and amps but I am talking creating an artificial environment.

I do not think I can just add resistance before the lights since that will just cause the whole thing to draw more power (the current the lights want and the current the added load wants)

Nerys, several people have tried to explain the relationship between supply and load along with the relationship between current, wattage, voltage and resistance. You refuse to accept these facts.

I doubt that any of us here are bona fide physics instructors and so I don't think we can be held responsible for our teaching methods having failed you.

I strongly suggest that you visit your local library and check out several textbooks on basic or elementary electricity. Maybe after you study these textbooks it will become clearer what we have been trying to explain.


BTW, I suggested that you could use a variable transformer to allow continuously variable lighting form your fixture. Here is a link that would work well for your situation.

http://www.iseincstore.com/index.asp?PageAction=VIEWPROD&ProdID=23

I know your trying I just have trouble accepting thats its impossible :-( I mean we create electricity it just seems odd we can not regulate both voltage and current :-(

What will that gadget do it looks like it alters voltage ? (output 0-132vac)

is this the same function a dimmer would perform? he did say they are compatible with dimmers (I asked)

So what happens if I reduce the voltage (I assume thats what the transformer does?) the wattage will go down with it ?

What is the difference between this gadget and a dimmer ? (finer control??)

Seeing as how I know little about this stuff what would be a good place to start reading ? anyone have a URL or other good site for "beginners" info on this kind of stuff ?

CAN I replace the resistor inside to alter the wattage comsumption? its only one resistor and its very large I think even my soldering skills can handle that one (and its through the PCB so no nasty surface mount stuff (nasty for one with moderate soldering skills) I was trying to avoid doing that but I think I can risk a single fixture to try.

My end result desired is this. I want to REDUCE how much power they consume to extend there lifespan.

At this point WHAT are the possible ways to do this (no matter what they entail) I can take apart one and take pictures of all the components if that will help. Just tell me what I need to do get or read.

So I can not do it with something external fine lets go internal. I can afford to risk a single $19 light to experiment with.

What do you need pictures of? etc...

I hope the ADMIN is ok with this just remove it if its not ok but I do not know how else to do this.

http://www.nerys.com/LED/ This is JUST pictures for this discussion - I do not see a way of attaching images to a post and they are 2mb in size total Since its directly relevant to this topic I am hoping that is within the rules??

Check out 06 for the circuit line pattern on the backside then goto Constructions folder for pictures of ALL the components on the board. there are 5 of those brown units 2 resistors 1 small one next to a black components and then the large one at the opposite end of the board. if you need a better picture of something let me know

I do NOT make or SELL these. I just buy them to USE and am trying to FIX them. I want to switch to LED really badly :-) I have no vested interest besides simply wanting them to work properly so I can use them.

This one is 300 LED's since I only see 1 major and 1 smaller resistor and not 300 I am assuming he is wiring in series ??

To see the problem I am having check out the last image before going to Construction. Problem LED.jpg the images were shot at identical camera settings so they are good relative to each other. the dimmer one USED to be as brights as the bright one about 4 weeks ago.

Since the original maker of these had no way of knowing the exact voltage in the location where they would be used, I would bet that some sort of voltage requlation is built into his design. Hence minor adjustments to the 120 supply voltge would have little or no effect on the lamps voltage.

In an electrical circuit voltage is the given. It is what comes from the utility or the transformer that you install.

The only thing that we can change in the field is resistance. We cannot change voltage, we cannot change current we can only add or subtract resistance. Amperage will be the result of the voltage divided by the resistance. SO to limit the amperage in a circuit we add resistance in series with the load. In physics when they call something a law it does not mean the same thing as what the government calls a law. Laws made by the governmet can be broken, and if you get caught they punish you. Laws of physics cannot be broken. It is impossible. It does not matter that they do not make sence to you yet, they are fixed, finite, unchangeable.

OHMS LAW dictates what can and cannot be done in an electrial circuit. SO if the resistance of the leds is the same the only way to lower the amperage would be to add resistance in series, and this would reduce the voltage at the diode, but not the total voltage. The reistor would convert the extra energy into heat.

http://www.allaboutcircuits.com/vol_1/index.html

You have said repeatedly that you want to decrease power to the LEDs. Power or Watts is also part of ohms law. P = E x I where P is power E is volts and I is amps.

