Questions about leds in series.

[Jtroutt19]

Hey guys I am having a hard time with this and it is making me feel stupid. So I am gonna ask for some help. Let's say I wanted to run 20 leds in series. 10 blue and 10 red.

Specs on leds

Blue 1000mA at 3.4v
Red 1000mA at 2.65v

Series circuit rules state that the total voltage is the sum of the individual voltage drops along the circuit. And current is the same through each component.

So due to the difference in voltage drops of the two individual LEDs is it possible to just run the leds in series with each other without and resistors.

I know it would work with resistors. As long as the resistor is made to handle the current flowing through it. But are leds different?

Using ohms law I can figure out the resistance of the two individual leds running at full bore but is that allowed. I'm confused I know just brought about this stuff to make myself even more confused. Will someone tell me if my mind is correct and I can do those and if not educate me. Thanks!

[stj]

Re: Questions about leds in series.

not good, run the setup as 2 series strings.

[petehall347]

Re: Questions about leds in series.

ideally vf should match closely .
or am i thinking parallel ?

[Jtroutt19]

Re: Questions about leds in series.

Would you be willing to explain why run them as two series strings?

[Jtroutt19]

Re: Questions about leds in series.

A lot of what I am reading elsewhere says that you can run different Vf leds in series and that in parallel they would need resistors. I did a little experiment last night and setup a little 15 led string of blue and red. I powered with right at 40v every led cane on and the string was pulling 400mA. Using an infrared temp gun each led was heating up the same. I am going to get an adjustable driver both current and voltage and try it again with corrrct voltage to drive them at 80%.

[stj]

Re: Questions about leds in series.

in series the voltage and current to each led will be equal, so one colour will be overdriven at the expense of the other.

[Jtroutt19]

Re: Questions about leds in series.

Why is that. Why the difference between resistors and diodes. Because resistors will drop an amount of voltage based on its resistance. Don’t leds have an internal resistance that allows it to drop a certain amount of voltage?

[Longbow]

Re: Questions about leds in series.

Of course you can run them in a series string. Prove it to yourself by putting them together on a bread board. I don't know why anyone would ask for trouble in this way, but it CAN be done if the voltage drops on all the individual LED's add up to the applied voltage. Ask the question: using a regulated voltage? Using a regulated current? Xmas tree LED strings are probably an example of just crossing your fingers and hoping for the best.

Since a typical LED will work over a wide range of currents, you can get away with sloppy designs. Just one faulty LED will shut it down. Brightness of different LED's will differ. If you run the LED close to its max current the lifetime will be short, as in any electrical system.

LED's are current driven devices. The actual voltage drop across the LED is incidental. Why not accept reality and use correct design methods?

[Jtroutt19]

Re: Questions about leds in series.

Ok maybe I am not explaining my thoughts correctly. I am trying to learn here.

I know a typical leds works over a range of currents. I am also aware they are current driven devices and no I am not wanting to drive the leds at full power. I am not wanting to do a sloppy design I am wanting to understand the why behind the correct design. Not oh it’s just that way. I want to know why it is that way.

I am not against accepting reality. But again I want to understand that “reality” I have thought I have made that perfectly clear in the above posts.

If I were to hook up resistors in series. Depending on their resistances I would see different voltage drops across each one. Do leds not work this way? If no why not.

In my case i would get an Appropriate driver meaning a constant current driver but I would get it where it would be adjustable. Might as well get it where the voltage can be controlled as well to match voltages. But you say this is poor design without explaining why that is so. Can you explain why that is so?

[Curious.George]

Re: Questions about leds in series.

Hey guys I am having a hard time with this and it is making me feel stupid. So I am gonna ask for some help. Let’s say I wanted to run 20 leds in series. 10 blue and 10 red.

Specs on leds

Blue 1000mA at 3.4v
Red 1000mA at 2.65v

Series circuit rules state that the total voltage is the sum of the individual voltage drops along the circuit. And current is the same through each component.

The potential (voltage) you will need to overcome (to cause meaningful current to flow) is determined by the SUM of all of the forward voltages of the LEDs in the string AT THE CURRENT YOU PLAN ON PASSING THROUGH THEM.

