Austin, TX
Offline
Faraday Member
Karma: 42
Posts: 5244
CMiYC
|
 |
« Reply #15 on: March 31, 2012, 11:15:56 am » |
What I thought when I said Would using the 12v adapter with a 7805 to power the TLC's cause more glitching, since i'd also be powering the led's with it was that the leds wil be going on and off a lot, so i thought that might interfere with other circuits running from the same 12v source.. or won't that be a problem? That's where decoupling capacitors can help. By the way, would, theoretically, adding 1000 capacitors to a power source, change anything about the output? Adding 1000 capacitors doesn't matter. What matters is how much capacitance you have added (or the value of each of those 1000). You need to strike a balance between having enough and not having too much. If you have too much, there will be a large amount of in-rush current when your circuit turns on, as all of the capacitance is charged. (Too much has a lot of "depends" in it. However, thousands of microfarads is probably a good line to watch.) A 0.1uF cap near each IC's VCC is helpful to smooth out high frequency demands. Sometimes, it may be necessary to put something larger in parallel, like a 1.0uF. FYI, Even if you have 20 of these caps in your design, that's still less than half the capacitance as the bulk decoupling caps on the Arduino board itself.
|
|
|
|
|
Logged
|
|
|
|
|
Offline
Newbie
Karma: 0
Posts: 48
|
 |
« Reply #16 on: March 31, 2012, 11:58:24 am » |
Sometimes, it may be necessary to put something larger in parallel, Aren't decoupling capacitors always parallel? Or is that the mistake i'm making?
|
|
|
|
|
Logged
|
|
|
|
|
Manchester (England England)
Offline
Brattain Member
Karma: 299
Posts: 26024
Solder is electric glue
|
 |
« Reply #17 on: March 31, 2012, 12:19:34 pm » |
It sounds like you are not sure what the decoupling capacitors are doing for you. Please read this:- http://www.thebox.myzen.co.uk/Tutorial/De-coupling.htmlAren't decoupling capacitors always parallel? Yes
|
|
|
|
|
Logged
|
|
|
|
|
Offline
Newbie
Karma: 0
Posts: 48
|
 |
« Reply #18 on: March 31, 2012, 02:20:49 pm » |
I'm now using a different power supply for the leds, and i encountered something odd.
When I change the voltage to a about 1.5v more, or 1.5v less than 12v, there is a lot less glitching.. Does this bring any more light to the problem?
|
|
|
|
|
Logged
|
|
|
|
|
Manchester (England England)
Offline
Brattain Member
Karma: 299
Posts: 26024
Solder is electric glue
|
 |
« Reply #19 on: March 31, 2012, 02:30:01 pm » |
Does this bring any more light to the problem? Not really, it just points up that it is a decoupling issue. It is probably that the regulation on the power supply is different for the different voltages enabling the existing decoupling to cope with the interference.
|
|
|
|
|
Logged
|
|
|
|
|
Offline
Newbie
Karma: 0
Posts: 48
|
 |
« Reply #20 on: March 31, 2012, 03:03:50 pm » |
Can you explain that in a bit easier to understand English? My English is not that good.
Note that this is only the power supply to the anodes of the leds, it's not powering the arduino/tlc's and it's only connected to each other through the ground
|
|
|
|
|
Logged
|
|
|
|
|
Manchester (England England)
Offline
Brattain Member
Karma: 299
Posts: 26024
Solder is electric glue
|
 |
« Reply #21 on: March 31, 2012, 03:09:31 pm » |
It is probably that the regulation on the power supply is different for the different voltages The regulation on a supply is a measure of how stable the voltage is as more or less current is drawn from it, and how much ripple there is on top of that voltage. If your power supply has a variable voltage it is likely that the change in voltage as the current changes is going to be different at different voltages. It is likely that at 12V this is worse on your power supply than it is a few volts either side of it. This is not unusual for a variable power supply.
|
|
|
|
|
Logged
|
|
|
|
|
Austin, TX
Offline
Faraday Member
Karma: 42
Posts: 5244
CMiYC
|
 |
« Reply #22 on: March 31, 2012, 03:52:52 pm » |
What is the current rating of your suppiles? And how much current do you expect the LEDs to draw?
|
|
|
|
|
Logged
|
|
|
|
|
Offline
Newbie
Karma: 0
Posts: 48
|
 |
« Reply #23 on: March 31, 2012, 04:57:36 pm » |
The regulation on a supply is a measure of how stable the voltage is as more or less current is drawn from it, and how much ripple there is on top of that voltage.
