Is it safe to run 12v DC LED from 8v AC bell transformer?

[haggis999]

Ah, OK - I'm ignoring MikeFromLondpn. I forget why, he must have posted something rude or otherwise annoying at some point in the past.
I now see his post number 61 where he makes various comments, which I guess I should respond to:

- It clearly won't get any hotter than a resistor doing the same job.
- With a resistive constant load like an LED it will work fine without a decoupling capacitor.
- People niormally use series resistors to regulate LEDs because they have a known supply voltage and a known load current. In this case we have an unknown supply voltage and an unknown load current, but a known load voltage.
- These devices cost pennies, and the implementation is very straightforward with no need to guess or even measure anything - so it's far from overkill.

Are you talking about linear voltage regulators or switching voltage regulators? I understand that the latter are more efficient but emit much more RF noise. I'm not sure I would want to get audible interference on my AV systems. Is that a real world risk?

I guess the other point is that if it all works fine without a voltage regulator then what exactly would be the benefit?

 

[JohnW2]

Are you talking about linear voltage regulators or switching voltage regulators?

He's talking about the former. No RF noise problem.

I guess the other point is that if it all works fine without a voltage regulator then what exactly would be the benefit?

Valid question - but I think the point being made is that, given the uncertainties, you would have to do some measuring/fiddling to get the right component values for it to 'work fine' if you didn't use a regulator, but if you had used a regulator, none of that would have been necessary.

Kind Regards, John

 

[haggis999]

I checked out the 78L12 voltage regulator you mentioned earlier and can now see that it is a linear design.

 

[haggis999]

Valid question - but I think the point being made is that, given the uncertainties, you would have to do some measuring/fiddling to get the right component values for it to 'work fine' if you didn't use a regulator, but if you had used a regulator, none of that would have been necessary.

Kind Regards, John

There are two points of view being expressed about voltage regulators in this thread. It may be time to call a referendum. Those for 'taking back control' are obviously in favour, while those against are happy with the status quo (we could call them the 'remainers'). Where have I heard this stuff before...

 

[JohnW2]

There are two points of view being expressed about voltage regulators in this thread. It may be time to call a referendum.

It's really just a matter of their being many ways of skinning a cat. If one were starting from scratch (with the limited information available to you) there is an argument that one approach is 'simpler' - but either will work, so there really is no great point in a debate.

Kind Regards, John

 

[haggis999]

It's really just a matter of their being many ways of skinning a cat. If one were starting from scratch (with the limited information available to you) there is an argument that one approach is 'simpler' - but either will work, so there really is no great point in a debate.

Kind Regards, John

I'm not a great supporter of referendums (in the world of politics, that is), so it was not really meant to be a serious suggestion. I also think that I've heard enough now on this topic to make a decision. Given that such regulators cost only pennies, I will probably get one for educational purposes. If it does indeed run hot as MikefromLondon suggested it might, then he wins the argument. If not, then it seems like an elegant and cheap way of feeding the LED the voltage it was designed for.

 

[JohnW2]

If it does indeed run hot as MikefromLondon suggested it might, then he wins the argument.

As as has been said, unless Mike has found a way of re-inventing the laws of physics it is not going to get any hotter than would a resistor doing the same job. When a device (be it a resistor, an IC regulator or anything else) drops the voltage by a certain amount, with a certain current flowing, the power dissipated in the device (whatever it is) will be the same.

If anything, I would expect the temperature rise (for the given power dissipated) to a bit less with the regulator, since it is likely to be physically a bit bigger, hence with greater thermal mass and surface area, than a resistor.

Kind Regards, John

 

[Mikefromlondon]

Biggest risk using a voltage regulator is if there were any shorts, they get really hot and can often catch fire. Resistors are much safer, resistors self limit short circuit current and as long as their power rating is correct they do not get red hot or catch fire.

This is the reason I condemn anyone suggesting using a voltage regulator just to drive an LED especially if it is not soldered into a circuit enclosed in a flame retardant box. Most regulators are current devices, they regulate voltage at specified currents, even a small 78L12 can provide 100mA, and if you had a short at 20 volts input, its temperature will get really hot and burn, often catch flame, I have actually seen regulators catch a flame, unless you go to extreme length of putting heat sinks and enclosing them in safe unvented boxes.

