LED Driver

[bernardgreen]

I wonder how easy is is to find LED bulbs/lamps (essentially just 'bare' LED elements) to be 'driven' by them?

The long strips of LED lamps are just LED elements and resistors. Most commonly the 12 volt strips are repeated sections each section having 3 white LEDs each of 3 volts Vforward in series ( 9 volts ) leaving 3 volts across the resistor whose value is chosen to set the current through the LED elements.

 

[JohnW2]

The long strips of LED lamps are just LED elements and resistors. Most commonly the 12 volt strips are repeated sections each section having 3 white LEDs each of 3 volts Vforward in series ( 9 volts ) leaving 3 volts across the resistor whose value is chosen to set the current through the LED elements.

Indeed - so they are designed to run from a 'constant'-voltage source.

If they were designed for a constant-current source, those resistors would not be there - they would just be a source of wasted power.

Kind Regards, John

 

[ianniann]

LED strips are inherently constant voltage, being a variable number (depending where you cut) of parallel mini-circuits with a small number of SMDs. That said, some of them have their own constant current circuitry built in, one or two small ICs for each group of LEDs - expensive though.

Plenty of bare LEDs out there that people are using with constant current drivers, but light bulbs usually have their own driver circuitry of sorts designed to work off a constant voltage. Usually quite a bit more sophisticated than the plain resistor found on the flexible LED strips and prone to fight with a constant current driver. There are some bulbs out there intended for constant current drivers, probably the best-known are the SORAA MR16s (also in constant voltage format).

 

[JohnW2]

LED strips are inherently constant voltage, being a variable number (depending where you cut) of parallel mini-circuits with a small number of SMDs.

Quite. As I wrote to bernard, that was what he was describing, even though his comment was in response tro my asking what 'was out there' to be driven byfrom a constant-current source.

Plenty of bare LEDs out there that people are using with constant current drivers ...

Are you talking literally about 'bare LEDs', like ...

... (sorry it's red!) if so, I wouldn't think many people use them for lighting.

There are some bulbs out there intended for constant current drivers, probably the best-known are the SORAA MR16s (also in constant voltage format).

Those are the sort I things I was saying that I'd never heard of. What sort of base do they have?

Kind Regards, John

 

[ban-all-sheds]

I'm not so sure about "less to go wrong". Every single such LED lamp has to contain some sort of (often quite crude) AC LV -> (usually DC) ELV 'power supply' - which is appreciably more complex than the electronics required for current control in those designed to be fed from a (single for several lamps) external power supply.

Yes, but surely the makers will have spared no expense to use the best design, and the best components, won't they?

 

[JohnW2]

Yes, but surely the makers will have spared no expense to use the best design, and the best components, won't they?

If you say so.

Kind Regards, John

 

[ericmark]

The first thing I noted when looking at the link was

LED bulbs require a constant current to work properly.

yet it seems to be showing a constant voltage supply, second is

This LED Driver is the perfect product to replace the 12v transformer which is powering your existing low voltage MR16 or MR11 downlight, and your G4 fitting

but it has a DC output and most the MR16 I have looked at are marked 50/60 Hz so are AC units.

I am sure many MR16 lamps will work with a DC supply, but I expect some which say 50/60 Hz actually do need an AC supply?

As the constant current main advantage is it auto compensates for volt drop, so one can install a line of lights down the side of a 2 mile runway and every lamp is the same brightness. In real terms the transformers are wired in series so losing one bulb will not cause all the lights to go out. In domestic I can see the advantage when running down a long garden, but 75 volt is maximum for extra low voltage DC and 3 volt approx for a white LED, so 25 LED's is maximum in real terms more like 15 LED's keeping under 50 volts as often not ripple free.

What I see as a problem is a fitting which can take a lamp anywhere between 1 and 50 watt AC or DC with nothing at the lamp to say what should be fitted. Add to that angles 15° to 36° and often under the care of just an instructed person. Unless we use the LII GU10 where a dimple is required in the lamp for it to fit in the holder we have no way to ensure an LED lamp is replaced with an LED lamp.

Some power supplies will auto shut down with over load, likely the linked ones will, but there is nothing in the spec that says it will auto shut down, so either you need to fit fuses or get one which says it will auto shut down.

