LED Tree kit question

[kalel]

There's this interesting tree kit called "Three-Dimensional 3D Christmas Tree LED DIY Kit Red/Green/Yellow LED Flash Circuit Kit Electronic Fun Suite"

It seems to have 4 flashing segments, according to the description:

This circuit is composed of 16 pcs LED (light emitting diode). The 16pcs LED divided into four groups, and each group has different flashing frequencies, when four groups flash together, this item looks like a small Christmas tree. The circuit is powered by a 9V battery.

I'm curious how simple would it be to bypass the flashing circuitry in order to put in slow changing RGB LEDs? I think it would look more relaxing than the flashing.

The seller I'm looking at doesn't provide a schematics. Maybe another does.

So the remaining question is if you believe this would be relatively simple to do based on the PCB images. If it's a complicated endeavor, it might not be worth it (plus, I have no idea other than the assumption if it actually looks better that way).

Edit: Based on the 9v battery thing (I see 4.5v from batteries or 5 from USB), I'm not sure if the rest of the description is correct either. I assume the images are probably correct.

According to some comments, not all LEDs work properly due to forward voltage difference and parallel  connections, and according a comment, this might be the schematics: http://www.haoyuelectronics.com/Attachment/CTR-30C/CTR-30C%20Flashing%20Christmas%20Tree%20Instruction.pdf

I understand that in such a circuit, the resistors and capacitors will determine the flashing frequency. But is there a simple way to make it zero, e.g. replacing the capacitors with resistors? And if that works, is there a better way?

[MK14]

There is no real point to the kit, if you are going to use your own automatic colour changing RGB LEDs.

Just use verostrip or something (maybe even your first simple custom PCB), placing the LEDs in your favorite Christmas Tree configuration.

If you insist on using the Kit, you can just leave out all the components, and make sure the RGB leds are connected to the ground and positive supply rails, via a suitable resistor (ideally one per Led).

[kalel]

There is no real point to the kit, if you are going to use your own automatic colour changing RGB LEDs.

Just use verostrip or something (maybe even your first simple custom PCB), placing the LEDs in your favorite Christmas Tree configuration.

If you insist on using the Kit, you can just leave out all the components, and make sure the RGB leds are connected to the ground and positive supply rails, via a suitable resistor (ideally one per Led).

That image looks neat. Still, the kit is well designed, green (if that's actually better) and at less than $3 you get the boards, the USB cable and probably the battery holder too. It's cheaper than or equal to the custom solution, although you can take more pride from working with a protoboard. Custom designing and ordering a PCB for that project is also not especially cheap in comparison with the kit, while it would be a relatively simple and fun  thing to design (I'm sure anyone can do it for just some automatic color changing LEDs), it's probably not worth it.

[MK14]

PCB's can be a pain to rework. I.e. if you need to cut tracks, especially if they are under solder resist. Drilling tiny holes can be tricky as well.

But sure, on cost grounds the kit makes lots of sense.

These days, with electronics as a hobby, it can easily cost more (or even a lot more), to make something yourself. Which you could have just bought. But if you are doing it for fun and a great learning exercise, and many other reasons. It can be well worth the extra cost.

In the old days, you could actually save money by building (electronic) things yourself.

[kalel]

PCB's can be a pain to rework. I.e. if you need to cut tracks, especially if they are under solder resist. Drilling tiny holes can be tricky as well.

But sure, on cost grounds the kit makes lots of sense.

These days, with electronics as a hobby, it can easily cost more (or even a lot more), to make something yourself. Which you could have just bought. But if you are doing it for fun and a great learning exercise, and many other reasons. It can be well worth the extra cost.

In the old days, you could actually save money by building (electronic) things yourself.

I agree that if it requires cutting tracks and adding jumper wires all over the place, I'm not sure it would even look so great in the end. Unless the wires are all green, of course. I was wondering if there was a simpler solution, e.g. replacing the caps with resistors, or shorting out the transistor traces in some way, or something similar that doesn't make physical modifications to the circuit PCB.

One advantage of the self color changing RGB LEDs is that they do a lot of the controlling themselves (or disadvantage if you want to run your own special color patterns, but that might be a different type of project with an affordable MCU). You could just feed them a reasonable voltage and that's about all that is necessary. Not that a resistor per each LED can hurt.