What you are not understanding is that the way to decrease power is to add resistance and therefore decrease voltage and thus amperage.

Somewhere you read that if you change the voltage to the LEDs you will cause them not to work, or damage them. This is a misnomer. It should have read that subtile changes in voltage to an LED will cause major changes in the amount of current that they will pass. So the resistance of an LED changes when the voltage applied changes. This is why you need to make very minor changes to the voltage.

So to clarify SMALL voltage changes will result in large current changes? I think we will be ok them because I also do know LED's have a voltage "range" they are comfortable in. if I lower the overall voltage by even as much as 10v I do not think they will care all that much.

So your saying I can do this 2 ways. Lower the external voltage which will reduce current?

Or add resistance (larger resistor) which will also lower voltage?

is that correct? At this point if you say it will work I am just going to try it the worst I can do is burn out a $19 lamp. I can afford to do that. Ive got little to loose and a lot to gain. I have 2 brand new ones so I can plug them both in one with lowered voltage and one without to a UPS as a control and see what happens after a few weeks. I think if I can keep the temps below 90' I will be ok (they are running 112 now)

Yea, but if you lower the voltage on the supply side of any voltage regulator, you could cause dammage to it if you lower too much.

I think you need to lower the voltage after the regulator and before the lamps. This is where I would insert the 20 turn pot. and experiment from there.

He did say it was dimmer compatible. Is that what a light dimmer does (lower the voltage??)

If so I would prefer that route as $84 is a bit much :-) I can goto walmart and buy a dimmer today.

I plan to only lower the voltage slightly. Just enough to "take the edge off" so to speak and see what effect it has on longevity. ie unnatural dimming over time and color change.

OK, here's my 2 cents worth.... You need a constant current power supply that is adjustable. No big deal. Just Google 'constant current power supply' for a lot of info. Or you can go to www.constant-current.com and see some examples. Some of the better power supplies are made by Lambda, but they are a little pricey.

Nashcat

After looking at the pictures I think that jwhite is correct in his thought that the fixture has a regulated power supply. If this is the case then a variable transformer on the input will not achieve the desired results.

On the other hand, since the supplier states that the fixture may be used with a standard lamp dimmer then it would not have a regulated supply.

At this point I have nothing more to offer.

Sorry.

ok, 10 volts variance!!!!! depending on how the leds are wired 10 volts can be a lot. If you went to the led page I gave you, you will see some voltage requirements for some typical leds. We're talking (at the most) 1 volt difference between design voltage and max voltage.


You just posted that the manufacturer said these things are dimmer compatible. Find out what type of dimmer is recommended and go from there. You can build a small box to plug into a wall recep that has the dimmer in it and facilities to plug in or attach the light string to that. (so your house will still remain code compliant)

then see what you have.

http://www.ledsupply.com/led-drivers.php

Those drivers are for just a few LEDs, your running hundreds. Im sure there exists LED Drivers for home use when your using dozens at a time.

well I have a few that burned out and I just compared a picture taken now 1 month after installation and its well wow. at least 10 times dimmer than when new.

So I grabbed a transformer 12.6v from radio shack gonna wire one up tonight and see what it does.

Also ordered a variac off ebay and will play with that when it gets here and see what happens.

I am going to try and adjust voltage till they are 50-60% of max brightness and then see if they "stay" that bright after a month. If they do I will be quite happy. this variac can run every fixture in my house if I were somehow able to wire them all together (they take about .113 amps each tops) so power won't be an issue. Got one for $40 shipped :-)

OK Got the variac. Works like a dream. I had to turn the voltage down to almost 75 volts!!! but at 75 volts I only lose 40% of the brightness and create almost no heat. (usually 8-11' higher than ambient) which is perfect!

A month from now I will know if it works (ie if it stays white and does not dim further on me)

So a question. In your guys opinion what is the simplest and hopefully cheapest way to straight up drop a 120 line to 75-80 volts ? something super simple and cheap :-) I just want to cut the voltage by 40 volts. What is the easiest way to do this on the AC side. IE cut the power cord and "insert" something inline or plug this into something and then that into an outlet. Small being preferred but cheap and easy being paramount ? For rooms where I use 4 or 5 tubes I just use a variac but in places like my bathrooms where I only have one tube variacs are a rather expensive solution :-( I do not need to be able to adjust the voltages. Just set it for 80v and go with it.

Suggestions?

