230 Volts LED

[RODALCO]

Hero999
It still protects the LED.
The 1 N4007 is a 1 kV diode, the LED is in series with it, and can only handle 6 Volts reversed max. so a nasty spike while the LED is not conducting and revere biased still could kill the LED, for the sake of a 5 cts diode and 20 seconds extra soldering time a extremely reliable potential indicator for 230 Volts is built.
I know the two series resistors will limit the available voltage across the LED anyway, but better reliable than popped LED and repair.
If reversed voltage sharing with two diodes is required both diodes have to have same voltage and current rating ideally, IMO.
In the 1500 or so meters I have installed 3 LED's each in makes 4500 * 230 Volts LED's with NO faillures. Try to beat that. !!

 

[eblc1388]

In the 1500 or so meters I have installed 3 LED's each in makes 4500 * 230 Volts LED's with NO faillures. Try to beat that. !!

This is a correct argument that the circuit and its element are sufficient to obtain the reliability mentioned.

I can also claim the same reliability if I add two more 1N4007s in series with the original 1N4007 but that does not implies that my two 1N4007 is "necessary" to order to obtain the same reliability.

The point is, in my opinion and maybe others also, same reliability will result if Diode D2 is omitted. This remains an opinion only as I don't have any data to prove it.

 

[audioguru]

The 1N4007 is darn slow to turn off and allows a good portion of the reverse voltage to zap the LED.

 

[eblc1388]

Assume turn off time of 1N4007 at 3mA forward current or less is 20us.

The 50Hz AC waveform timing is 20ms per cycle and so the max reverse voltage that fed through the 1N4007 diode after zero crossing is.....a few volts ????

 

[audioguru]

... so the max reverse voltage that fed through the 1N4007 diode after zero crossing is.....a few volts ????

I don't know exactly how slow a 1N4007 rectifier turns off.
The absolute max reverse voltage for an LED is only a few volts.

 

[JimB]

Rodalco designed this indicator circuit for use in an industrial environment.
He wanted good reliability.
The cost of the extra 1N4148 diode is nothing compared with the cost of replacing a faulty indicator unit.

If you were designing these things to be built by the million in China, and sell on the domestic market as cheaply as possible, you may well consider missing out the 1N4148.
If the LED fails in 2 months time, what the hell, you can sell them another one!

Horses for coarses.

JimB

 

[fever]

well thanks RODALCO.
i built it for a desk crystal ball,and its awesome.
but i want that LED(i am using green) to be little more bright.ofcourse iam using ultra bright one only.so what shld be modified to get little more brightness by keeping the LED safe.

 

[RODALCO]

fever

Use a higher efficiency LED e.g 15000 Millicandela instead of 5000, or put 2 or 3 LED's in series with their own antiparallel diode each.

We have an interesting discussion going here about these indicators, great.

 

[eblc1388]

Rodalco designed this indicator circuit for use in an industrial environment.
He wanted good reliability.
The cost of the extra 1N4148 diode is nothing compared with the cost of replacing a faulty indicator unit.

Yes, this is really an very interesting discussion.

Of course I would not advocate to Rodalco to remove the diode in all his existing circuits.

The argument is based purely on the function of the diode as an diode in circuit theory. But there is still no solid proof, to me at least, via circuit simulation or calculation or otherwise, that the diode D2 is doing its share to enhance the reliability of the circuit. If that is the case, one cannot simply says that the diode is there for good reliability as I can also safely say two 1N4007s in series will also enhance reliability.

I remembered seeing diodes placed in series to semiconductor base to prevent breakdown of the semiconductor junction so these circuits might share some similarity here. There is no anti-parallel diode connected also in those cases.

And, sorry Rodalco I would disagree again with your suggestion that separate anti-parallel diodes are needed instead of using just one across the LED chain.

