Hi!, I have an TLC5947 that I use to drive RBG led strips. My problem is that when I use the library to tell the IC to output 100% power (4095) the led strips don't glow as much as it should, Im sure i did something (or many things) wrong.
What I do is to give the TIP122 5v (with a 1K resistor) to the B, and I have the TLC5947 channel 0 connected to the E. The C goes to the ground (red channel) of the RGB strip and the +12v of the strip goes to 12v from the power supply.
- When I give channel 0 a pwm value of 4095, it glows but not as much as it's supposed to.
- Connecting my board's pin that usually goes to the channel 0 of the TLC5947 to ground makes RGB glow perfect (Connecting C to ground instead of the TLC5947 channel 0).
- ground of the TCL5947 is connected to the main ground just like the arduino.
ps: ignore the text bellow the image, lol.
What I'm I doing wrong?
Here's a basic schema of the circuit for one rgb strip channel. and some pictures of the board i made and the testing board with one transistor where it all works well.
I use tip122's to handle the 12vols.
I have the TLC5947 outputs conected to TIP122 transistors. My problem is that when I use the instruction: pwm_1.setPWM(0,0,0,4095)) the led strip for that channel should light up a lot and I dosen't.
TCL5947 - led strip not as bright as it should be?
Moderators: adafruit_support_bill, adafruit
Please be positive and constructive with your questions and comments.
- acanessa
- Posts: 18
- Joined: Fri Oct 24, 2014 2:45 pm
- adafruit_support_mike
- Posts: 67485
- Joined: Thu Feb 11, 2010 2:51 pm
Re: TCL5947 - led strip not as bright as it should be?
Hmm.. what you have now is an elegant solution to a problem that doesn't actually exist.
The circuit you've created is called a 'cascode', and is an excellent way to make one transistor protect another from overvoltage:
For the sake of talking through the ideas, assume:
- Q1 is smart but weak.. it's connected to a circuit that does something interesting, but can only operate between 0v and 5v.
- Q2 is dumb but strong.. its base is connected to a fixed reference voltage, but it can operate from 0v to 100v.
- We want to control a load that operates outside Q1's range, but inside Q2's range.
Under those conditions, a cascode is the right tool for the job. The voltage at Q2's emitter will always be about 0.65v below the voltage at its base, and you can put the base voltage wherever you want. Q1 will see Q2's emitter as a current source operating at a voltage Q1 can handle, and Q1 will control that current.
Your cascode's reference voltage is a bit too high.
The TIP122 is a Darlington transistor, so its emitter voltage will be about 2.5v below its base voltage. At the same time, its collector voltage will be about 2v higher than its emitter voltage. Putting those together, the collector will be about half a volt below the base.
If you take the base voltage to 5v, the collector voltage will be about 4.5v. That leaves 7.5v for the LED strip, or 2.5v per LED (assuming you have three LEDs in series). If you're trying to run blue or white LEDs, that will restrict the headroom pretty badly.
You could change the TIP122's base voltage to give the LEDs more headroom, but you don't need to..
The TCL5847's output pins can handle load voltages up to 30v.
The TLC5947's output pins don't need to be protected from a 12v supply, so in this case a cascode won't do anything useful. Just connect the LED strip directly to the TLC5947 and things will be fine.
The circuit you've created is called a 'cascode', and is an excellent way to make one transistor protect another from overvoltage:
For the sake of talking through the ideas, assume:
- Q1 is smart but weak.. it's connected to a circuit that does something interesting, but can only operate between 0v and 5v.
- Q2 is dumb but strong.. its base is connected to a fixed reference voltage, but it can operate from 0v to 100v.
- We want to control a load that operates outside Q1's range, but inside Q2's range.
Under those conditions, a cascode is the right tool for the job. The voltage at Q2's emitter will always be about 0.65v below the voltage at its base, and you can put the base voltage wherever you want. Q1 will see Q2's emitter as a current source operating at a voltage Q1 can handle, and Q1 will control that current.
Your cascode's reference voltage is a bit too high.
The TIP122 is a Darlington transistor, so its emitter voltage will be about 2.5v below its base voltage. At the same time, its collector voltage will be about 2v higher than its emitter voltage. Putting those together, the collector will be about half a volt below the base.
