Names Brucey
New member
You are the man. I initially tried using one of the leads cut from a cap to bridge the 5v one. The PWM one I am lost as to what to do.
DIY LED driver for reef lighting
- Thread starter der_wille_zur_macht
- Start date
der_wille_zur_macht
Team RC
Brucey, any updates?
Names Brucey
New member
Nothing good. I tried all suggestions, including some others from Fishman, to no avail. The pots appear to be perfectly fine (although admittedly at a very odd level compared to where they were set. So the long and short is that after removing the caps and trying just a bare board I was still getting strange issues. Some of the outputs were showing voltage to them without having a 5v source. Some of the outputs were giving no power still and this is with all new CATs. Unless I somehow ruined 12 brand new CATS then I think something has just gotten screwed up on the board itself.
Names Brucey
New member
I want to start by saying thank you to everyone for their help and interest. Mostly, I want to extend my extreme gratitude to Fishman. He opened his doors to me last night to sit down and trouble shoot my issues. He was successful in trouble shooting the boards and got me going. I can not begin to tell you how gratefull I am and how much I appreciate fine folks like himself who are willing to help.
der_wille_zur_macht
Team RC
Any chance you guys can describe what was wrong with the boards? It might help others troubleshooting. Plus, I'm curious.
Makes me wonder if we should consider some sort of reverse polarity protection by default on these designs.
Makes me wonder if we should consider some sort of reverse polarity protection by default on these designs.
Names Brucey
New member
Several factors, most of which installer driven I think. The reversed polarity fried both my CAT's and my on-board 5v regulators. My initial attempts to simply replace just the CATS proved useless as we know. When I then determined that the 5v regulators were blown, I changed out the one on a board with all new CATS and for some reason, assuming manufacturing issue, it was giving a constant 10. It appears as though that constant 10 fried almost all of my CATS again. When I had tried replacing the 5v on another board with all new CAT's, I apparently did not get a sufficient or thorough solder through the entire thru-hole. It was making a connection on the lower plane but not the surface plane so no 5v, even though it appeared to be soldered in-place. The long and the short is, reversed polarity will definitely kill the CATS as well as in my case, the 5v regulators. My own stupidity for not realizing or detecting the 5v solder issue...well that was just stupid.
That's my take at least. I will let Fishman tell you what else he knows as most of his testing and tracing was above my head. I learned alot, that is for sure.
That's my take at least. I will let Fishman tell you what else he knows as most of his testing and tracing was above my head. I learned alot, that is for sure.
Names Brucey
New member
Please understand too, that this was after nearly 2.5 hours of troubleshooting and the like. I have not re-populated the other 2 boards that were affected by the reverse polarity issue although we both presume that the root cause/effect was dead CAT and dead 5v regulator and we hope that fresh new parts will solve those 2 boards as well. There was a 5th board we worked on but that was a whole different issue and that board was not subjected to the reverse polarity.
TheFishMan65
New member
I think that summed it up. Reversed polarity fried the CATs and 5 volt regulator. As I recall my regulator did not blow, but maybe I just did have the reversed voltage on long enough or I got lucky. The 3 CAT version won't have this problem 
Removal of parts i easier than it sounds. I think the reason for some of the bad connections was that some of the pads had been lifted or torn out on removal. I have not done a lot of removal, but until these board don't think I had ever lifted a pad.
Some things when debugging.
Pin 1 - PWM - should be 5 volts
Pin 2 - Vcc - should be 5 volts
Pin 3 - ground - should be zero volts
Pin 4 - Rsense - should be 1.2 if it read zero you have a bad CAT from our experimenting
PIN 5 - will read 0 with no LEDs we thought this meant bad until we remembered that CAT controls the voltage of the LEDs by adjusting negative reference to the string. Ideally this should read 0.5 volts. The 0 has a stumped for a little while.
I think that about covers it.
Removal of parts i easier than it sounds. I think the reason for some of the bad connections was that some of the pads had been lifted or torn out on removal. I have not done a lot of removal, but until these board don't think I had ever lifted a pad.
Some things when debugging.
Pin 1 - PWM - should be 5 volts
Pin 2 - Vcc - should be 5 volts
Pin 3 - ground - should be zero volts
Pin 4 - Rsense - should be 1.2 if it read zero you have a bad CAT from our experimenting
PIN 5 - will read 0 with no LEDs we thought this meant bad until we remembered that CAT controls the voltage of the LEDs by adjusting negative reference to the string. Ideally this should read 0.5 volts. The 0 has a stumped for a little while.
I think that about covers it.
kcress
New member
Sometimes it behooves you to check things incrementally while repairing. For instance once the original CATs were removed - since you estimated they were the only thing blown - you could power up the board and check the 5V before installing new CATs.
I've learned to always stuff and test a new switching supply on a board I've designed -before- installing the next $200 dollars worth of parts. Or alternatively I include a jumper from the supply to the board. I don't install the jumper until I've proved the supply.
I've learned to always stuff and test a new switching supply on a board I've designed -before- installing the next $200 dollars worth of parts. Or alternatively I include a jumper from the supply to the board. I don't install the jumper until I've proved the supply.
TheFishMan65
New member
Funny, I just got an email and tried to do a step by step.
1) Replace teh 5 volts and make sure it is correct
One other recommendation. I have decided on my builds to test them before I solder down the CATs tab. Then if you need to take it out it is much easier. If you do this the voltage difference becomes more critical because you don't want to damage the CAT from heat.
1) Replace teh 5 volts and make sure it is correct
One other recommendation. I have decided on my builds to test them before I solder down the CATs tab. Then if you need to take it out it is much easier. If you do this the voltage difference becomes more critical because you don't want to damage the CAT from heat.
jcgd:
Not to critique your pcb board design, and I am no electrical engineer, but think you may run into problems with using pots on the Rsense. Data sheet specifies a Kelvin Connection with the Rsense resister.
http://www.cypress.com/?id=4&rID=37694
jcgd:
Not to critique your pcb board design, and I am no electrical engineer, but think you may run into problems with using pots on the Rsense. Data sheet specifies a Kelvin Connection with the Rsense resister.
i contacted the company and asked if a pot could be used instead of a resister on the rsense for adjustments and they replied it would be fine but i think they where talking about a digital pot instead of a plan pot and sense i know nothing about a digital pot i went ahead and put a trimpot on mine its been running fine for 3 months know
Not to critique your pcb board design, and I am no electrical engineer, but think you may run into problems with using pots on the Rsense. Data sheet specifies a Kelvin Connection with the Rsense resister.
i contacted the company and asked if a pot could be used instead of a resister on the rsense for adjustments and they replied it would be fine but i think they where talking about a digital pot instead of a plan pot and sense i know nothing about a digital pot i went ahead and put a trimpot on mine its been running fine for 3 months know
TheFishMan65
New member
I have been running with a trim pot also - no problem. I think the main issue is to keep the leads short.
Oh jeez guys, I don't even know what the rsense is. If someone can spell it out for me I'll ask my buddy about it and see what he says. Everyone using the boards so far said they work great, but I haven't used mine yet so no personal experience.
Is the rsense the resister between the cat and the pot dimmer? What is a Kelvin Connection?
Question about the CAT PWM inputs, can they be all conencted together or is a seperate PWM required for each CAT. I plan on driving them all with one PWM signal. Should I isolate them with a diode for each CAT.
TheFishMan65
New member
Yes PWMs can share a common source.