So, if you plan on passing "1000mA" through the string, you will need to develop a potential of 10*3.4 + 10*2.65 = 60.5V. Therefore, you will need a 1000mA current source with ~60.5V "compliance". As the current source regulates its output current (and lets its voltage vary within that compliance), if the Vf's of this particular set of 20 LEDs happens to be 59.9V -- or 61.3V -- AT THAT 1000mA forward current, the output voltage of the power supply (current source) will climb to ensure the 1000mA output current is maintained.

[Note that a normal power supply is a voltage source. The output CURRENT varies to ensure the VOLTAGE remains at the desired potential].

So due to the difference in voltage drops of the two individual LEDs is it possible to just run the leds in series with each other without and resistors.

The resistor in a typical LED driver acts as a ballast -- it limits the current through the LED to a particular value. Otherwise, imposing anything > Vf across an LED will emit an incredibly brief (and terminal!) flash of light.

The resistor's role remains that of ballast in the 20 LED string -- if you choose to build your current source out of a voltage source behind a resistor!

For example, if your Vf actually ends up being 60.5V, you could design a 70V power supply (constant voltage) that can handle "up to ~1000mA". Then, put a 9.5 ohm resistor in series with the power supply's output. When connected to the LED string, the LED string will drop 60.5 of those 70 volts impressed and the remaining 9.5 volts will appear across the resistor. As the resistor has been conveniently chosen to also be 9.5 ohms, exactly 1000mA will flow through the resistor (and the LED string).

[Of course, the resistor will get awful hot as it will be dissipating ~10W; but, the LED sting will also be hot as it will be dissipating 60.5W!][/QUOTE]

When driving larger loads (like these LEDs), its usually more efficient to design a better current source that doesn't "waste" as much power (as heat in the resistor). You can typically also add capabilities like being able to easily vary the drive current -- so you can "dim" the LEDs (by passing less current through them or duty-cycle modulating them... things that may ALSO affect the Vf of the string]

If I were to hook up resistors in series. Depending on their resistances I would see different voltage drops across each one. Do leds not work this way? If no why not.

Hooking up the LEDs in series means each sees the same current ("sum of currents into a node equals sum of currents out of node"). While a resistor exhibits a linear relationship (ignoring temperature and tolerances) of voltage across to current through (V=I*R), a diode is a non-linear device. So, a small change in current can make a big difference in voltage across. Add to this the part-to-part variations and you would expect to see a bunch of different voltage drops in such a string -- even though all are being driven with the same current.

[For resistors driven with a constant current, the differences in voltage drops would be a direct reflection of the "value error" (tolerance) of the individual resistors in the string... 10%, 5%, etc. This is easy to wrap your head around and simple to figure with basic arithmetic. Non-linear devices, not so much! :< ]

Might as well get it where the voltage can be controlled as well to match voltages.

YOU don't control the voltage! YOU set the output current (just like you would normally set the output VOLTAGE on a power supply). The current source "controls" the voltage in response to the characteristics of the load so it can maintain the desired output CURRENT. Again, this is just like a (voltage mode) power supply "controlling" the current in response to the load's characteristics to maintain the desired output VOLTAGE.

A (voltage mode) power supply designer sits down knowing he needs to develop X volts at a range of output currents that will be dictated by the load. A (current mode) power supply designer sits down knowing he needs to develop X amps at a range of output voltages that will be dictated by the load.

[Jtroutt19]

Re: Questions about leds in series.

Ok great info thank you.
This is the link to the driver I was thinking of using

https://www.ledsupply.com/power-supp...ell-hlg-series the type A one which from the description says this

The A-type HLG products are IP65 rated and include on-board internal potentiometers where the output voltage and current can be adjusted. These models have the two black circles that you see on the top side of the product. This is where the adjustment is made. The range of adjustment is noted for each model in the tables below, but the voltage can typically be adjusted around ±10%, while the current range is typically between 50% to 100%

The way I read this is that I would be able to adjust voltage out to where I want it but also be able to adjust current out. Allowing me to set it right where I need it to get the performance I want out of it. All the while allowing me the option of dimming the enitre string with a 0-10v pwm signal if need be. That is why I said control voltage. Or am I reading the description of this driver wrong?

[budm]

Re: Questions about leds in series.