If you take the base voltage to 5v, the collector voltage will be about 4.5v. That leaves 7.5v for the LED strip, or 2.5v per LED (assuming you have three LEDs in series). If you're trying to run blue or white LEDs, that will restrict the headroom pretty badly.
You could change the TIP122's base voltage to give the LEDs more headroom, but you don't need to..
The TCL5847's output pins can handle load voltages up to 30v.
The TLC5947's output pins don't need to be protected from a 12v supply, so in this case a cascode won't do anything useful. Just connect the LED strip directly to the TLC5947 and things will be fine.
- acanessa
- Posts: 18
- Joined: Fri Oct 24, 2014 2:45 pm
Re: TCL5947 - led strip not as bright as it should be?
Thank you very much for your reply!
I think I get what you mean (I could be wrong ha!). Also I was thinking, if in my circuit the TCL channel was the "ground" for the Darlington AND the led strip, when I am applying the concept of PWM, I am limiting the voltage for the transistor, but ain't I also, sort of preventing the led strip V+ to go fully to ground? Is that why I am not getting full brightness?
Just to be sure (as I'm sort of new in the electronics area) you say that the breakout/IC can handle up to 30v. My plan is to have 8 RGB strips of about 2 meters long each for each channel. I have a power supply more than capable to handle that, but my (perhaps ignorant) question is;
- Can the TCL handle the amps for something like that? What you say is to connect 2 meters of RGB led strip directly into the TCL? It sound too good to be true? :P
Also, another question; If I put the RGB led strip directly into the breakout, I should use 12v as the input for the breakout, is this correct? If so, is this also correct?;
V+ breakout -> 12v PSU
GND breakout -> GND PSU
SPI -> SPI
But the LEARN page for the TCL states:
"If you need V+ voltages higher than 12v, you will need to use a separate power supply. If you use a separate supply, be sure to connect the ground wire to the Arduino ground." So, I should not only connect the breakout GND to the PSU GND, but also the breakout GND to the arduino GND? As in;
V+ breakout -> 12v PSU
GND breakout -> GND PSU
GND breakout (also to) -> Arduino GND
SPI -> SPI
Lastly, if it is not too much trouble, can you provide me with a diagram so that I am extra sure that what I am doing is correct?
Thank you very much for you help!!!! :)
Augusto
I think I get what you mean (I could be wrong ha!). Also I was thinking, if in my circuit the TCL channel was the "ground" for the Darlington AND the led strip, when I am applying the concept of PWM, I am limiting the voltage for the transistor, but ain't I also, sort of preventing the led strip V+ to go fully to ground? Is that why I am not getting full brightness?
Just to be sure (as I'm sort of new in the electronics area) you say that the breakout/IC can handle up to 30v. My plan is to have 8 RGB strips of about 2 meters long each for each channel. I have a power supply more than capable to handle that, but my (perhaps ignorant) question is;
- Can the TCL handle the amps for something like that? What you say is to connect 2 meters of RGB led strip directly into the TCL? It sound too good to be true? :P
Also, another question; If I put the RGB led strip directly into the breakout, I should use 12v as the input for the breakout, is this correct? If so, is this also correct?;
V+ breakout -> 12v PSU
GND breakout -> GND PSU
SPI -> SPI
But the LEARN page for the TCL states:
"If you need V+ voltages higher than 12v, you will need to use a separate power supply. If you use a separate supply, be sure to connect the ground wire to the Arduino ground." So, I should not only connect the breakout GND to the PSU GND, but also the breakout GND to the arduino GND? As in;
V+ breakout -> 12v PSU
GND breakout -> GND PSU
GND breakout (also to) -> Arduino GND
SPI -> SPI
Lastly, if it is not too much trouble, can you provide me with a diagram so that I am extra sure that what I am doing is correct?
Thank you very much for you help!!!! :)
Augusto
- acanessa
- Posts: 18
- Joined: Fri Oct 24, 2014 2:45 pm
Re: TCL5947 - led strip not as bright as it should be?
