LED strip intermittent flickering

[lepton]

Hello,

I've recently made a LED strip consisting of 10 parallell modules, each module containing 4 LED's in series running on a bench supply set to 12.5V and limited to 190mA.
My problem is that sometimes the bench supply freaks out for some reason, and the voltage and current falls for around 10.3V at around 5mA, it then randomly jumps to the normal voltage, stays there for a short while (maybe 1-200 milliseconds), then falls down again. The weird thing is that it's highly inconsitent, it can run for hours without problems sometimes, and sometimes if refuses to work propely at all and freaks out instantly. There isn't any loose connection anywhere, the bench supply works absolutely fine otherwise and the modules themselves seem fine. Might it just be that I'm getting what I paid for, really cheap LED's?

The supply is a Velleman PS3005D and the LED's can be found here if anyone is interested http://www.ebay.co.uk/itm/302098225565?euid=c5ac5faec330493cb94cf7fac24c1acd&bu=44343428079&cp=1&sojTags=bu=bu

Thank you.

[danadak]

If you are typing 4 leds in series, 10 of these series elements
in parallel, then set the short circuit current of the supply to
190 mA, you have no guarantee what led gets what current.
Some will be off, some on hard, because of the led to led
threshold variation. In fact you could be exceeding the max
LED rating of 150 mA.

LEDs like to be driven with constant current.

Cant explain the PS behavior. Try a dummy load, eg Rload
 = 12.5 V / .19A, and see if the PS has issues.


Regards, Dana.

[frozenfrogz]

Ideally you would include a current regulator in every parallel strand. As danadak stated, you run into problems, because there is no way the current will flow equally through all strands - even if you have expensive LEDs with a very tight binning. Voltage drop will vary even more when the LEDs heat up. LEDs in parallel are incredibly bad at sharing current.

[lepton]

After some testing it does appear that the supply is faulty, it's doing the jumping with another fixed load as well, strange that it never happened when I used it with other circuits...
Admittedly the design is flawed, but I just wanted some nice lighting on the cheap, so any sort of regulation went out the window.

Thanks for the help,
lepton

[danadak]

Inspect the mechanical connections (like to binding posts) as well as the PCB
in the power supply for crack free solder connections. Could be you are seeing
thermal effects on marginal connections and traces. Even remove and re-insert
any socket-ed devices. Lead oxidation can cause issues.

Also look at test leads, I just bought a set with banana plugs on one end,
alligators on the other, alligator ends had poor crimping which caused
issues.


Regards, Dana.

[lepton]

I've checked the power supply from top to bottom, can't see (or smell) anything out of the ordinary. Tried refitting all the connectors, nothing. What I can hear is that the two relays inside are trying to switch for maybe 10 seconds or so, then stop. They do this after I turn on the output. This isn't the same relay-going-crazy business as previously mentioned, because then the relays did switch, with the voltage and current going mental along with it. Any ideas? (Maybe it's proper conduct for a new thread?)

Thanks,
lepton