Go Down

Topic: TLC5940/Uno interface problem (Read 8733 times) previous topic - next topic

Av8ter

Here are new images.  I'm going to take your advice and tear everything down tomorrow and re-wire it with different jumper wires.  I'll post how it goes.  Thanks again.

raschemmel

you know what ? I think I know what the problem is ? Did you actually put a meter on pin 21 to verify +5V ?
You might have one of those breadboards that don't have continuity all the way across the bus. You have to put little jumpers in the spaces between the gourps on the bus. I would get a meter an take measurements in different places, starting with pin-21.
Put a meter ground where the arduino ground connects to the bus and move the positive meter lead to various points to measure the voltage. It's a long shot but it's worth checking. If you've tried more than one chip then there is something wrong with your setup because it can't be software. When you wired it up did you use a checklist to check off each connection ? Do you have any other leds (ordinary red ) that you can substitute before you tear down the circuit just to rule out the RGB leds ?

Av8ter

Well, it wasn't the board, it was about how I had misused the board.  I had checked and quadruple checked the board wiring and overlooked my mistake each time.  I had the 2K2 resistor on pin 19 going to +5V.  It should have gone to GND, thus,  I think I fried all three TLC5940s.  I really appreciate all your time and effort.  I'm sure I will need guidance when I get into the coding.  Thanks again.

raschemmel

#18
Mar 24, 2014, 03:02 pm Last Edit: Mar 24, 2014, 03:23 pm by raschemmel Reason: 1
that shouldn't hurt them
see page 8 of datasheet ; equilent circuit for Viref.  Pin 20 is the V- input of an op-amp.

Av8ter

My mistake.  I should have said that I had pin 20 (IREF) going to +5V instead of GND as it should.  Wouldn't the reverse polarity damage the op amp.  Pin 19 is to +5V correctly.  The board checked out correctly with 4.98V to all  the proper points.

raschemmel

#20
Mar 24, 2014, 08:49 pm Last Edit: Mar 24, 2014, 09:02 pm by raschemmel Reason: 1
No. The I ref and V ref are based on a resistor connected from V- of the op amp to GND. ( in the schematic on page 8. ) So the voltage reference for the op amp that controls the current is provided by the voltage drop across the Rref (the 2k ). So as I pointed  out before , the WORSE CASE is NO RESISTOR to GND, bringing the voltage to lowest point. The Opposite of  WORST CASE is 2K RESISTOR TO +5V which should limit the current to ZERO . (I think; that's just a guess). You know that NO current was SUNK by LEDS because they NEVER LIT UP. IF THEY HAD ALL LIT UP SUPER BRIGHT FOR AN INSTANT AND THEN GONE DARK, then that would be bad, but that never happened. I can't see why connecting R(IREF) to +5V would damage the chip. If I am right, it should work when you get it wired correctly. You need to replace at least a few of the RGB leds with ordinary red leds first to do the test. We need to eliminate the RGBs from the equation.

Are you saying it STILL DOESN'T WORK after correcting the resistor ? (and the chip doesn't get hot ?)

Av8ter

Thanks very much for your patience.  I got rid of the pre-made jumper wires and replaced them with 22 AWG  homemade
jumper wires like the ones from the TLC5930 to the RGBs, and all three TLCs work just fine. TheRGBs were part of the problem, they weren't seated very well and after adjusting them, they worked.    I learned something out of this "goat rope" thanks to you and your sharing of your expertise.

raschemmel

So you are confirming that my analysis of the mis-wired )R(ref) resistor to +5V instead of ground resulting in zero current , rather than maximum current was correct because NOT ONE of the chips were damaged ? Did you replace any of the RGBs with standard RED leds to test the chip as I suggested ? If you were going to write a FINAL ASSEMBLY CHECKLIST for other users for the TLC5940 what would it say ?

CrossRoads

You can have a lot of fun with a 328P and TLC5940's.
See the video ~1/2 down here
https://www.kickstarter.com/projects/954220292/the-infinity-clock-all-digital-analog-clock?ref=live
ClocKore Circuit Board
Designing & building electrical circuits for over 25 years.  Screw Shield for Mega/Due/Uno,  Bobuino with ATMega1284P, & other '328P & '1284P creations & offerings at  my website.

Av8ter

Yes, raschemmel, you were absolutely correct. The RGB leds were not damaged in any way. I removed all male/male jumper wires and replaced them with my own 22 AWG solid wire jumpers.  These fit more snugly in the breadboard, making for better contact.  Yes, I started to change out the RGBs with standard RED leds. After I replaced the first RED, I re-set the Uno, and the RED led lighted.  So I re-seated the remaining RGBs.  As I did each one, it  would light up, as I went down the row.  In a    The first thing I would say is,  If one is having problems with the chip,  get on this forum and ask for help.

Av8ter

   Also, be very careful with the wiring.  Thanks again for your patience and your help.

raschemmel

#26
Mar 26, 2014, 01:43 am Last Edit: Mar 26, 2014, 01:46 am by raschemmel Reason: 1
Now do you see why I said this back in Reply#16 ? (We are on Reply#25 now)
Quote
Do you have any other leds (ordinary red ) that you can substitute before you tear down the circuit just to rule out the RGB leds ?


Then I mentioned it again in Reply#20:
Quote
If I am right, it should work when you get it wired correctly. You need to replace at least a few of the RGB leds with ordinary red leds first to do the test. We need to eliminate the RGBs from the equation. 

Av8ter

Next time,  I will  follow the suggestions of you and others who know a whole lot more about  this than I do.  Once again, thanks for all you help.

Go Up