Still hoping to find a solution for this.

Its been over a year since I installed the LED's with the variac. So far so good. No Dimming No color change and NO failures. Even the fixture that were partially failing have STOPPED failing.

I just need a cheap way to get the same result as a variac without having to use a variac.

I have found out that a variac is just a transformer that is variable. SO what I need is a "fixed" transformer or set of transformers that will net me around 80volts. if its a little off thats ok.

Current capacity is not an issue as they draw so little current. around 3.5-4.0 watts per lamp. (measured on the 110v side)

Suggestions?

I have found out that a variac is just a transformer that is variable.

It was back in post #18 that I suggested you use a variable transformer. In post #25 I again suggested the variable transformer with the link to the catalog page with the variacs. At the top of the page it has the catalog flow and it states variable transformers>variacs.
Variac is just General Radio's (or whoever bought out General Radio) name for their line of variable transformers.


Now that you know what you want is approximately eighty volts to supply these lamps the rest is easy. Take the difference between 120 volts and 80 volts and you get 40 volts. (if your line voltage is something other than 120 volts adjust the number accordingly).

You previously stated these lamps draw 0.113 amperes at 120 volts. Using Ohm's law (R=E/I) that means that each lamp has an internal resistance of 120/0.113=1,062 (rounded up) You need to drop the 120 volts by 40 so we now take Ohm's law to find the necessary resistance that must be added to reduce the 120 volts to 80 volts. Again, it is R=E/I only in this case the voltage we use is the voltage we desire to "drop" across the newly inserted resistance. So it looks like this: R=40/0.113=354 (rounded up).

So for one of your LED lamps you need to insert a 354 ohm resistor in series with one of your 120 volt supply leads. Before you can do this you also have to determine the watt rating of this resister and for that we use the formula P=IE where P=power in watts, I=current in amperes and E=voltage or 0.113x40=4.5 watts. So you need a resistor rated at 354 ohms and 5 watts.

If you want to run several lamps off of a single resistor then you would add the current flow for each lamp (0.113 amperes) and the use this combined amperage draw in Ohm's law as previously described. Don't forget to also figure the wattage of the resistor as that will increase significantly.

Most likely you will be using adjustable wirewound resistors to achieve both the resistance required and the power (watt) rating. You will also need to mount the resistor where it has a the ability to transfer the heat (generated by reducing the voltage) to the air and at the same time cover the exposed wiring connections.

Hmm I would prefer to find a transformer solution.

If I understand how resistors work if I take an 11watt lamp and use a resistor to make the lamp run at 80v 4 watts then I am just transferring those extra 7 watts to the resistor?

Is that right? ie will the whole system (light and resistor) still consume the same power? or am I missing something?

I eventually want to "remove" the lighting for my house from the grid. LED's give me the potential to go solar/battery since they consume so little power. So wasting watts is really a bad idea since I will just have to pay for that in solar/battery capacity.

What if I were to "replace" the resistor thats already in the lamp? if I upload some images of the circuitry of this lamp. Can you determine if its even practical for me to try it?

I think the longest but easiest solution might be to put a potentiometer in place of the original resistor and then I use the thermometer and individually adjust each light.

Also what happen if a resistor "FAILS" does it continue to feed power to the light IE killing it or does it die "open" and the light just turns off and I just replace the resistor?

If it dies closed is there a way to build a safeguard into it?

I really appreciate the patience. I am slowly but surely "grasping" the knowledge needed for this.

Also if there is some good turtorial sites I can visit to expand my knowledge on this stuff please let me know :-)

Variac's are minimally practical. the problem with variac's is simple. Besides being Large and no good for smaller rooms where there is no place to hide them they are EXPENSIVE as all hell :-)

The only way I can get them is to take "pot shots" at cheap ones on ebay (I have acquired 3 so far)

will the whole system (light and resistor) still consume the same power?
Yes.

I eventually want to "remove" the lighting for my house from the grid. LED's give me the potential to go solar/battery since they consume so little power. So wasting watts is really a bad idea since I will just have to pay for that in solar/battery capacity.
Then you do not want the 120 volt LED lamps. Instead what you want to do is to add low-voltage (12 volt) direct current wiring and 12 volt D.C. LED lamps. I'm actually doing a similar yet less extensive night light/standby system in my house. I am using small LED clusters already packaged for 12 volt operation.

http://www.mpja.com/prodinfo.asp?number=17356+OP

These give out more than enough light for night lighting or during power outages although I would never consider them suitable for replacing conventional lighting. You may want to buy bulk high output LEDs and make your own clusters with printed circuit boards and 1/8 watt current limiting resistors. Bulk LEDs and resistors are quite reasonable compared to already manufactured assemblies.

