Fundamentals of DC power?

twinkleclaus

New elf
Joined
Jan 6, 2011
Messages
44
Hi,
I have run into an issue that I find confusing regarding DC power to display elements. To set the stage, here is one example:
- (8) 3ch DMX controllers, hooked up to various blinky things
- Mean Well 12v 50a power supply (or a professional benchtop adjustable power supply)


If I hook up all 8 controllers to the power supply (basically vampire-tap the positive and negative), with each drawing (according to my bench power supply) between .03 -.3 amps @ 12v, I get weird results. That is, some channels will sometimes respond, others will flicker and some will not respond at all. I measured the voltage at the end of the 12v run and it was 12 volts. Current draw (according to my benchtop PS) was 1.7a, well below the 5a of this power supply, and certainly well below that of the Mean Well.


After a lot of troubleshooting, I cut up some cheap 12v 1a wall-warts and hooked one up to every 2 controllers. Everything worked as expected.


Another example:
I have qty 20 LED popcorn strobes. The are a very simple construction, using a 555 timer and 4 piranha white LEDs in series, run at 12 volts (regulator is the Mean Well 12v 50a power supply). I attach these into a standard straight-through (in parallel) C9 cord, and attached to the power supply. The more I add, the dimmer they become, and their behavior appears to be not that of running off the 555 timer, but rather that of the circuit resetting. Again, I checked voltage at the end and it was 12v, and the amp draw was well under the 50 amps (in the order of single digits). Again, if I set an individual power supply to 2-3 of them, they worked perfectly, but once I add more in parallel, it was erratic.
[SIZE=78%] [/SIZE]
My question is why? Why can't I treat a leg of 12v power kinda like AC, where one can just tap into it as long as you don't overrun the total amps for the power supply?


Second question I have is related to the one above, depending on the answer. If, in the case above, I need to use a constant-current power supply, what happens when power gets manipulated through PWM on a voltage regulator? I've seen many DMX lighting controllers that use a 7912 with PWM, so doesn't that kinda defeat the purpose of constant-current?


I'm totally confused.
 

David_AVD

Grandpa Elf
Community project designer
Generous elf
Joined
Jun 12, 2010
Messages
4,681
Location
Victoria Point (Brisbane)
Maybe a picture of the setup that doesn't work would help. Maybe there's something else at play here that isn't obvious from the text.

Second question I have is related to the one above, depending on the answer. If, in the case above, I need to use a constant-current power supply, what happens when power gets manipulated through PWM on a voltage regulator? I've seen many DMX lighting controllers that use a 7912 with PWM, so doesn't that kinda defeat the purpose of constant-current?

I'm guessing you meant LM317 instead of 7912 here? The LM317 sets the current, whilst the PWM pulses the power on and off (at that current) many times per second. So, the current is constant during the ON times and zero during the OFF times. The ratio of ON/OFF times is what gives you the perceived dimming.
 

AAH

I love blinky lights :)
Community project designer
Joined
Dec 27, 2010
Messages
4,188
Location
Eaglehawk
My thoughts are that the problems that you are experiencing are due to ground loops and the fact that the 3 channel controllers don't have isolated dmx inputs on them but are sharing the common supply. When you are changing to a bunch of separate supplies you are isolating them and removing any ground loops.
A diagram like David says may clear up confusion as there may be something that we envision from your description but it doesn't match what you've actually done.
 

adski

Dave Brown
Joined
Feb 15, 2012
Messages
203
Location
Two Wells, SA
Again, not knowing the exact configuration, there's a bit of guesswork here, but in general, if you tap into a longish run of light gauge wire, the wire resistance will drop the voltage and the situation will get worse the further down the run you get.

Ah! you say, but I measure 12V at the end of the run. This is still possible if you're dimming. Yet again, not knowing exactly how your controllers work, but if their PWM cycles are all synchronised and they're dimmed to say 10%, you could drop a few volts during the ON time (when they're all drawing maximum current) while barely noticing any effect on the average supply voltage. Even worse, if your DC voltmeter responds to the peak voltage on a variable waveform, you could still read 12V when the average is much lower. You'd need an oscilloscope to know for sure.

I'd try testing it with separate cables back to the power supply.

Dave
 

twinkleclaus

New elf
Joined
Jan 6, 2011
Messages
44
Thanks for the help. I think I sent you all on a wild goose chase. I realized that my bench PS is 5a up to 8v, but 2.5a from 8-20v. For my singing heads, it was hovering around 1.7-1.9a when running a sequence, but would jump up over 2a if 20 of my led modules fired in any given head (so I misstated the current draw in the original post -sorry!) I assume it might have peaked into overload protection. I also noticed that I needed to power on the heads first with the Rays 3ch DMX controllers before powering on the DMXREN8, if not, things were erratic. The heads have 2 Rays 3ch controllers, 20 RGB rectangle nodes, a 3m 12v el-wire inverter and opts/triacs for controlling each el wire strand.


As for my popcorn strobes, I think I was seeing an artifact of the tolerance drift built into them, giving the impression of not working right when 20 of them fire (according to my PS, a continuous .3a@12v). These strobes are brain dead simple. A 555 timer, trimmer pot and caps/resisters to trigger the 555 with 4 piranha white 5mm LEDs in series in each bulb. They are small enough to fit into a 2 inch gumball machine prize capsule. Pic is here:https://www.facebook.com/photo.php?fbid=405688352807620&set=a.270583379651452.63893.270420659667724&type=3


(club Jameco even picked them up! http://www.jameco.com/webapp/wcs/stores/servlet/ProductDisplay?freeText=Popcorn&langId=-1&storeId=10001&productId=2159007&search_type=jamecoall&catalogId=10001&ddkey=http:StoreCatalogDrillDownView )


Anyway, thanks all!
 

adski

Dave Brown
Joined
Feb 15, 2012
Messages
203
Location
Two Wells, SA
No probs.

They say that exercise is good for the brain, and dreaming up unlikely solutions for tricky problems is always good fun. If nothing else, it helps us to focus on where potential problems may lie.
 
Top