UPDATE:
OK, so I gave it a try. I conected the PSU the arduino and the TLC like this:
V+ breakout -> 12v PSU
GND breakout -> GND PSU
GND breakout (also to) -> Arduino GND
SPI -> SPI
VIN Arduino -> 5v
GND Arduino -> GND PSU
RGB Strip 12+ -> 12v from TLC board
RGB Strip Color -> TLC Channel 7
On the picture, red cable is 5v, orange is 12v, black ground. Led strip 12v goes to the orange 12v and color goes to the pins you see conectes to rgb "channel 7".
On the sketch:
tlc.setLED(7, 4095, 4095, 4095);
Result: The led strip turns on, but still NOT AS BRIGHT.
It's the same result as using the TIPs
What I am doing wrong? Thank you and sorry for the basic questions!!
OK, so I gave it a try. I conected the PSU the arduino and the TLC like this:
V+ breakout -> 12v PSU
GND breakout -> GND PSU
GND breakout (also to) -> Arduino GND
SPI -> SPI
VIN Arduino -> 5v
GND Arduino -> GND PSU
RGB Strip 12+ -> 12v from TLC board
RGB Strip Color -> TLC Channel 7
On the picture, red cable is 5v, orange is 12v, black ground. Led strip 12v goes to the orange 12v and color goes to the pins you see conectes to rgb "channel 7".
On the sketch:
tlc.setLED(7, 4095, 4095, 4095);
Result: The led strip turns on, but still NOT AS BRIGHT.
It's the same result as using the TIPs
What I am doing wrong? Thank you and sorry for the basic questions!!
- adafruit_support_mike
- Posts: 67485
- Joined: Thu Feb 11, 2010 2:51 pm
Re: TCL5947 - led strip not as bright as it should be?
Yep that's it.acanessa wrote:but ain't I also, sort of preventing the led strip V+ to go fully to ground? Is that why I am not getting full brightness?
In most circuits, the two things you need to care about are voltage and current. You can connect up to 30v to one of the TLC5947's LED control pins safely, and the circuits inside the chip control how much current will flow through the LEDs and the control pin when the pin is active.. usually about 30mA, which is enough to light most LEDs brightly.acanessa wrote:Can the TCL handle the amps for something like that? What you say is to connect 2 meters of RGB led strip directly into the TCL? It sound too good to be true? :P
Blue LEDs want about 3-1/2v to operate, so from a 12v supply you can stack three of them in series and control them with a single TLC5947 pin.
No, you want to connect the TLC5947's VCC pin to 5v, like you have it now.acanessa wrote:Also, another question; If I put the RGB led strip directly into the breakout, I should use 12v as the input for the breakout, is this correct?
The pins that control the LEDs are a lot stronger than the parts that handle the logic and pulse generation.
You have the basic idea. Connect the 12v supply's positive connection to the VCC end of your LED strip and its negative connection to the Arduino/TLC5947's GND.acanessa wrote:So, I should not only connect the breakout GND to the PSU GND, but also the breakout GND to the arduino GND?
Sure:acanessa wrote:Lastly, if it is not too much trouble, can you provide me with a diagram so that I am extra sure that what I am doing is correct?
- acanessa
- Posts: 18
- Joined: Fri Oct 24, 2014 2:45 pm
Re: TCL5947 - led strip not as bright as it should be?
Thank you for your reply!
So, some things are clearer now. But, I feel there is still something I am missing here:
You said: "The TLC5947's output pins don't need to be protected from a 12v supply, so in this case a cascode won't do anything useful. Just connect the LED strip directly to the TLC5947 and things will be fine."
What I interpreted by that was:
PSU 12v -> TLC V+
PSU GND -> TLC GND
Arduino VIN -> PSU 5v
Arduino GND -> PSU GND
RGB strip 12V -> TLC V+ channel 7
RGB strip GND -> TLC "PWM" channel 7
This gave me the same result, RGB strip not as bright as expected.
If I understand correctly what you say now is;
- Only 5v to V+ pin on "V+,GND,DIN,CLK/,OE,LAT" row in TCL.
- GND of TLC to common ground with arduino / PSU
The thing is, I did this. After I removed the TIP 122's following you advice, I did this. (of course, I might have misunderstand you!) Your diagram makes absolute sense. But, again, on the latests set-ups, I did that....after testing I came to the conclusion that for some reason the TLC was not letting the 12 volts from the RGB strip fully go to ground. When I was using the TIP's I thought that was the problem, but having removed them, the problem remains.