To my mind it's a watch my back exercise, if the unit says it auto shuts down on over temperature and over current then if some daft twit fits a 35W quartz bulb then if it goes wrong it's not your fault. Be it needing a class II sticker or having a stated overload device as electricians we have to ensure they are fitted. As a DIY guy you can take a chance, if it does go wrong your unlikely to be taken to court, and if DIY in your own house you are unlikely to try fitting a spare 35W quartz as you know transformer only rated 25W, but as an electrician you have to expect at some time some one will say, I'll just borrow a bulb out of this table lamp until I can get a replacement.

 

[ban-all-sheds]

I am sure many MR16 lamps will work with a DC supply, but I expect some which say 50/60 Hz actually do need an AC supply?

Surely those MR16s will be LV ones, not ELV?

What's the frequency of the AC output of a switched mode PSU?

 

[ban-all-sheds]

http://www.ledbenchmark.com/faq/Transformers-for-MR16s.html

http://www.ledbenchmark.com/faq/Transformers-Output-and-Compatibility.html

 

[JohnW2]

Surely those MR16s will be LV ones, not ELV?

http://www.ledbenchmark.com/faq/Transformers-for-MR16s.html

I may have misunderstood your point, but have you perhaps answered your own question? We've discussed this before, and it does indeed seem that at least some ELV LEDs are designed for AC. Indeed, as we've also discussed, even those which don't mention (and perhaps are not 'intended' to be used with) DC will often/usually contain a bridge rectifier (so as to make them polarity-insensitive) - so may well actually work with AC. As you will be aware, the first bit of your first link reads:

This article assumes you are considering retrofitting existing low voltage halogen downlights with LED MR16 lights (MR16 lights are also known as GU5.3 sometimes). LED MR16 lamps require external driver circuitry in order to step down from the 120/240VAC mains supply to 12VAC. The LED chips inside LED lamps are driven by direct current (DC). As a result of this, MR16 LED lamps contain electronics to convert from AC to DC (generally a bridge rectifier). .... There is also a different class of low voltage LED lights that do not contain a a bridge rectifier and are DC only

What's the frequency of the AC output of a switched mode PSU?

I imagine that the AC->DC conversion within most of these LED lamps will be very crude (little more, if anything, more than the rectifier itself), so quite probably will not be appreciably concerned about frequency. However, having said that, as eric has indicated, some (of the ELV ones) appear to be labelled "50/60 Hz", so I don't know what that's about.

Kind Regards, John

 

[bernardgreen]

some (of the ELV ones) appear to be labelled "50/60 Hz", so I don't know what that's about.

They use the impedance of a capacitor in series with the incoming supply to control the amount of power that is available to power the LED element

 

[JohnW2]

They use the impedance of a capacitor in series with the incoming supply to control the amount of power that is available to power the LED element

That's obviously one possibility. However, when we've discussed this before, some people have reported that some of the ones labelled "50/60Hz" actually do work on DC - in which case there obviously cannot be a series capacitor.

If/when they do have a series capacitor for current limiting, then things will obviously be frequency-dependent, and one designed for 50/60Hs would presumably be quite likely to 'blow up' if supplied by a SMPSU with a high frequency AC output. It seems a bit odd that such things should be around, given that wirewound transformers for ELV lighting (the only practical way to get 50/60Hz) seem to be a relative rarity in the consumer marketplace.

Kind Regards, John

 

[ban-all-sheds]

I may have misunderstood your point, but have you perhaps answered your own question?

Ah - this is what happened.

1. I typed the reply to Eric, but then before posting I thought "I ought to be able to find that out", so I Binged to see what I could find. And I found those articles, which I thought were interesting.
2. I posted my existing reply instead of the URLs.
   
3. I replied to my post instead of editing it to replace it with the URLs.
   
4. I edited that reply to strip out the quote.
5. I thought "I should have replaced my bogus reply with a ".", and gone back and edited the #1 reply to Eric.
6. I decided to have a cup of tea instead of doing any more law-of-diminishing-returns tweaking.

As for the answer - I saw, as I expected, that the basic output of the supplies is HF AC. I didn't know that that was shaped into a sort-of-sine wave envelope. Nor do I know why that is done, given that the supplies are designed for halogen lamps which don't need to be fooled into thinking that they are getting 50/60Hz AC.

 

[ban-all-sheds]

It seems a bit odd that such things should be around, given that wirewound transformers for ELV lighting (the only practical way to get 50/60Hz) seem to be a relative rarity in the consumer marketplace.

Less rare in the USA?

 

[JohnW2]

Less rare in the USA?

I wouldn't know - you may be right.

Kind Regards, John