As for the point that buying is cheaper than making, I do wish it was the other way around as an extra incentive. But when you save on cost, you can work on more projects. If we look at it a different way, both making and buying are cheaper than before, perhaps sacrificing some quality. At least when making for fun, you don't have to worry about longevity or use more expensive quality components (unless that is your goal - to make something of superior quality).

[MK14]

You should be able to hide the 'new' wiring, UNDERNEATH the PCB, where it won't normally show. You might be able to cut one lead of the LEDs shorter than the other, and then solder a resistor on it. Hence the PCB would then just need to supply the supply rails.

Like this:



Although that would NOT look so neat. It might look better, if you bend the resistor by 90 degrees. So the LED looks almost like it did, before messing with it.

Anyway, there are lots of ways of changing it. I've only shown one or two of the ways.

[kalel]

You should be able to hide the 'new' wiring, UNDERNEATH the PCB, where it won't normally show. You might be able to cut one lead of the LEDs shorter than the other, and then solder a resistor on it. Hence the PCB would then just need to supply the supply rails.

Like this:

That's a neat and simple idea for adding some resistance. Thanks for your tips so far.
Edit: Well okay, the neat is relative, but functional and simple for sure.

[MK14]

You could probably hide the resistor, by covering the new lead (with resistor), with heat-shrink, or insulation (sleeves). If you want it to look as neat as possible.

[kalel]

You could probably hide the resistor, by covering the new lead (with resistor), with heat-shrink, or insulation (sleeves). If you want it to look as neat as possible.

I assume a 1/8 watt resistor is recommended (for the size)?

Not sure about the exact calculations for an RGB LED, but 1/8 should still be sufficient.

SMD are also out there, but probably a bit complicated to use.

[MK14]

You could probably hide the resistor, by covering the new lead (with resistor), with heat-shrink, or insulation (sleeves). If you want it to look as neat as possible.

I assume a 1/8 watt resistor is recommended (for the size)?

Not sure about the exact calculations for an RGB LED, but 1/8 should still be sufficient.

SMD are also out there, but probably a bit complicated to use.

At a guess (size-wise), the picture looks like a 1/4 watt. But an 1/8 watt resistor should be fine. Since you are likely to limit the current to no more than 20 milliamps (if not less than that), and it is running at 5 volts (more like very approx 3 volts because of the voltage drop across the LEDs).

So at most it is:
Power = Voltage X Current = 5 x 0.02 = 0.1 watts.
But a fair bit less in practice, because of the voltage across the LED(s) and the current you choose, may be a fair bit less than that.

[kalel]

Youtuber BigClive designed a colour changing LED tree that readers of this thread may find interesting. He sent some out to other Youtubers and Micah Scott (YT channel scanlime) recently did a fabulous video of her build. The camera never does them justice but she manages to convey the charm of it quite well nonetheless.

And Big Clive's original video

That's an awesome kit well worth mentioning. I don't think he is offering those kits yet, but people have asked.

At a guess (size-wise), the picture looks like a 1/4 watt. But an 1/8 watt resistor should be fine. Since you are likely to limit the current to no more than 20 milliamps (if not less than that), and it is running at 5 volts (more like very approx 3 volts because of the voltage drop across the LEDs).

So at most it is:
Power = Voltage X Current = 5 x 0.02 = 0.1 watts.
But a fair bit less in practice, because of the voltage across the LED(s) and the current you choose, may be a fair bit less than that.

I agree. Something like 100 ohm 1/8 might work. Additional resistance through the wires will add some more (even the USB cable can have a noticable voltage drop). There are always losses that add some resistance, including the simple USB cable.

Edit: Or if at night it seems too much, perhaps 220 ohm or such instead. As long as it is sufficient to allow all colors to show decently. But that can be tested first.

[MK14]

I agree. Something like 100 ohm 1/8 might work. Additional resistance through the wires will add some more (even the USB cable can have a noticable voltage drop). There are always losses that add some resistance, including the simple USB cable.

Edit: Or if at night it seems too much, perhaps 220 ohm or such instead. As long as it is sufficient to allow all colors to show decently. But that can be tested first.

The resistance of the wires etc, can usually be neglected, unless the currents are very high and/or the wires are extremely long (e..g. many metres).