 

[RODALCO]

BUMP !! Just get this

 

[RODALCO]

BUMP !! Just get this LED topic relivened

Cheers, Raymond

 

[kchriste]

Hero999
It still protects the LED.
The 1N4007 is a 1 kV diode, the LED is in series with it, and can only handle 6 Volts reversed max. If reversed voltage sharing with two diodes is required both diodes have to have same voltage and current rating ideally, IMO.

I agree with you. The reverse leakage of the 1N4007, while very small, could cause the LED to be reversed biased during the negative half of the cycle. When rectifier diodes are placed in series for a higher PIV, then balancing resistors are required:
**broken link removed**
Since the LED is a diode, the same issues arise and hence the inclusion of the extra diode. You could probably accomplish the same thing with a resistor (10K) across the LED.

 

[Hero999]

I doubt it would damage the LED, especially if it's a red LED. When I was a child I hooked up a red LED and a 10k resistor to the mains with no diode in reverse and it didn't damage the LED. White and blue LEDs seem to be more sensitive and are destroyed when subjected to similar abuse.

 

[RODALCO]

Interesting Hero999, I did the same when i was about 13 years young and it worked with a LED on 220 Volts in The Netherlands.

Somehow the older inefficient LED's stand more abuse. I did the LED with an about 68 k 1/2 watt resisitor which went up in smoke after a while. a 1 watt worked but got very hot but lasted.

In that experiment I used a yellow LED with probably 5 mCad light output, now these days the LED's are 1000 + times brighter.

Ok back to the topic, In case of indicator LED's The cost of a servicecall for a failed indicator far outweights the costs of a 5 cts 1N4148 diode.

You probably get away with it but RELIABILITY is the issue for me.

 

[RODALCO]

Bump !

in case the link doesn't work

 

[Mr RB]

The mains voltage is not 50Hz, at least the noise spikes are not, they are MUCH faster and will trash the LED without the 1N4148 diode.

I've done many similar LED circuits in mains equipment I have built over the years, but I use a large cap in parallel with the LED (100uF, maybe 220uF) and no 1N4148. The cap absorbs any of the fast spikes that get through the 1N4007 and has the added advantage of absorbing positive spikes as well as negative, and improves LED brightness and appearance due to the reduced ripple at the LED.

 

[RODALCO]

The mains voltage is not 50Hz, at least the noise spikes are not, they are MUCH faster and will trash the LED without the 1N4148 diode.

I've done many similar LED circuits in mains equipment I have built over the years, but I use a large cap in parallel with the LED (100uF, maybe 220uF) and no 1N4148. The cap absorbs any of the fast spikes that get through the 1N4007 and has the added advantage of absorbing positive spikes as well as negative, and improves LED brightness and appearance due to the reduced ripple at the LED.

That is exactly the reason why I prefer the resistive voltage drop option.
Transformer tapchanger (TCOL) switching spikes and or lightning interference, line clash etc. can create short duration spikes less than 1 or 2 cycles and basically fly through the series voltage drop capacitor and can damage the LED, as they are of a different time constant.

 

[oslinux23@yahoo.gr]

thank you so much

 

[Mr RB]

That is exactly the reason why I prefer the resistive voltage drop option.
Transformer tapchanger (TCOL) switching spikes and or lightning interference, line clash etc. can create short duration spikes less than 1 or 2 cycles and basically fly through the series voltage drop capacitor and can damage the LED, as they are of a different time constant.

Good point that a resistor will reduce the spikes much more than a R+C system.

However I've still lost leds over the years where I just had them connected with a resistor and diode. Here in Australia we have 240v mains which has bigger spikes and we probably have a noisier mains in general due to the long distance distribution systems and many areas have quite old equipment. My mains here fluctuates in voltage a lot and gets nasty 1+ second brownouts too.

But after using a decent size electro cap across the LED I haven't had a blown mains led yet.

 

[bailey45]

The anti-parrallel diode is require due to high voltage spikes. The 1N4007 diode is also a capacitor if you consider that a 1000-2000v spike on a power can be a common occurance the 1N914 will protect the LED when the capacitance of the 1N4007 passes the spike.