I still don't get why the RGB strip does not glow fully, in your diagram, when the channel is fully open to ground (4095) all 12v from the RGB strip should flow via the channel to the GND pin and ergo, to the common ground where the PSU GND is connected...
When I had the TIP122 transistor alone, connected via 1K to 5v and the led strip 12+ connected to 12v and GND to the TIP122, and I connected the TIP122 3rd pin to ground, the RGB strip was at FULL BRIGHTNESS
Am I making sense? Please let me know what you think, and I will send you a picture as clear as I can of the physical set-up. But I am 99% that I tried your design and the leds still didn't glow fully.
Thank you for your patience.
So, some things are clearer now. But, I feel there is still something I am missing here:
You said: "The TLC5947's output pins don't need to be protected from a 12v supply, so in this case a cascode won't do anything useful. Just connect the LED strip directly to the TLC5947 and things will be fine."
What I interpreted by that was:
PSU 12v -> TLC V+
PSU GND -> TLC GND
Arduino VIN -> PSU 5v
Arduino GND -> PSU GND
RGB strip 12V -> TLC V+ channel 7
RGB strip GND -> TLC "PWM" channel 7
This gave me the same result, RGB strip not as bright as expected.
If I understand correctly what you say now is;
- Only 5v to V+ pin on "V+,GND,DIN,CLK/,OE,LAT" row in TCL.
- GND of TLC to common ground with arduino / PSU
The thing is, I did this. After I removed the TIP 122's following you advice, I did this. (of course, I might have misunderstand you!) Your diagram makes absolute sense. But, again, on the latests set-ups, I did that....after testing I came to the conclusion that for some reason the TLC was not letting the 12 volts from the RGB strip fully go to ground. When I was using the TIP's I thought that was the problem, but having removed them, the problem remains.
I still don't get why the RGB strip does not glow fully, in your diagram, when the channel is fully open to ground (4095) all 12v from the RGB strip should flow via the channel to the GND pin and ergo, to the common ground where the PSU GND is connected...
When I had the TIP122 transistor alone, connected via 1K to 5v and the led strip 12+ connected to 12v and GND to the TIP122, and I connected the TIP122 3rd pin to ground, the RGB strip was at FULL BRIGHTNESS
Am I making sense? Please let me know what you think, and I will send you a picture as clear as I can of the physical set-up. But I am 99% that I tried your design and the leds still didn't glow fully.
Thank you for your patience.
- acanessa
- Posts: 18
- Joined: Fri Oct 24, 2014 2:45 pm
Re: TCL5947 - led strip not as bright as it should be?
UPDATE:
what if I use a TIP122 like I did at the beginning, but instead of NPN, PNP? I suspect that was my original problem using the TIP122, that they were NPN and not PNP, am I correct?
I still don't get why without the TIP's the LED strip don't glow as much as they should....
what if I use a TIP122 like I did at the beginning, but instead of NPN, PNP? I suspect that was my original problem using the TIP122, that they were NPN and not PNP, am I correct?
I still don't get why without the TIP's the LED strip don't glow as much as they should....
- adafruit_support_mike
- Posts: 67485
- Joined: Thu Feb 11, 2010 2:51 pm
Re: TCL5947 - led strip not as bright as it should be?
Oh ugh.. dumb mistake on my part: you're talking about an LED *strip* and I was just reading "LEDs".
That changes the rules of the game. The TLC5947 is what's called a 'constant current LED driver', which means it only allows a certain amount of current to flow from its LED control pins to GND.. usually about 60mA, but you can adjust the exact value. That's great if you're controlling a single LED or a set of LEDs in series. It doesn't work so well for analog LED strips because those have serveral rows of LEDs in parallel.
For that kind of circuit, you do need a Darlington or mosfet that can handle more current than the TLC5947.
The trouble is that the TLC5947 isn't very good at controlling the kind of Darlington or mosfet you'd need in this case.
NPN transistors, N-type Darlingtons, and N-mosfets all allow current to flow when their input voltage is positive (above a certain threshold). When the input voltage is zero (or below the threshold), they prevent current from flowing.
