Some VERY interesting LED modules

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Jim Butler 1 on 09/01/2023 00:01:12:

Posted by Michael Gilligan on 08/01/2023 22:59:35:

Posted by Robert Atkinson 2 on 08/01/2023 21:17:24:

Interesting that there is a bridge rectifier on the board.

.

… especially as the data-sheet appears to say it should not be reverse connected.

MichaelG.

Looking at the track pattern in the picture, the AC terminals of the bridge rectifier seem to be routed to the power input connector.

In which case, reversed polarity supply would not be a problem.

JimB

 

.

Exactly my point, Jim

[would not be a problem … except that it is ]

… which is why it is essential to read the data-sheet.

MichaelG.

.

Edited By Michael Gilligan on 09/01/2023 08:49:48

[Ian PParticipant @ianp]

The information in the manufacturers datasheets should sometimes be taken with pinch of salt.

I saw the bridge and the diode packages on Michael's first picture but did not (and still have not, followed the track layout) but from the text in the Plessey datasheet one might infer that the diode was before the bridge, but in that case why even have a bridge rectifier.

If the bridge rectifier is across the incoming supply then why the warning about reverse polarity I wonder.

Ian P

[Douglas JohnstonParticipant @douglasjohnston98463]

Right then-cut to the chase-has anybody got them working yet. Being somewhat chicken I have a couple of them on the workbench while I fret over how to power them without killing them.

Doug

[Ian PParticipant @ianp]

Whether you need to fret or not depends on what power supply you have, also depends on whether you just want to see them light or operate them continuously.

For continuous operation at 700mA they would need a heatsink.

Ian P

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Douglas Johnston on 09/01/2023 14:26:40:

Right then-cut to the chase-has anybody got them working yet. Being somewhat chicken I have a couple of them on the workbench while I fret over how to power them without killing them.

Doug

.

There is a suitable driver module winging its way to me, Doug

… I will report back a.s.a.p.

MichaelG.

[Robin GrahamParticipant @robingraham42208]

Posted by Douglas Johnston on 09/01/2023 14:26:40:

Right then-cut to the chase-has anybody got them working yet. Being somewhat chicken I have a couple of them on the workbench while I fret over how to power them without killing them.

Doug

This is a 60 degree version running in an otherwise unlit room (white ceiling):

It's pretty bright!

It's running from a lab PSU in CC mode, current clamped at 700mA , voltage was 43.4V at that current.

Not having a suitable heatsink I just perched it on top of a piece of 2.25" Ali bar, fresh from the workshop so at about 10C. I ran it for maybe 2 minutes – the aluminium back of the LED was stone cold after that but the metallised plastic housing was warm.

I'll rig up a thermocouple and get proper measurements anon.

Robin.

 

Edited By Robin Graham on 09/01/2023 22:12:38

[Robin GrahamParticipant @robingraham42208]

PS. I guess the fact that the array didn't blind the camera at that angle shows that the 'optics' work pretty well in directing the beam – the overall illumination comes mostly from reflection from the ceiling, about 6ft above the device.

Does anyone know how to calculate heatsink requirements for continuous running?

Thanks for sharing this discovery Michael – I think I might have to order more if they're still available at this price.

Robin.

[Robert Atkinson 2Participant @robertatkinson2]

I just got my 25 degree one and it is very bright. I expected this as they are spotlight replacements.
The "turn-on" voltage on mine is 34-35V (1mA @ 35.5V) .
Heatsink is easy
Maximum operating temperature is 90 deg C. and the power is 31W @ 0.7A. The heatsink requiement is 90 – ambient temperature /31. So for a consertaive 35 deg. ambient it is (90-35) / 31 = 1.77 deg / W.
Which is quite big. An example:

https://uk.rs-online.com/web/p/heatsinks/1898280

Robert G8RPI.

[Les Jones 1Participant @lesjones1]

The two that I ordered arrived today. I found a suitable heatsink to use.

Here are some measurements of voltage against current

100 mA 40.24 V
200 mA 41.21 V
300 mA 42.07 V
400 mA 42.80 V
500 mA 43.29 V
600 mA 43.76 V
700 mA 44.15 V

Even at 100 mA it was quite bright.
Many thanks to Michael for pointing them out.

Les.

[Ian PParticipant @ianp]

I too received a pair of these modules and thank Michael for spotting them.

Regarding the voltage/current values just posted by Les, I would like to point out to (less electronically aware readers) that it is current rather than voltage which is the main consideration when selecting a power source.

The voltage across an illuminated string of LEDs (which varies slightly with temperature anyway) may be of interest but its not critically relevant, with these modules a 'constant current' supply (with an unloaded output higher than 45 volts (ish) is how they should be used.

I have not powered up mine yet but I not try to connect them direct to a regulated DC supply even if it was set to 44.15 Volts

Ian P

[Les Jones 1Participant @lesjones1]

Ian makes a good point as beginners may be tempted to drive them with a constant voltage rather than a constant current. As well as the voltage for a given current changing with temperature it can be different between samples of LEDs. This is why it is a bad idea to drive LEDs in parallel. (Unless they are matched.)

Les.

[Robin GrahamParticipant @robingraham42208]

Posted by Robert Atkinson 2 on 12/01/2023 16:14:26:

I just got my 25 degree one and it is very bright. I expected this as they are spotlight replacements.
The "turn-on" voltage on mine is 34-35V (1mA @ 35.5V) .
Heatsink is easy
Maximum operating temperature is 90 deg C. and the power is 31W @ 0.7A. The heatsink requiement is 90 – ambient temperature /31. So for a consertaive 35 deg. ambient it is (90-35) / 31 = 1.77 deg / W.
Which is quite big. An example:

https://uk.rs-online.com/web/p/heatsinks/1898280

Robert G8RPI.

