LED Driver issue / Startup voltage issue

Forums › Home Workshop DIY › LED Driver issue / Startup voltage issue

wonderstuff

110 posts

Master Geek

#282599 28-Feb-2021 14:52

Hi,

I am having an issue with a MeanWell ELG-150-12 which I purchased from Jaycar.
This power supply is driving 6x5W LEDs (rated at 12V) and 6x1.5W LEDs (rated at 12V).
They are all connected in parallel.

On start up, the LEDs start flickering and I measure 8-9V on the DC output. If I disconnect and reconnect the DC output, the LEDs are driven correctly and 12V is supplied.

On this PSU, there is a DIM- and DIM+ input which I have tried a number of resistances and open circuit with, but I still have the same issue.

It is not an option in this configuration to have the PSU on constantly with a switch on the DC output. An ideas how I correct this or configure it correctly to get 12V?

Thanks

Mike

1 | 2

2487 posts

Uber Geek
Inactive user

#2665767 28-Feb-2021 15:46

Sounds like the inrush on the LEDs' is too high and the power supply is browning out.
Disconnecting the DC side will allow the power supply to charge the capacitors and then it can cope with the inrush.

k1w1k1d

1519 posts

Uber Geek

#2665770 28-Feb-2021 15:56

What happens if you try to start up with less LED's connected?

If it can't start with a 39W load, you may have to add a short delay timer relay to the output?

Have you advised Jaycar of the issue you are having?

pih

649 posts

Ultimate Geek

Lifetime subscriber

#2665772 28-Feb-2021 15:59

Does the same thing happen with a single LED? I'm not an electrical engineer but I'm thinking it might have something to do with the over voltage or over current handling. If it still happens then you may have a faulty driver. If not, try running the 5W and 1.5W LEDs separately and see if you can work out whether it's just one type or the combination that is causing the issue.

Gordy7

1906 posts

Uber Geek

ID Verified

Lifetime subscriber

#2665779 28-Feb-2021 16:09

Can you provide a link to the model Meanwell power supply you are using from Jaycar?

The link you gave to the Meanwell power supplies data sheet covers too many models.

Can you provide links to the specs for the LEDs you are using?

Gordy

My first ever AM radio network connection was with a 1MHz AM crystal(OA91) radio receiver.

neb

11294 posts

Uber Geek

Trusted

Lifetime subscriber

#2665811 28-Feb-2021 17:07

andrewNZ: Sounds like the inrush on the LEDs' is too high and the power supply is browning out.
Disconnecting the DC side will allow the power supply to charge the capacitors and then it can cope with the inrush.

If that is the problem, the OP could then look at inrush current limiting, for a DC supply rather than the usual mains-voltage input you could probably get by with an NTC thermistor and a MOSFET to switch it out of circuit once it's passing the full load.

wonderstuff

110 posts

Master Geek

#2665830 28-Feb-2021 17:37

Thanks for all the responses. Comments:

> Sounds like the inrush on the LEDs' is too high and the power supply is browning out.
> Disconnecting the DC side will allow the power supply to charge the capacitors and then it can cope with the inrush.

I am surprised a PSU built for driving LEDs does not have a timer built in to address this scenario. Could be the issue though.

> What happens if you try to start up with less LED's connected?

If I disconnect 4 of the LEDs (any 4, not a specific 4), then it starts working from power up correctly.

> If it can't start with a 39W load, you may have to add a short delay timer relay to the output?

Do you have a recommended delay relay timer?

> Have you advised Jaycar of the issue you are having?

Not yet.

> Can you provide a link to the model Meanwell power supply you are using from Jaycar?

https://www.jaycar.co.nz/120w-12v-10a-dimmable-led-power-supply/p/MP3380

ELG-150-12B

> Can you provide links to the specs for the LEDs you are using?

They are purchased from a supplier on Ali-Express. The specs are incomplete.

https://www.aliexpress.com/item/32822270789.html?spm=a2g0s.9042311.0.0.59774c4dzI0J3s

https://www.aliexpress.com/item/32843410569.html?spm=a2g0s.9042311.0.0.59774c4dzI0J3s (12VDC 5W)

Thanks all

richms

28175 posts

Uber Geek

Trusted

Lifetime subscriber

#2665834 28-Feb-2021 17:45

Are some of the LEDs ones that are made to take AC or DC - like MR16 replacement ones? They will have capacitance in them and will screw up running off another driver.

Richard rich.ms

Tradepub

Free professional, reference and technical white papers (affiliate link).

Gordy7

1906 posts

Uber Geek

ID Verified

Lifetime subscriber

#2665901 28-Feb-2021 19:50

All very tricky...

The power supply obviously handles the 40 watt load ok and is stable when you connect the 'already on' power supply output to the LEDs.

Things to try, these ideas (one or the other) will waste a bit of power (about 2 watts), but are passive options which you are looking for:

If the power supply has a minimum load current requirement or the LED capacitance is upsetting the power supply startup stability, then you could try putting about a 50 ohm resistor across the power supply output and effectively in parallel with all the LEDs.

As suggested, there may be an inrush current issue.... Put about 0.15 ohm resistor in series from the power supply output to feed all the LEDs.... this may limit peak current enough.