The TLC5947's LED control pins can't provide a positive voltage. Their operating states are "very high resistance" and "current-controlled path to GND". They'd be great a turning N-type transistors off, but that isn't what you want here.
Thing is, you probably need N-type transistors to control the LED strip because it has a single connection to 12v and separate connections to GND for the red, green, and blue channels.
To make everything work, you need a few PNP transistors.
PNPs, P-type Darlingtons, and P-mosfets all allow current to flow when their control voltage is negative. The TLC5947 provides exactly the kind of signals those want.
The circuit you want looks like this:
When the TLC5947's control pin opens a connection to GND, it will turn the PNP transistor on. That will allow current to flow to the base of the N-type Darlington, turning it on. That will allow a whole lot of current to flow through the LED strip. When the TLC5947's control pin closes the connection to GND, it cuts off the current to the PNP, which shuts off the Darlington, which shuts off the LED strip.
The PNP transistor doesn't have to carry a lot of current. Darlingtons have huge current gain, so a couple of milliamps to the base will control a couple of amps on the main current path. I listed a 2N3906 PNP transistor because that's one of my favorites (they've been around since the 1950s, work really well, and are cheap as dirt), but any small-signal PNP will do.
The TIP122's current gain is 1000, so to control 2A through the LED strip you'd need 2mA at its base. Its base voltage is about 2.5v when it's on, so a 1k resistor with a couple volts across it will provide about the right amount of control current.
I listed the resistor between the TLC5947 and the PNP as 1k too, but it can be quite a bit larger. The 2N3906's current gain is between 150 and 200 in the 1-10mA range, so you could go up to 100k and still have the circuit work.
I didn't draw the power supplies in this diagram, but you want to connect their GND lines as before. The Darlington needs connections to both.. the control current from the PNP needs to flow back to the 5v power supply and the current through the LED strip needs to flow back to the 12 power supply.
That changes the rules of the game. The TLC5947 is what's called a 'constant current LED driver', which means it only allows a certain amount of current to flow from its LED control pins to GND.. usually about 60mA, but you can adjust the exact value. That's great if you're controlling a single LED or a set of LEDs in series. It doesn't work so well for analog LED strips because those have serveral rows of LEDs in parallel.
For that kind of circuit, you do need a Darlington or mosfet that can handle more current than the TLC5947.
The trouble is that the TLC5947 isn't very good at controlling the kind of Darlington or mosfet you'd need in this case.
NPN transistors, N-type Darlingtons, and N-mosfets all allow current to flow when their input voltage is positive (above a certain threshold). When the input voltage is zero (or below the threshold), they prevent current from flowing.
The TLC5947's LED control pins can't provide a positive voltage. Their operating states are "very high resistance" and "current-controlled path to GND". They'd be great a turning N-type transistors off, but that isn't what you want here.
Thing is, you probably need N-type transistors to control the LED strip because it has a single connection to 12v and separate connections to GND for the red, green, and blue channels.
To make everything work, you need a few PNP transistors.
PNPs, P-type Darlingtons, and P-mosfets all allow current to flow when their control voltage is negative. The TLC5947 provides exactly the kind of signals those want.
The circuit you want looks like this:
When the TLC5947's control pin opens a connection to GND, it will turn the PNP transistor on. That will allow current to flow to the base of the N-type Darlington, turning it on. That will allow a whole lot of current to flow through the LED strip. When the TLC5947's control pin closes the connection to GND, it cuts off the current to the PNP, which shuts off the Darlington, which shuts off the LED strip.
The PNP transistor doesn't have to carry a lot of current. Darlingtons have huge current gain, so a couple of milliamps to the base will control a couple of amps on the main current path. I listed a 2N3906 PNP transistor because that's one of my favorites (they've been around since the 1950s, work really well, and are cheap as dirt), but any small-signal PNP will do.
The TIP122's current gain is 1000, so to control 2A through the LED strip you'd need 2mA at its base. Its base voltage is about 2.5v when it's on, so a 1k resistor with a couple volts across it will provide about the right amount of control current.
I listed the resistor between the TLC5947 and the PNP as 1k too, but it can be quite a bit larger. The 2N3906's current gain is between 150 and 200 in the 1-10mA range, so you could go up to 100k and still have the circuit work.