Thanks for the heatsink calculation Robert – I thought it would be more complicated than that!.

Can you – or anyone else – rationalise this from the datasheet:

"Do not expose the module to liquids such as adhesives or thermal paste. No thermal paste
should be used as a TIM between the module and the heatsink. Do not remove the product
barcode".

My bold/italics. It seems common sense to me to use thermal paste. Am I missing something?

Obviously no one in their right mind would remove the barcode. One shudders to think of the consequences…

Robin.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Robin Graham on 12/01/2023 23:16:34:

[…]

My bold/italics. It seems common sense to me to use thermal paste. Am I missing something?

Obviously no one in their right mind would remove the barcode. One shudders to think of the consequences…

Robin.

.

It confused confuses me too, Robin …

As for the barcode [the little titchy one on the edge] … it’s interesting to note that it’s not the part number, so it might be something to do with batch traceability.

MichaelG.

[Robert Atkinson 2Participant @robertatkinson2]

I'd not spotted the note about thermal paste. Most of these are based on silicone oil and a filler (typically zinc oxide).
The Silicone oil tends to bleed out of the paste and creep across surfaces. I can only assume that some component of the assembly is not compatible with silicone oil. Given the large surface area of the assembly and low power rating I would not have thought compound was needed anyway. Its more typivcally used on small components or those with a lot more power.
My heatsink calculaton do not allow for the thermal interface as I conservatively assumed all the input power turned to heat. In rality only about 60% is heat so more like 18W.

Robert G8RPI.

[John HaineParticipant @johnhaine32865]

Silicone polishes were banned from strowger telephone exchanges after a contractor used some on a floor instead of beeswax. The oil crept all over the contact surfaces and oxidised to silica if there was any arcing. Result, one exchange out of service. As the design of the light just has some kind of mechanical contact between bare wire and some sort of barb arrangement that may be susceptible to the same problem?

[Jouke van der VeenParticipant @joukevanderveen72935]

Can somebody tell if this LED system is dimmable with, for instance, an Euchips 30 Watt Constant Current Dali and1-10 Volt Driver 550-900mA, with dip switch settings 700mA-43VDC?

Or even simpler with a 0-10V dinmming with a 100kOhm potmeter?

Regards,

Jouke

[Robert Atkinson 2Participant @robertatkinson2]

That dimmer specification is marginal. The 43V is the problem. It would work but you would probably not get full output. The LED is specificed as 44V typical, 46.5V max so the driver may not reach the rated 700mA curent.
This one:

file:///C:/Users/rober/Downloads/EUP30D-1HMC-0_102100251101_EN%20SPE%20DALI%2030WATT.pdf

has a 650mA 45V seting which would be better. and is 97.5% of the rating. You won't notice the missing 2.5%.

Robert G8RPI.

[V8EngParticipant @v8eng]

Posted by Robert Atkinson 2 on 13/01/2023 12:33:11:

That dimmer specification is marginal. The 43V is the problem. It would work but you would probably not get full output. The LED is specificed as 44V typical, 46.5V max so the driver may not reach the rated 700mA curent.
This one:

file:///C:/Users/rober/Downloads/EUP30D-1HMC-0_102100251101_EN%20SPE%20DALI%2030WATT.pdf

has a 650mA 45V seting which would be better. and is 97.5% of the rating. You won't notice the missing 2.5%.

Robert G8RPI.

I think you may have linked to a local file on your C Drive?

[Robert Atkinson 2Participant @robertatkinson2]

Doh!

[Jouke van der VeenParticipant @joukevanderveen72935]

But the recommended type is clear: EUP30D with Dali.

I understood that with a 0-10V dimming you can do this just with something like a 100kOm potmeter connected to the driver and that with a 1-10V dimming you need an active adjustable voltage supply connected? Is this really the case?

[Jouke van der VeenParticipant @joukevanderveen72935]

Double

 

Edited By Jouke van der Veen on 13/01/2023 16:00:01

[Jouke van der VeenParticipant @joukevanderveen72935]

Perhaps the difference between 0-10V and 1-10V dimming I mentioned above is wrong.

Perhaps a CC Led driver with a “sink” dimming system only needs a potmeter for dimming while a driver with “source” dimming needs a dimmer which feeds a DC voltage (0/1-10V) to the driver?

[Jouke van der VeenParticipant @joukevanderveen72935]

Hallo Robert G8RPI,

The dip-switch values for the EUP30A and EUP30D are the same, so to see.

650mA/9-45V and 700mA/9-43V for both.

And I found in a datasheet that the EUP30A is dimmable by means of a passive resistor (simple potmeter?).

Somewhere else it was said that you should not apply a led driver above approx. 80% of its maximum power, higher would have a negative influence on its lifetime.

[Michael GilliganParticipant @michaelgilligan61133]

The driver module that I ordered is one of these: **LINK**

https://uk.rs-online.com/web/p/led-drivers/7772724/

If you look at the ‘series’ data-sheet, there are several variants available, compatible with different types of dimmer.

Can’t comment on value-for-money because mine was a very reasonably priced ebay purchase

”bought for a project and never used”

I will try it with a suitable dimmer when I can, and will report back.

MichaelG.

.

Edit: __ see also

https://www.meanwell.co.uk/led-drivers/constant-current-led-drivers

Edited By Michael Gilligan on 13/01/2023 20:23:12

[Michael GilliganParticipant @michaelgilligan61133]

Apologies …

I should have mentioned that the ‘Guide’ linked on that RS page is a useful download.

MichaelG.

[Author]

Posts