Gordy

My first ever AM radio network connection was with a 1MHz AM crystal(OA91) radio receiver.

1101

3122 posts

Uber Geek

#2666137 1-Mar-2021 12:15

"On start up, the LEDs start flickering and I measure 8-9V"

are you measuring voltage at the PSU, or at the LED's
Perhaps voltage drop in the wiring to the LEDS , if its 12v at the PSU & 8v at the LEDs

LED inrush current issue ?
https://adlt.com.au/led-driver-inrush-currents/

if the power supply drops well below 12v , you have issues

Measure the startup current with your multimeter , if it has a max hold (and fast sampling) even better .

Perhaps have a timer & relay circuit so spread out the LED startup , rather than all at once .

wonderstuff

110 posts

Master Geek

#2666149 1-Mar-2021 12:32

> are you measuring voltage at the PSU, or at the LED's
> Perhaps voltage drop in the wiring to the LEDS , if its 12v at the PSU & 8v at the LEDs

Its at the terminals on the PSU.

> Perhaps have a timer & relay circuit so spread out the LED startup , rather than all at once .

I suspect that this is the solution. Do you have a recommendation for a timer and relay?

MadEngineer

4278 posts

Uber Geek

Trusted

#2666153 1-Mar-2021 12:44

Constant current or constant voltage. Choose.

You’re going to need a second driver for the LEDs with the different rating.

You're not on Atlantis anymore, Duncan Idaho.

richms

28175 posts

Uber Geek

Trusted

Lifetime subscriber

#2666238 1-Mar-2021 13:41

1101:

if the power supply drops well below 12v , you have issues

Measure the startup current with your multimeter , if it has a max hold (and fast sampling) even better .

Perhaps have a timer & relay circuit so spread out the LED startup , rather than all at once .

That one mentions constant current and constant voltage, so if it has a current limit then it will hit that and slowly charge the capacitors in the LEDs, but once the LEDs start to conduct and take all the current, the voltage will stop going up as its at the current limit. Basically if the PSU is constant current its not suitable to put LEDs with smooting capacitors on the output.

Richard rich.ms

frankv

5680 posts

Uber Geek

Lifetime subscriber

#2666253 1-Mar-2021 14:17

The datasheet says a B model has "3 in 1 dimming function (0~10Vdc, 10V PWM signal and resistance)". Further on it shows that you can put a PWM, voltage, or resistance across the DIM+/- pins. For resistance, increasing resistance (10K-100K if I read it right... what resistances did you try?) -> increasing current, so open circuit = full brightness.

It also says it has protection against over-current, which works in "hiccup mode" (restart immediately after shutting down the output) when below 50% voltage and 100% current. This sounds like what's happening when you first turn it on with load on.

You could add a little RC circuit with a transistor to the DIM circuit, so that initially the DIM+/- pins are connected together (0 resistance = 0 output current), but after a few seconds (or maybe after the output is over 6V (i.e. outside the hiccup area)) they are disconnected (infinite resistance = full brightness), or connected via a variable resistor if you want manual dimming control.

How are you measuring the output voltage? If with a typical multimeter, you'll get the average of the voltage, not the actual voltage. I suspect you're getting 12V for about 60% of the time, 0V for the rest to average out to 8-10V. Some meters have a max/min function, or you could use an oscilloscope or similar to see what the output signal actually is.

(I am a somewhat self-taught electronic somewhat engineer... check the above out with someone who knows this stuff before you try it, and don't ask me to design the the RC circuit).

wonderstuff

110 posts

Master Geek

#2666354 1-Mar-2021 17:09

> The datasheet says a B model has "3 in 1 dimming function (0~10Vdc, 10V PWM signal and resistance)". Further on it shows that you can put a PWM, voltage, or resistance across the DIM+/- pins. For resistance, increasing resistance (10K-100K if I read it right... what resistances did you try?) -> increasing current, so open circuit = full brightness.

I have tried 90K, 120K, 150K, and open circuit across the DIM+/- pins. Makes no difference. Lower resistor values and it does dim the LEDs.

> You could add a little RC circuit with a transistor to the DIM circuit, so that initially the DIM+/- pins are connected together (0 resistance = 0 output current), but after a few seconds (or maybe after the output is over 6V (i.e. outside the hiccup area)) they are disconnected (infinite resistance = full brightness), or connected via a variable resistor if you want manual dimming control.

Thanks for the multiple recommendations for adding Rs and RC circuits. I am trying to avoid adding discrete components if possible.

> How are you measuring the output voltage? If with a typical multimeter, you'll get the average of the voltage, not the actual voltage. I suspect you're getting 12V for about 60% of the time, 0V for the rest to average out to 8-10V. Some meters have a max/min function, or you could use an oscilloscope or similar to see what the output signal actually is.

I have a very basic multimeter and no access to a decent one or scope. I couldn't say if it were a PWM like driver.

wonderstuff

110 posts

Master Geek

#2668931 6-Mar-2021 12:41

Thanks for all your help.

As speculated, it appears to be an inrush current issue as there probably is a dc-dc driver in the LED modules themselves.

I purchased a timer-relay from trademe (https://www.trademe.co.nz/Browse/Listing.aspx?id=2997473050) which allows the PSU to stabilise at 12V and this has solved the problem.

1 | 2