I didn't draw the power supplies in this diagram, but you want to connect their GND lines as before. The Darlington needs connections to both.. the control current from the PNP needs to flow back to the 5v power supply and the current through the LED strip needs to flow back to the 12 power supply.
- acanessa
- Posts: 18
- Joined: Fri Oct 24, 2014 2:45 pm
Re: TCL5947 - led strip not as bright as it should be?
Hey, thanks for your reply!
OK, now it all makes a lot more sense :) I knew darlingtons had to be part of the equation and I had the feeling that I should go PNP somewhere. I could not read your reply until now, but yesterday I did some testing with TIP 127 (or 177? can't remember now) PNP and I did manage to have it all work, but your circuit makes a lot more sense than the one I made, that did not include the 2N3906.
Thank you for your help, I will give all this a try and let you know! :)
OK, now it all makes a lot more sense :) I knew darlingtons had to be part of the equation and I had the feeling that I should go PNP somewhere. I could not read your reply until now, but yesterday I did some testing with TIP 127 (or 177? can't remember now) PNP and I did manage to have it all work, but your circuit makes a lot more sense than the one I made, that did not include the 2N3906.
Thank you for your help, I will give all this a try and let you know! :)
- acanessa
- Posts: 18
- Joined: Fri Oct 24, 2014 2:45 pm
Re: TCL5947 - led strip not as bright as it should be?
I tried the circuit and now it all works perfect! Thanks a lot for the help!
- acanessa
- Posts: 18
- Joined: Fri Oct 24, 2014 2:45 pm
Re: TCL5947 - led strip not as bright as it should be?
Hi, there is another question that I have, maybe you can help me?
The circuit you gave me works fantastic. There is one thing tough, I've found that some times, well often, when I am chancing colours by sending information the the TLC, an unexpected colour would appear. I did some testing, and found some behaviours:
- turning on/off the A/C will make the TLC change colours.
- turning on/off the fridge will make the TLC change colours.
- if TLC is set to all channels off, turning some mayor appliance will make it turn on the channels.
You can imagine where this is going...I'm far from knowing about this, but...I have reached the conclusion that my PSU (ATX PSU) is susceptible to external changes in the AC power, and some how is introducing some sort of ripple noise in the TLC signal and maybe corrupting the data?
Also, I have a couple of DS18B20 temperature sensors, and every time I read them, the TLC turns on a bit the led strips.
What do you think? is the a way you can recommend to resolve this? I've heard of capacitors in the power lines? Can your point me to some literature that might help me regarding this problem? thank you!!
The circuit you gave me works fantastic. There is one thing tough, I've found that some times, well often, when I am chancing colours by sending information the the TLC, an unexpected colour would appear. I did some testing, and found some behaviours:
- turning on/off the A/C will make the TLC change colours.
- turning on/off the fridge will make the TLC change colours.
- if TLC is set to all channels off, turning some mayor appliance will make it turn on the channels.
You can imagine where this is going...I'm far from knowing about this, but...I have reached the conclusion that my PSU (ATX PSU) is susceptible to external changes in the AC power, and some how is introducing some sort of ripple noise in the TLC signal and maybe corrupting the data?
Also, I have a couple of DS18B20 temperature sensors, and every time I read them, the TLC turns on a bit the led strips.
What do you think? is the a way you can recommend to resolve this? I've heard of capacitors in the power lines? Can your point me to some literature that might help me regarding this problem? thank you!!
- acanessa
- Posts: 18
- Joined: Fri Oct 24, 2014 2:45 pm
Re: TCL5947 - led strip not as bright as it should be?
Additionally, I see on the datasheet that there is a capacitor between V+ and GND, is this a decoupling capacitor? is this to help filtering noise from the power line?
if what I am saying makes some sense (doubtful ha ha), what capacitor should I use?
Thank you!
And....i see on the board, a "opt. V+ cap" space....is this where the decoupling capacitor should go?
if what I am saying makes some sense (doubtful ha ha), what capacitor should I use?
Thank you!
- acanessa
- Posts: 18
- Joined: Fri Oct 24, 2014 2:45 pm
Re: TCL5947 - led strip not as bright as it should be?
