ughh so,
TLC outputs Voltage applied to output (OUT0–OUT15) max 17v per pin? My white LED has Vf of 2,8-3,4v @20mA RGB LED has 1,8-2,4V for RED and G.B. 2,9-3,6v @20mA
I measured resistance of longest 5m wire is 2ohm. Can't find any information on these no-brand wires.
So V=IxR = 0,02 x 2 = 0,04v
seems ok? or should i calculate the voltage drop from all the wires together? in that case for 1 cahnnel I have 300m of total wire,which adds up to 1200m for 4 channels (not counting anode channels). If 5m drops 0,04v, 1m = 0,008x1200 = 9,6v and I'm much more over 5v...
But these are common cathode RGB LEDs if I remember. So in the +V wire you could have 3 times that much current and so 3 times more voltage drop. Plus the voltage drop in the anode wires, so it could be around 0.16V in total.
PaulRB:
But these are common cathode RGB LEDs if I remember. So in the +V wire you could have 3 times that much current and so 3 times more voltage drop. Plus the voltage drop in the anode wires, so it could be around 0.16V in total.
yeah, you're right. And should I calculate that for all the total length of wires for all the LEDs in total? Because in total I have 300m per channel = (rgb 3cat. + 1an., white 1cat+1an.) = 300x6 1800meters (1,1miles)
But with that sort of length between TLC and LED you have a lot of inductance and you have to treat things like a transmission line problem.
This might involve pull up resistors at each end and a capacitor or two. You are going to need an oscilloscope to make sure the signals are correct even if you get functionality.
Grumpy_Mike:
But with that sort of length between TLC and LED you have a lot of inductance and you have to treat things like a transmission line problem.
with 5meters (16feet)? the length from TLC to LED is ranging 1-5meters (3,1-16feet). Like: TLC out0-3.1feet, out1-4 feet, out2- 6feet, out3-3feet etc.
1800meters (1,1miles) is when i count all totals of these out0-15 wires for all 25 TLCs, for all separate anode and catode wires together.
That looks amazing!! :o
But I tested 3m wire with no problems, it was one led in TLC row, worked fine along with others.. maybe because it was one..
And what about overall wire resistance? Should I add all sum of wires of all TLCs? or if they are in parallel it doesn't matter?
Did you use an oscilloscope? Like I said it might appear to function but not be reliable long term if you don’t.
Don’t add up any resistors just calculate the voltage drop from the source to the LED and don’t forget to include the resistance of the ground return wire.
Grumpy_Mike:
Don’t add up any resistors just calculate the voltage drop from the source to the LED and don’t forget to include the resistance of the ground return wire.
whoooo..... i was so worried.. so basically voltage drop is 0.16V for 5m and i can use 0.22mm wires for shorter wires like 1m i guess. And 1,5mm for longer 3-5m.
Grumpy_Mike:
Did you use an oscilloscope? Like I said it might appear to function but not be reliable long term if you don’t.
no.. maybe i could add 100pF cap by the LED as you suggest, seems very nice idea!
maybe i could add 100pF cap by the LED as you suggest, seems very nice idea!
No you should only do this if it is necessary, and you can only tell if it is with an oscilloscope.
Basically you are pushing what electronics are designed to do. In fact they are definitely not designed to do what you are trying.
It is not the voltage drop that is a problem, it is long leads acting like antenna for both emissions and susceptibility. And the transmission line effect with reflections from un / miss terminated lines.
Grumpy_Mike:
Basically you are pushing what electronics are designed to do. In fact they are definitely not designed to do what you are trying.
It is not the voltage drop that is a problem, it is long leads acting like antenna for both emissions and susceptibility. And the transmission line effect with reflections from un / miss terminated lines.
Sad to hear.. I really hope it does work..
So how else could I control LEDs from distance?
When electronics have to work over a distance it is normal to use a buffer on the control signals. The buffer turns a digital level signal into a differential pair. That is there are two wires, one is high and the other low at voltages at or above the normal 0 to 5V level.
This means at the receive end you have to turn this back into a normal digital signal with a buffer. Because the buffer is looking at the difference between two pins any interference picked up by the wires is cancelled out. Often twisted pairs of wires are used.
One example looks like this:-
This is for bidirectional communications needed by I2C but it shows the principle.
You could take this control signal and pass it to a string of WS2812 LEDs that is a long way away. If you had to use the TLC chip then you would need one of these buffer send / receive pairs for each signal. So that would not be as convenient, but would ensure that you had no problems from long distances. In the case of groups of LEDs you would only need one sender buffer and have a receive buffer on each group for common signals.
yeah.. thank you for information, very useful! this should have been the right way, unfortunately i have invested too much in this setup and time is against me, so I have to hope for best this time. On next project I will definitely use your suggestion!
So, I manage to draw scheme of my setup as it is now on breadboard.
It kind of works. Except for GPRS freezing loop().. I havent added yet 0,1uF capacitors to SN74HC04N as drawn, but i plan to. I hope it's a little better to understandable as previous..
Grumpy_Mike:
Waste of time trying it without them.
Actually, it works without them on breaboard. Maybe because I added 1uF and 100uF caps between +V -V railings between them. As VCC and GND are so apart I didn't find a way to physically add 0,1uF cap. As I understand it has to be as close as possible to the IC pins.
Grumpy_Mike:
the wiring of the inverters is a bit hard to follow.
yeah it's bad.. but the main thing is: BLANK, XLAT, SCLK, GSCLK goes through them. All empty inverter IN pins are connected to +V. Let's start soldering, yay!
ehhhh... i have soldered all and now testing boards. When I set #define NUM_TLCS 25 everything breaks.. kind of blinks for a second and then goes off.. Just this rgb test function..
It kind of works until i set NUM_TLCS to 16, then it lights some up, some not, and at 25 it's all gone...
what could be the problem? I'm testing to one TLC so frustrating..
The problem with just a snippet is that you can't tell where other things come from. For example the variable activeChannel seems to over ride the channel variable when it is zero. What is going on here.
I'm testing to one TLC
Don't know what that means, sorry.
Have you got the decoupling capacitors on each chip?