Using solar and batteries in a low voltage (12 volt) arrangement can be fairly economical although it will never be less expensive than power from a public utility. It will certainly be nice if you experience frequent or long power outages. However, if you plan on using a large array of photovoltaic panels with batteries and inverters to produce utility quality power (and maybe even utility interactive to allow selling power back to the utility) you had better have some pretty deep pockets because even with governmental incentives your out-of-pocket costs are going to be sizable.

What if I were to "replace" the resistor thats already in the lamp? if I upload some images of the circuitry of this lamp. Can you determine if its even practical for me to try it?
Probably not. :(

I think the longest but easiest solution might be to put a potentiometer in place of the original resistor and then I use the thermometer and individually adjust each light.
This will certainly work but it has the same drawback as using a fixed resistor, i.e. the wasted power dissipated by the potentiometer which is nothing more than a variable resistor.

Also what happen if a resistor "FAILS" does it continue to feed power to the light IE killing it or does it die "open" and the light just turns off and I just replace the resistor?
The resistor would fail open and the lamp would fail to operate.

There are few possibilities with transformers but they either will cost as much as the Variacs or will require a fair amount of "fiddling" to accomplish the operation of the LED lamps you have on standard 120 volt utility power. They would also have the problem of requiring extensive re-wiring of the lighting circuits in the house and because of the voltage this would come under the electrical codes.

Think about my idea for a completely separate low voltage system. It will be simpler, easier to install, bypass most electrical code issues and be readily usable with a solar / battery system.

Those ideas can not work. The cardinal issue here is price. The lamps you show are $9 plus shipping. I am currently buying 198 LED tubes for $12 shipped and these ARE bright enough for primary lighting.

The math for the solar is also a little off. Remember by going to LED I am using far far less power. Examples.

Lighting my bedroom ALL lights on. 35watts

Upstairs bathroom 18watts. Smaller Bathrooms (2) 7.5 watts each

Kitchen 30watts. Dining Area 9 watts

see the picture here? I figure I can illuminate the ENTIRE house with a margin at under 400 watts and thats if I turned on every single light in the house (after led conversion)

Suddenly the size of the battery pack and solar array becomes much much smaller and MUCH more affordable. instead of a $10,000 solar panel a $500 solar panel might do the job. (remember JUST for the lights)

No sell back just yet grid tie ins are still to expensive at about $1600

I can not use 12v since any savings anywhere else will be negated 1000 fold in the cost of wiring and buying the 12v bulbs. (anyway inverters are cheap and very efficient over 95%)

As for the resistors taking power.

Can someone explain that to me. I mean there is a resistor IN the lamp right now (its just the wrong size) its certaintly not drawing 15amps max from my outlets. its drawing 12-14 watts.

IE I want to REPLACE the WRONG resistor with the RIGHT resistor. would that not work? I originally disregarded this path as too complicated for me. but there are only a few components on the board and my soldering skills have improved slightly. I also have enough "failing" lamps on hand (before I got the variac) that I can TEST my skills with near impunity.

I can understand how daisy chaining another resistor inline will add to power consumption but what about REPLACING the existing resistor?

thanks!

Ii don't know what your utility rate schedule is but I just ran some numbers. If you run an aggregate of 400 watts for 12 hours a day at a power cost of 20 cents per kilowatt hour (my rate is a little more than 8 cents in the winter) the monthly cost is less than $30. Most likely your usage is less than half of the example maybe less than a quarter of the example along with you likely having a lower cost per kilowatt hour.

Assuming that you can get a suitably sized photovoltaic array to keep the battery charged during all kinds of weather you still need the deep cycle battery AND an inverter. I doubt that you would need a pure sine wave inverter for just the lights so a 500 watt modified sine wave inverter would set you back maybe fifty bucks. Now all you need is the battery. At 200 watts output the inverter has an input of 12 volts at about 18 amps. Figure maybe eight hours minimum usage at that rate with no input from the PV and you get something like 144 ampere-hours. That's still going to be a pretty hefty battery and it is going to have an equally hefty price tag. And that is probably a minimum, I would like to see at least double the capacity in the battery. Add to that maintenance and eventual replacement cost and you may find that utility power isn't as expensive as you once thought.