The potential (voltage) you will need to overcome (to cause meaningful current to flow) is determined by the SUM of all of the forward voltages of the LEDs in the string AT THE CURRENT YOU PLAN ON PASSING THROUGH THEM.

So, if you plan on passing "1000mA" through the string, you will need to develop a potential of 10*3.4 + 10*2.65 = 60.5V. Therefore, you will need a 1000mA current source with ~60.5V "compliance". As the current source regulates its output current (and lets its voltage vary within that compliance), if the Vf's of this particular set of 20 LEDs happens to be 59.9V -- or 61.3V -- AT THAT 1000mA forward current, the output voltage of the power supply (current source) will climb to ensure the 1000mA output current is maintained.

[Note that a normal power supply is a voltage source. The output CURRENT varies to ensure the VOLTAGE remains at the desired potential].



The resistor in a typical LED driver acts as a ballast -- it limits the current through the LED to a particular value. Otherwise, imposing anything > Vf across an LED will emit an incredibly brief (and terminal!) flash of light.

The resistor's role remains that of ballast in the 20 LED string -- if you choose to build your current source out of a voltage source behind a resistor!

For example, if your Vf actually ends up being 60.5V, you could design a 70V power supply (constant voltage) that can handle "up to ~1000mA". Then, put a 9.5 ohm resistor in series with the power supply's output. When connected to the LED string, the LED string will drop 60.5 of those 70 volts impressed and the remaining 9.5 volts will appear across the resistor. As the resistor has been conveniently chosen to also be 9.5 ohms, exactly 1000mA will flow through the resistor (and the LED string).

[Of course, the resistor will get awful hot as it will be dissipating ~10W; but, the LED sting will also be hot as it will be dissipating 60.5W!]

When driving larger loads (like these LEDs), its usually more efficient to design a better current source that doesn't "waste" as much power (as heat in the resistor). You can typically also add capabilities like being able to easily vary the drive current -- so you can "dim" the LEDs (by passing less current through them or duty-cycle modulating them... things that may ALSO affect the Vf of the string]



Hooking up the LEDs in series means each sees the same current ("sum of currents into a node equals sum of currents out of node"). While a resistor exhibits a linear relationship (ignoring temperature and tolerances) of voltage across to current through (V=I*R), a diode is a non-linear device. So, a small change in current can make a big difference in voltage across. Add to this the part-to-part variations and you would expect to see a bunch of different voltage drops in such a string -- even though all are being driven with the same current.

[For resistors driven with a constant current, the differences in voltage drops would be a direct reflection of the "value error" (tolerance) of the individual resistors in the string... 10%, 5%, etc. This is easy to wrap your head around and simple to figure with basic arithmetic. Non-linear devices, not so much! :< ]



YOU don't control the voltage! YOU set the output current (just like you would normally set the output VOLTAGE on a power supply). The current source "controls" the voltage in response to the characteristics of the load so it can maintain the desired output CURRENT. Again, this is just like a (voltage mode) power supply "controlling" the current in response to the load's characteristics to maintain the desired output VOLTAGE.

A (voltage mode) power supply designer sits down knowing he needs to develop X volts at a range of output currents that will be dictated by the load. A (current mode) power supply designer sits down knowing he needs to develop X amps at a range of output voltages that will be dictated by the load.[/QUOTE]

+10
BTW, OP should also study
Vf VS If graph of the LED.
Vf VS Temperature.
Etc.

[redwire]

Re: Questions about leds in series.

You can run the LED's entirely in series, this needs the highest voltage and lowest current.
Not sure what you have for a power supply, constant current or constant voltage?

LED's have a negative temperature coefficient so their voltage drops a bit with heat. About -0.11V red, -0.2V blue, from 25->85C. This effect adds up the most when all are in series.

So your expected string voltage is 34V blue + 26.5V red = 60.5V at 25C.
As the LED's heat up this drops to 32V blue + 25.4V red = 57.4V at 85C.

A constant-voltage power supply alone would not work. If you set it to 60.5V for 1,000mA... as the LED's warm up, current would skyrocket because the LED's are conducting hard at say 58V but your power supply is trying to force 60.5V into them.