Hey, I saw this post just now. (viewtopic.php?f=19&t=52569&p=265741&hil ... or#p265622) "TLC5947 - what capacitor for the reserved space on the board"
- So, I guess...i should use a capacitor to filter out the noise from the power line?
- If that's the case, I saw in above post that the capacitor is used together with the "Optional V+"....? I am not clear about that....are those the holes in the back of the board? - When should I use that "optional V+"?
- If I install the capacitor in the board without using the "optional V+", will the capacitor still filter noise coming from the PSU to the V+ standard pin on the beard?
I also saw this: "I've found an answer in a LED's forum.. Connect the /OE to the clock and the flicker disappears." (http://forums.adafruit.com/viewtopic.php?f=47&p=259745)
Thank you again and sorry for the bombardment of questions!
- So, I guess...i should use a capacitor to filter out the noise from the power line?
- If that's the case, I saw in above post that the capacitor is used together with the "Optional V+"....? I am not clear about that....are those the holes in the back of the board? - When should I use that "optional V+"?
- If I install the capacitor in the board without using the "optional V+", will the capacitor still filter noise coming from the PSU to the V+ standard pin on the beard?
I also saw this: "I've found an answer in a LED's forum.. Connect the /OE to the clock and the flicker disappears." (http://forums.adafruit.com/viewtopic.php?f=47&p=259745)
Thank you again and sorry for the bombardment of questions!
- adafruit_support_mike
- Posts: 67485
- Joined: Thu Feb 11, 2010 2:51 pm
Re: TCL5947 - led strip not as bright as it should be?
Yep.. the problem you're having is clearly related to noise in the power lines, and a cap will remove that noise.acanessa wrote:So, I guess...i should use a capacitor to filter out the noise from the power line?
"Optional V+" means you connect a power supply other than the Arduino's Vin pin to the breakout's V+ pin.acanessa wrote:If that's the case, I saw in above post that the capacitor is used together with the "Optional V+"....? I am not clear about that....are those the holes in the back of the board? - When should I use that "optional V+"?
Yes. The cap will smooth out ripples no matter what power supply you use.acanessa wrote:If I install the capacitor in the board without using the "optional V+", will the capacitor still filter noise coming from the PSU to the V+ standard pin on the beard?
There's only one problem with that: the TLC5947 doesn't have an /OE pin.acanessa wrote:I also saw this: "I've found an answer in a LED's forum.. Connect the /OE to the clock and the flicker disappears." (http://forums.adafruit.com/viewtopic.php?f=47&p=259745)
- acanessa
- Posts: 18
- Joined: Fri Oct 24, 2014 2:45 pm
Re: TCL5947 - led strip not as bright as it should be?
Thank you for your reply. Indeed, the PSU is the problem. Is an ATX PSU that I made into a workbench PSU. I solved the problem by:
1) Connecting the power pins of the TLC to the arduino instead of the PSY 5v / common GND. Still had noise, but much less.
2) Used the Cap on the board. This apparently had no effect, but I think the noise from the PSU was too much for the cap, that's why perhaps? I used a 10uF 25v and a 100uF 25v.
3) Connected the PSU power in (220v) to a UPS. This took away 95% of the noise.
Now it all works perfect.
Thank you for your help! I love Adafruit's products. You can't find them in my country (Argentina) and you can't buy them on the internet. (Our government sucks) so I asked a friend that lives in the States that was visiting to buy the lot and bring them over! I have learned a lot since then. :)
1) Connecting the power pins of the TLC to the arduino instead of the PSY 5v / common GND. Still had noise, but much less.
2) Used the Cap on the board. This apparently had no effect, but I think the noise from the PSU was too much for the cap, that's why perhaps? I used a 10uF 25v and a 100uF 25v.
3) Connected the PSU power in (220v) to a UPS. This took away 95% of the noise.
Now it all works perfect.
Thank you for your help! I love Adafruit's products. You can't find them in my country (Argentina) and you can't buy them on the internet. (Our government sucks) so I asked a friend that lives in the States that was visiting to buy the lot and bring them over! I have learned a lot since then. :)
Please be positive and constructive with your questions and comments.