Don't forget, if you want to use the PV>battery>inverter route you will STILL need to rewire all the lighting circuits separate from any receptacle circuits and include the pro;er overcurrent protection...all in accordance with your local electrical code.

I know. I already have the inverter so thats no problem.

My guess is I would actually have maybe 100watts burning at any one point in time. but I would shoot for minimum double as you suggested anyway. I am eventually hoping to be able to go a "week" on battery power alone.

Sure its not going to be ultra cheap but I can do is a piece at a time. I would actually need 230amps of batter power not 114 since its VERY damaging to leads to drop them below 50% DOD. so to get 114amps of capacity I would need double that to keep the batteries over 50% dod. but I can initially do that on the cheap. I can get car batteries for $15 a pop from the junkyard. so figure $100 for the battery pack.

Eventually I want to use NIMH batteries. 230 D cells would give me a 200amp 13.8v power pack and with nimh I can safetly use ALL of that 200amps. so I could make it a 115cell pack for 100amps and expand it later. That would only cost about $500 at current prices for 10amp D cells. the trick would be the charger :-) charging such a beast would be interesting to say the least.

The other nice thing about an inverter is that its 110v power. SO I can put this whole rig right next to a grid outlet. If something goes wrong I can just unplug from the inverter and plug it into the grid.

Its more a proof of concept. the idea is that we can lower out lighting power usage SO much that it suddenly becomes DOABLE for the average person to have "zero" grid load for lighting. that is something that impossible with incandescents where one room can use more power than I want to use to light the whole house. and still impossible with CFL.

Ehh I have no idea. Its just an idea. For now I just want to get the lights working.

There has got to be a CHEAP way to drop the voltage to 80v. I see transformers in radioshack for $4. are there no transformers or combinations of transformers that will get me what I want?

I am betting there is but I lack to knowledge to even know WHAT to look for. If radioshack can sell me a 18v AC power supply for $9 why can I not build one to do 80v for even double that price.

how are they dropping the 110v to 18v?

Ok here are some pictures. there are only 8 components not counting the LED's themselves.

http://www.nerys.com/ledlights/ its the DSCN pictures

4 capacitors 1 smaller capacitor. 2 resistors (opposite ends of the unit) and 1 gadget which I assume to be the rectifier?

I am assuming he is straight up converting 120vac to 120vdc and just runs them in series. takes about 33leds (typical forward voltage of 3.7v) so I would "guess" its 6 parallel groups of 33 led's in series or something like that. At least the numbers appear to point to that.

OR hem might be running them off AC on the opposing cycles one group of 3 this cycle another group next cycle. BUT I do not notice the "glimmer" you usually see when they do this.

I am hoping I can just replace those resistors to get what I want. would be more work but it would cost close to nothing even radio shack sells resistors 6 for a buck.

That "gadget" in the fourth picture with the markings DB107S is a full wave bridge rectifier. It converts the AC input to a pulsing DC output. A single phase, full wave rectifier has an average output of 63.6 % of the AC input. At least one of the capacitors is for filtering the ripple from the DC output.

Why don't you use a paper and pencil to trace out the circuit of the lamp? As I previously wrote I didn't think I could tell much from pictures of the circuit board. Those resistors seem to have a different color code than I am used to or maybe the colors are not reproducing well over the Internet.

Try to find a copy of Ugly's Electrical References for a small pocket-sized reference to many of the questions you are asking. I think it is about $10 and if I remember correctly I got mine at a Home Depot or Lowes.

Oh, car batteries, properly known as starting and ignition batteries, will not have the same characteristics as deep cycle batteries, especially junkyard batteries. Even with the relatively light load of a 500 watt inverter they will crap out long before a deep cycle model.

Once you are able to draw the equivalent circuit for one of these lamps I may have some more to offer. If I'm in a charitable mood (not tonight) I'll give you some transformer experimenting tips. You do have a digital volt-ohm-milliammeter don't you?

The paths are somewhat visible but critical parts are hidden underneath components. I will see what I can figure out. I will also see if maybe he will send me the schemes for it. Who knows he might.

I will also stalk a local lowes and look for that book.