Of course it is a case by case thing, in some applications regulators are safe, in this particular circumstance, I do not find it necessary and so it removes any remote possibility of regulator getting hot enough to burst into flames.

The police stops you speeding because you might kill yourself or someone, but you haven't actually killed anyone in your 20,30,40 years of driving history, so are police mad telling us that speed kills?

Every situation demands its own merit, so on the other hand one is allowed to drive at extreme high speeds on racing tracks where other safety parameters have been catered for against loss of life.

So I suggest if you must use a voltage regulator, please provide flame proof housing, and mount the regulator on a heat sink against short circuit currents, many such regulators have build in short circuit protection, but they often fail to work, same as a F1 car can breach safety barriers and kill spectators.

Using a resistor is a much much safer design criteria, it is a passive component with which very little can go wrong, using regulators is an active component which are prone to failure.

 

[endecotp]

 

[haggis999]

Why not do away with electrical illumination and use a tritium powered Beta Light. May not be the cheapest option but no wires and a half life of over 10 years. ( samples may be available )

http://www.betalight.nl/en/light-source.html

These are interesting light sources, which I've not heard of before, but I'm not convinced that they provide a better bell push solution. This is partly because I have already purchased an engraved stainless steel plate with an illuminated pushbutton, but primarily because (to my eyes) it looks much smarter than any BetaLight-based design I can imagine. I also don't expect my current LED solution to dim significantly after 10 years, but even if it did, I could always boost the brightness by fitting a lower value series resistor at a cost of about 10p.

 

[JohnW2]

Biggest risk using a voltage regulator is if there were any shorts, they get really hot and can often catch fire. Resistors are much safer, resistors self limit short circuit current and as long as their power rating is correct they do not get red hot or catch fire.

It really depends upon how high the LED current is (at 12V), and the higher that current, the less difference would there be between the resistor and regulator in the (very unlikley) event of a short....

If we assume an LED current of 50mA (at 12V) and a supply of 24V, then one would have a 600Ω resistor, and the current through it (with 24V across it) would be 100mA if the LED was short-circuited. With a 7812, the S/C current is limited to 230mA, so only just over twice that with the resistor. If the LED drew 100mA or so (at 12V), there would be virtually no difference between the power dissipated in resistor and regulator under S/C conditions.

Indeed, because of the built-in current limiting in the regulator, if the load drew more than 115 mA at 12V, the power dissipated in the regulator under S/C conditions would be less than that dissipated in a resistor under under teh same conditions.

Kind Regards, John

 

[Mikefromlondon]

save me writing here, Read warnings on that data Sheet at No 1 towards the bottom of the page, which means those ratings are only valid if the regulator is used properly, with some sort of heat sink, often thick and wider copper track areas of a printed circuit board. free flow of air around the device, but also at the same time avoiding in supply of fresh oxygenated air, hence device enclosed in box big enough to circulate internal air, yet stop more oxygen entering the box if should the device burst into flames, built in current fold back or limiting peak currents can make the device really hot, and if this heat is not dissipated away, the device will become even more hotter, unless heat sinks are used, and if not the device's current limit system will fail due to thermal run away, and the device will go pop or catch fire.

 

[Mikefromlondon]

The fact is John, nobody has ever used a voltage regulator to drive just one LED, I haven't come across any gadget or contraption where anyone other than the endocopt using a voltage regulator to drive an LED. It is not industry standard, its like saying I can put a F1 tyre on my family car, of course it will work, but it will also look stupid and cost a lot more and then there will be other implications, you won't be able to go past the first police car that sees you, and you may have brand new sleek F1 tyre but the police will charge you for a bald tyre with no threads!

 

[bernardgreen]

The 78Lxx regulators made by Texas Instruments have internal thermal protection

Current limiting is included to limit the peak output current to a safe value. Safe area protection for the output transistors is provided to limit internal power dissipation. If internal power dissipation becomes too high for the heat sinking provided, the thermal shutdown circuit takes over preventing the IC from overheating.

 

[JohnW2]

The fact is John, nobody has ever used a voltage regulator to drive just one LED ....

One of the very first things I was taught in my higher education is that any assertion containing words such as "always", "never", "no-one", "everyone" etc. were almost inevitably wrong

Kind Regards, John