You need a ballast resistor to absorb the difference (60.5 to 57.4V) , or instead use a constant-current power supply. The larger power supplies on fleabay are capable of high output voltage (62V) and would work fine with your string, like MeanWell IDLC-65-1050 or -700.
Smaller CC power supplies at 1,000mA go to max. 37V or so, not enough voltage for this 20-LED string. You could use two small PSU's and split the string into two of 10 LED's each.

[Jtroutt19]

Re: Questions about leds in series.

I looked st those drivers the meanwel you mentioned but came across the MeanWell HLG type A ones as well figured it would suit my purposes better

[budm]

Re: Questions about leds in series.

Specs on leds

Blue 1000mA at 3.4v
Red 1000mA at 2.65v

Can we see the actually spec sheets of these LED's?
Should have pretty good size heatsink and good thermal management to go along with them.

[Jtroutt19]

Re: Questions about leds in series.

https://cdn.badcaps-static.com/pdfs/...2fe64ea173.pdf

Data sheet for them

[Curious.George]

Re: Questions about leds in series.

The way I read this is that I would be able to adjust voltage out to where I want it but also be able to adjust current out. Allowing me to set it right where I need it to get the performance I want out of it. All the while allowing me the option of dimming the enitre string with a 0-10v pwm signal if need be. That is why I said control voltage. Or am I reading the description of this driver wrong?

The string of LEDs represents a "potential" (voltage) that you need to overcome before you can get current to flow through them. Imagine it's a hose that is 60.5 ft tall. You've got a bucket of water that you want to run through the hose. In order to get it into the hose, you've got to climb to a height of at least 60.5 feet! Any less and nothing will flow through the hose.

So, the first criteria for your "power source" is that it must be capable of creating a potential of at least 60.5 volts between its + and - leads.

You want to push 1000mA through the LEDs. So, the second criteria is that your power source must be capable of delivering at least 1000mA -- AT THAT 60.5V POTENTIAL.

Now, the trick is that you want EXACTLY 1000mA to flow through the string, regardless of what the ACTUAL Vf of the LED string is at this point in time (it WILL vary). But, you don't want anything MORE than that to flow as it would exceed the capabilities of each of the LEDs in the string (the weakest one will fail first)

Once you've exceeded the ~60.5 volts, the LED string looks like a short circuit. So, a traditional (voltage mode) power supply would just pump out as much power (amps) as it could before it -- or the load -- blew.

With this in mind, the "poor man's" way of driving an LED string is to set a power supply to something >= the highest Vf expected for the string (look at component tolerances and expected die temperatures during operation).

THEN, remembering that the power supply will try to put out as much current as it can into this "short circuit", adjust the current limit on the power supply to the value that you want -- 1000mA. As long as the Vf never exceeds the voltage setting for your power supply, the power supply will operate in current limited mode and effectively act as a current source.

[minor fudging, here, as it depends on how the current limiting is implemented]

Note that you can take advantage of the fact that the power supply will WANT to output something greater than Vf (say, 63V?) but is being constrained to only output Vf (~60.5V) due to the current limiter's action.

If the load suddenly does NOT appear as a short circuit (e.g., one of the LEDs fails "open"), then the current limiter will not constrain the output and the output VOLTAGE will rise to the actual voltage setting (63V in this example). So, if you see a voltage > Vf on the output of the power supply, you know that the "bulb is blown"... that there is NO current flowing through it! (it's usually easier to monitor voltages than currents so, here, the voltage tells you that the current is likely 0)

[budm]

Re: Questions about leds in series.

https://cdn.badcaps-static.com/pdfs/...2fe64ea173.pdf

Data sheet for them

I will NOT drive them at 1000mA. I would drive it at 800mA or less with really really good thermal management.

[Jtroutt19]

Re: Questions about leds in series.

The reason I was just going buy a driver made for driving leds. I understand what you are saying though.

[redwire]

Re: Questions about leds in series.

At 60W, you're gonna need forced-air cooling.
XPE LED's are tiny and getting rid of 2-3W each is very difficult. They are mounted to MCORE (aluminium-core PC board) and then mount that on a large heatsink.
The stuff for sale from china is misleading, it looks like a 30W or 60W LED plate but has no heatsink, so it will quickly overheat and die.