It depends on which RGB colour changing LEDs you use (but 20mA is a common limit). I.e. where you buy them from, should have a datasheet and/or specification on the maximum recommended current for them. When you have them, you can try a few different values of resistance, on a breadboard, to fine tune how bright they are, relative to your room ambient light levels.

Yes, 220 might be right. It depends on the LEDs max allowable current and how bright you want it.

[tronde]

I have played with some of the Chinese "rainbow" colour changing LEDs. They should make a nice tree. You must have one resistor on each LED. They don't fancy being connected in series because of the internal logic circuitry.

If you are lucky, the holes in the PCB are large enough for you to have a thin insulating sleeve (strip something from thin wires) on one of the LED's legs. If so, you can mounth the LED's flush to the PCB and use either 1/8W axial resistors or some SMD resistors on the other side.

I think you will get the best visual effect if you can find LED's with a milky-white lens instead of the more common water-clear types.

[Zero999]

Edit: Based on the 9v battery thing (I see 4.5v from batteries or 5 from USB), I'm not sure if the rest of the description is correct either. I assume the images are probably correct.

According to some comments, not all LEDs work properly due to forward voltage difference and parallel  connections, and according a comment, this might be the schematics: http://www.haoyuelectronics.com/Attachment/CTR-30C/CTR-30C%20Flashing%20Christmas%20Tree%20Instruction.pdf

The circuit in the PDF, which is probably the same as the one used in the kit, is a ring oscillator. It's true that all those LEDs connected parallel won't light uniformly. They should each have their own series resistor.

I understand that in such a circuit, the resistors and capacitors will determine the flashing frequency. But is there a simple way to make it zero, e.g. replacing the capacitors with resistors? And if that works, is there a better way?

Omit all the capacitors and resistors, except for those in series with the LEDs and replace the transistors with a wire link between the emitter and collector connections.

You could probably hide the resistor, by covering the new lead (with resistor), with heat-shrink, or insulation (sleeves). If you want it to look as neat as possible.

I assume a 1/8 watt resistor is recommended (for the size)?

Not sure about the exact calculations for an RGB LED, but 1/8 should still be sufficient.

SMD are also out there, but probably a bit complicated to use.

I thought colour changing LEDs generally have the resistor built-in, so you don't need to add an external resistor, but I could be wrong. Check the data sheet for the LEDs you're using.

If you don't need resistors, then the series resistors can be replaced with wire links. If you do, then that's a pain because you want one resistor per LED. You could replace the series resistor with a link and add your own SMD resistors by cutting the appropriate tracks and scraping off the solder resist.

[MK14]

I thought colour changing LEDs generally have the resistor built-in, so you don't need to add an external resistor, but I could be wrong. Check the data sheet for the LEDs you're using.

I don't know which specific LEDs the OP was planning to use, so it is possible they don't need a resistor.

Taking the following one as an example, it says you DO need resistors (except at a specific, low voltage, please see below):

EDIT: On reflection. If it doesn't need a resistor at a specific voltage. It makes sense to either get ones rated for 5V (if that is your intended supply voltage), or use a voltage regulator (or something), to give them the exact 3.6V they need. Then you won't need to mess about fixing all those tiny resistors, into the circuit.

http://www.ebay.co.uk/itm/10-ULTRA-BRIGHT-5mm-COLOUR-CHANGING-RGB-LEDs-5000-MCD-/400409667498?epid=1888255197&hash=item5d3a46a7aa:g:ffIAAMXQydtTOl4r

PACK OF 10
ULTRA BRIGHT CLEAR COLOUR CHANGING RGB 5mm LEDs
THE LEDs HAVE A CIRCUIT BUILT IN
THE COLOUR CHANGE RATE IS APPROX  3-5  SECOND's
THE COLOUR OF THE LEDS CHANGES SLOWLY AND GRADUALLY
THE COLOURS ARE RED, BLUE, GREEN, ORANGE, YELLOW
AND A MIXTURE OF THESE COLOURS AS IT CHANGES
THESE LEDS ARE EXALLENT FOR ARTISTIC DISPLAYS etc

5000 MCD
FORWARD VOLTAGE  3.6V
MAX CURRENT 30 mA

YOU CAN USE A HIGHER VOLTAGE
IF THE CORRECT CURRENT LIMITING RESISTOR IS USED IN SERIES
SUGGESTED RESISTOR VALUES FOR DIFFERENT VOLTAGES
5V  100  OHM
6V   150  OHM
12V   470  OHM

[Zero999]

I thought colour changing LEDs generally have the resistor built-in, so you don't need to add an external resistor, but I could be wrong. Check the data sheet for the LEDs you're using.