If your power supply has a variable voltage it is likely that the change in voltage as the current changes is going to be different at different voltages. It is likely that at 12V this is worse on your power supply than it is a few volts either side of it. This is not unusual for a variable power supply. So that would mean the glitching comes from the led power source, and thus I need to decouple that power source instead of the TLC's?
|
|
|
|
|
Logged
|
|
|
|
|
Manchester (England England)
Offline
Brattain Member
Karma: 299
Posts: 26024
Solder is electric glue
|
 |
« Reply #24 on: March 31, 2012, 05:20:53 pm » |
and thus I need to decouple that power source instead of the TLC's? No the TLCs are where the damage is being done, decouple them tightly and don't let the interference in.
|
|
|
|
|
Logged
|
|
|
|
|
Salem, Oregon
Offline
Full Member
Karma: 0
Posts: 175
Arduino rocks but sometimes it rolls
|
 |
« Reply #25 on: March 31, 2012, 10:27:53 pm » |
[ That's probably a bad idea. It would be better to use something like a 7805 fed from 12v and then passed to VIN on the arduino and TLC5940's. Don't misunderstand how the TLC5940 operates. It does not draw current from its Vcc to power the LEDs. It controls the amount of current allowed to flow through its outputs. All of the "high current" comes from whatever supply the LEDs are connected to, not the TLC5940. There's no reason to add another regulator when the Arduino's on-board regulator (or USB supply) can handle powering dozens of TLC5940s. James. Thanks for the correction. I was under the impression from what I've read on-line that several TLC5940's would draw too much power from the Arduino (not to power the LEDs but for the chips themselves). Since the boards I always use aren't genuine Arduino's, I power the TLC5940's from the same 5v source that drives the boards themselves and just assumed that was correct. Thanks, Brad.
|
|
|
|
|
Logged
|
|
|
|
|
Austin, TX
Offline
Faraday Member
Karma: 42
Posts: 5244
CMiYC
|
 |
« Reply #26 on: March 31, 2012, 11:05:45 pm » |
James. Thanks for the correction. I was under the impression from what I've read on-line that several TLC5940's would draw too much power from the Arduino (not to power the LEDs but for the chips themselves). You can't always believe what you read online. Most people don't understand how LEDs work, let alone the TLC5940. You have to look to the data sheet to understand.
|
|
|
|
|
Logged
|
|
|
|
|
Offline
Newbie
Karma: 0
Posts: 48
|
 |
« Reply #27 on: April 01, 2012, 11:07:35 am » |
I'm still waiting for the new TLC's to arrive. In the meanwhile I encountered something interesting. When I use the power source from the led's to power a walkietalkie, and broadcast with it, I can hear a noise-frequencie which matches the pattern of the leds.. Is there anything useful i can do with that? 
|
|
|
|
|
Logged
|
|
|
|
|
Offline
Newbie
Karma: 0
Posts: 48
|
 |
« Reply #28 on: April 02, 2012, 04:16:52 am » |
I've received new TLC's and tried most of the suggestions, but there's still glitching:
|
|
|
|
|
Logged
|
|
|
|
|
Offline
Newbie
Karma: 0
Posts: 48
|
 |
« Reply #29 on: April 02, 2012, 05:13:18 am » |
Stuff is acting even weirder: void loop() { for (int channel = 0; channel < NUM_TLCS * 16; channel++) { Tlc.set(channel, 4095); Tlc.update(); } delay(1000); for (int channel = 0; channel < NUM_TLCS * 16; channel++) { Tlc.set(channel, 0); Tlc.update(); } delay(1000); }
|
|
|
|
« Last Edit: April 02, 2012, 05:41:32 am by Inevitableavoidance »
|
Logged
|
|
|
|
|
|