I don't know which specific LEDs the OP was planning to use, so it is possible they don't need a resistor.

Taking the following one as an example, it says you DO need resistors (except at a specific, low voltage, please see below):

EDIT: On reflection. If it doesn't need a resistor at a specific voltage. It makes sense to either get ones rated for 5V (if that is your intended supply voltage), or use a voltage regulator (or something), to give them the exact 3.6V they need. Then you won't need to mess about fixing all those tiny resistors, into the circuit.

http://www.ebay.co.uk/itm/10-ULTRA-BRIGHT-5mm-COLOUR-CHANGING-RGB-LEDs-5000-MCD-/400409667498?epid=1888255197&hash=item5d3a46a7aa:g:ffIAAMXQydtTOl4r

PACK OF 10
ULTRA BRIGHT CLEAR COLOUR CHANGING RGB 5mm LEDs
THE LEDs HAVE A CIRCUIT BUILT IN
THE COLOUR CHANGE RATE IS APPROX  3-5  SECOND's
THE COLOUR OF THE LEDS CHANGES SLOWLY AND GRADUALLY
THE COLOURS ARE RED, BLUE, GREEN, ORANGE, YELLOW
AND A MIXTURE OF THESE COLOURS AS IT CHANGES
THESE LEDS ARE EXALLENT FOR ARTISTIC DISPLAYS etc

5000 MCD
FORWARD VOLTAGE  3.6V
MAX CURRENT 30 mA

YOU CAN USE A HIGHER VOLTAGE
IF THE CORRECT CURRENT LIMITING RESISTOR IS USED IN SERIES
SUGGESTED RESISTOR VALUES FOR DIFFERENT VOLTAGES
5V  100  OHM
6V   150  OHM
12V   470  OHM

Unfortunately, as with many Chinese products, no proper datasheet is a available.

I had a quick look for data sheets and found this. It isn't clear, whether they need a constant voltage source of 3.8V to 5V or the current should be limited to under 20mA and that's the forward voltage. I'd recommend using a resistor to be on the safe side.
http://www.optosupply.com/uppic/2016812904937.pdf

[kalel]

Finding the proper datasheet may be possible, but more times than not, you don't have it with the cheapest sources.

Some might work on 5v directly as intended.
For others, I guess that a step down can/could work.

A step down + some resistance might be the safest bet, but also the most complicated/costly. Not that a step down is very costly these days, just comparing.

Anyway, without using a resistor per each LED (e.g. with the stepdown solution) would them being in parallel still cause an issue? I mean, if we have 2 LEDs, one is turning "RED" and the other "BLUE" at the same time, will the "RED" "lock" the voltage of the "BLUE" so it doesn't show up fully?

Having a step down would probably be a good idea anyway. Most modules have a potentiometer (usually a tiny trimmer, but this can potentially be changed - or left as is), and you can fine tune the brightness. There is nothing bad about being able to make things less bright, or being able to change the brightness a bit without needing an MCU, which really isn't/shouldn't be needed for this case.

[Zero999]

Finding the proper datasheet may be possible, but more times than not, you don't have it with the cheapest sources.

Some might work on 5v directly as intended.
For others, I guess that a step down can/could work.

A step down + some resistance might be the safest bet, but also the most complicated/costly. Not that a step down is very costly these days, just comparing.

What do you mean by step-down? Reducing the voltage? This is different to just adding a series resistor which limits the current.

Anyway, without using a resistor per each LED (e.g. with the stepdown solution) would them being in parallel still cause an issue? I mean, if we have 2 LEDs, one is turning "RED" and the other "BLUE" at the same time, will the "RED" "lock" the voltage of the "BLUE" so it doesn't show up fully?

If two LEDs with different forward voltages are connected together in parallel, the one with the lower voltage hogs the power and the higher voltage one will either not light at all or be very dim. Red LEDs have a lower forward than blue so, the red would just glow.

[kalel]

Finding the proper datasheet may be possible, but more times than not, you don't have it with the cheapest sources.

Some might work on 5v directly as intended.
For others, I guess that a step down can/could work.

A step down + some resistance might be the safest bet, but also the most complicated/costly. Not that a step down is very costly these days, just comparing.

What do you mean by step-down? Reducing the voltage? This is different to just adding a series resistor which limits the current.

Yes, if we wanted to get e.g. 3.x volts from USB, a step down seems like a reasonable method.

Anyway, without using a resistor per each LED (e.g. with the stepdown solution) would them being in parallel still cause an issue? I mean, if we have 2 LEDs, one is turning "RED" and the other "BLUE" at the same time, will the "RED" "lock" the voltage of the "BLUE" so it doesn't show up fully?

If two LEDs with different forward voltages are connected together in parallel, the one with the lower voltage hogs the power and the higher voltage one will either not light at all or be very dim. Red LEDs have a lower forward than blue so, the red would just glow.

That might be another reason to use series resistors per each LED in any case.

[Zero999]

Finding the proper datasheet may be possible, but more times than not, you don't have it with the cheapest sources.

Some might work on 5v directly as intended.
For others, I guess that a step down can/could work.

A step down + some resistance might be the safest bet, but also the most complicated/costly. Not that a step down is very costly these days, just comparing.

What do you mean by step-down? Reducing the voltage? This is different to just adding a series resistor which limits the current.

Yes, if we wanted to get e.g. 3.x volts from USB, a step down seems like a reasonable method.

Like a voltage regulator? If so, that makes sense. You'll need something with a low dropout voltage, such as the TPS76201 but it comes in a small SMT package, which will make soldering a pain but is good because it's easy to hide.

http://www.ti.com/lit/ds/symlink/tps76201.pdf

[kalel]

Finding the proper datasheet may be possible, but more times than not, you don't have it with the cheapest sources.

Some might work on 5v directly as intended.
For others, I guess that a step down can/could work.

A step down + some resistance might be the safest bet, but also the most complicated/costly. Not that a step down is very costly these days, just comparing.

What do you mean by step-down? Reducing the voltage? This is different to just adding a series resistor which limits the current.

Yes, if we wanted to get e.g. 3.x volts from USB, a step down seems like a reasonable method.

Like a voltage regulator? If so, that makes sense. You'll need something with a low dropout voltage, such as the TPS76201 but it comes in a small SMT package, which will make soldering a pain but is good because it's easy to hide.

http://www.ti.com/lit/ds/symlink/tps76201.pdf

I meant more along the lines of a buck step down module. They cost less than $0.5 shipped now, and you can set the output voltage with a trimmer. I think that if the source voltage falls or increases, it will affect the output voltage. But with USB, we expect to have a reasonably regulated voltage to begin with, so this should be all right. They claim a maximum (at a specific setting of course) efficiency of 96%.

Here's a short clip of what happens when 2 diodes are in parallel, but only 1 (1k in this case) resistor is used in series. So USB > 1k Resistor > 2 LEDs in parallel.  The test was done to see if we can see red with other colors at the same time with these. And the answer is yes as expected, at certain points blue becomes very dark.

https://streamable.com/3vdxf

The flickering is not visible with eyes.

[MK14]

I meant more along the lines of a buck step down module. They cost less than $0.5 shipped now, and you can set the output voltage with a trimmer. I think that if the source voltage falls or increases, it will affect the output voltage. But with USB, we expect to have a reasonably regulated voltage to begin with, so this should be all right. They claim a maximum (at a specific setting of course) efficiency of 96%.

Here's a short clip of what happens when 2 diodes are in parallel, but only 1 (1k in this case) resistor is used in series. So USB > 1k Resistor > 2 LEDs in parallel.  The test was done to see if we can see red with other colors at the same time with these. And the answer is yes as expected, at certain points blue becomes very dark.

https://streamable.com/3vdxf

The flickering is not visible with eyes.

(Taking assumptions, since you seem to be saying there is no datasheet as such) Then you would need a slightly higher voltage (e.g. 5V, enough to ensure consistent brightness between LEDs and their colours), than the bare minimum they need and a resistor for each to share the current.

What voltage was the 1K resistor connected to (if you already said, I did not notice, sorry. Maybe it's 5V, as you mention USB) ?
And does one LED on its own, properly light up as regards all colours (specially Blue) at that voltage with the same resistor ?

[tronde]

You must have one series resistor per LED.
Mine is 2.9V to 3.6V depending on colour. Calculate for 2.9V and 20mA, and you should be safe. I have been running one from 12V with 470 ohm for several years. They are usually bright, so no problem choosing even a lower current.

[kalel]

You must have one series resistor per LED.
Mine is 2.9V to 3.6V depending on colour. Calculate for 2.9V and 20mA, and you should be safe. I have been running one from 12V with 470 ohm for several years. They are usually bright, so no problem choosing even a lower current.

Yes, one resistor per LED is needed of course, just wanted to see what happens when they are in parallel with a single resistor (in this case, e.g. RED and BLUE LEDs just wouldn't work together - Blue would turn off).

Several years sounds great. Are you using it in some ornament?

I meant more along the lines of a buck step down module. They cost less than $0.5 shipped now, and you can set the output voltage with a trimmer. I think that if the source voltage falls or increases, it will affect the output voltage. But with USB, we expect to have a reasonably regulated voltage to begin with, so this should be all right. They claim a maximum (at a specific setting of course) efficiency of 96%.

Here's a short clip of what happens when 2 diodes are in parallel, but only 1 (1k in this case) resistor is used in series. So USB > 1k Resistor > 2 LEDs in parallel.  The test was done to see if we can see red with other colors at the same time with these. And the answer is yes as expected, at certain points blue becomes very dark.

https://streamable.com/3vdxf

The flickering is not visible with eyes.

(Taking assumptions, since you seem to be saying there is no datasheet as such) Then you would need a slightly higher voltage (e.g. 5V, enough to ensure consistent brightness between LEDs and their colours), than the bare minimum they need and a resistor for each to share the current.

What voltage was the 1K resistor connected to (if you already said, I did not notice, sorry. Maybe it's 5V, as you mention USB) ?
And does one LED on its own, properly light up as regards all colours (specially Blue) at that voltage with the same resistor ?

Single LED looks decent with 1k and USB (~5v).  Of course, it's not as bright, but in some cases that might be suitable.
1k was just for testing, it's probably a bit too much. It (1k) would be good if the thing was battery powered (might also be good if it's running in the dark).

[Zero999]

Finding the proper datasheet may be possible, but more times than not, you don't have it with the cheapest sources.

Some might work on 5v directly as intended.
For others, I guess that a step down can/could work.

A step down + some resistance might be the safest bet, but also the most complicated/costly. Not that a step down is very costly these days, just comparing.

What do you mean by step-down? Reducing the voltage? This is different to just adding a series resistor which limits the current.

Yes, if we wanted to get e.g. 3.x volts from USB, a step down seems like a reasonable method.

Like a voltage regulator? If so, that makes sense. You'll need something with a low dropout voltage, such as the TPS76201 but it comes in a small SMT package, which will make soldering a pain but is good because it's easy to hide.

http://www.ti.com/lit/ds/symlink/tps76201.pdf

I meant more along the lines of a buck step down module. They cost less than $0.5 shipped now, and you can set the output voltage with a trimmer. I think that if the source voltage falls or increases, it will affect the output voltage. But with USB, we expect to have a reasonably regulated voltage to begin with, so this should be all right. They claim a maximum (at a specific setting of course) efficiency of 96%.

Here's a short clip of what happens when 2 diodes are in parallel, but only 1 (1k in this case) resistor is used in series. So USB > 1k Resistor > 2 LEDs in parallel.  The test was done to see if we can see red with other colors at the same time with these. And the answer is yes as expected, at certain points blue becomes very dark.

https://streamable.com/3vdxf

The flickering is not visible with eyes.

How does that compare to one LED and a 1k8 to 2k2 resistor (half the current)? The eye is less sensitive to blue so it's often darker than the other colours but that can depend on the type of semiconductor used to make the LED too.

Regarding the step-down: a buck regulator is a good idea and will certainly be more efficient than the linear regulator I suggested but make sure the drop-out voltage isn't too high.