Replacing Neon Lamp with LED

[paul_g_787]

I have a device here that I wish to replace the failed Neon power indicator.

It has no low voltage DC as it is a mains appliance, so it needs to run from the mains voltage.

I have done a lot of research and learned about capacitive drop supplies.  I have dismantled and reverse engineered a few LED light bulbs and learned the equations for capacitive reactance.

I have finally settled on this design which supplies the LED with 2V 10mA from the 240V AC Mains.
It works brilliantly and the LED should last a long time on 10mA.

Is there any improvement that can be made to this design?

My main concern about putting a fuse on it, as it is directly across the mains. Could I perhaps move R2 to the other side of C1 and change it to a fusible resistor? How would I calculate which fusible resistor?

[mikeselectricstuff]

The 100uf cap is overkill.
Do you need as much as 10mA?
For a power indicator, a low-current LED and series resistor  for 1-2mA could be fine. Use a fusible reistor for R2. Fusible resistors don't have a current rating, they are just specced to not catch fire if they pop.

[paul_g_787]

The 100uf cap is overkill.
Do you need as much as 10mA?
For a power indicator, a low-current LED and series resistor  for 1-2mA could be fine. Use a fusible reistor for R2. Fusible resistors don't have a current rating, they are just specced to not catch fire if they pop.

100µF is just one I had lying around. I also have 10µF would that suffice for C2?

The LED is a standard 20mA 2V LED. Running it at 20mA was a bit too bright but I found at 10mA it looks about right in a dark room.

If I replaced R2 for a fusible resistor would it be better to put it before C1?
And how would I spec the correct wattage of fusible resistor?

I notice in the regular LED light bulbs they just use 0.25W fusible metal film resistors, but they are before the dropper capacitor.

[tunk]

You could also take a look at this: https://youtu.be/Q23uh7AjjXw

[paul_g_787]

You could also take a look at this: https://youtu.be/Q23uh7AjjXw

Yes I did see Big Clive's video already That is where I got some of my information from.

I decided to add the smoothing capacitor as the flickering (although minor) was annoying and noticeable in the dark.
And I came across the idea for the Zener Diode just to help protect the circuit from High Voltage spikes on the mains.

[james_s]

Wouldn't it be easier to just replace it with another neon lamp? This LED retrofit is a lot more complicated with many more potential failure points, for what benefit?

[paul_g_787]

Wouldn't it be easier to just replace it with another neon lamp? This LED retrofit is a lot more complicated with many more potential failure points, for what benefit?

Perhaps it would be, but then I wouldn't learn anything that way

The benefit... Education.

[TimFox]

In general, the 100 uF capacitor to reduce flicker would be more effective if it saw more series resistance feeding it.  In your drawing, on positive cycles after the bridge, it is almost shorted out by the forward current of the LED (or the Zener).  You could put the 100 uF directly after the bridge, reduce the existing series resistor and move the difference to between the 100 uF capacitor and the diodes.

[Benta]

If you think that a major circuit for lighting an indicator lamp is great, fine with me.
I'd go for the neon lamp plus resistor every time.

Anyway: your circuit is not that bad but is has some flaws. The input circuit is OK, provided the bleed resistor is a high voltage type (eg, Vishay VR37), and the series resistor is current pulse resistant (eg, wire wound).

Bridge rectifier is also OK. But from there on it goes downhill.

1: why the 2 V Zener? (where on earth did you find that one?). LEDs are not voltage controlled, but current controlled.
2: 100 uF is way overkill, but you've already explained that point.

A good design would be, say 4.7 uf, a 4.7 V Zener and a series resistor for the LED to set the current.

But I'm not a friend of cap-drop powering of anything. The only isolation you have in this design is the resin of the LED.
On top of that, I believe this forum has a ban on cap-drop mains powering.

[David Hess]

Do not really need the bridge rectifier except for efficiency.  Keep the anti-parallel diode to protect the LED from reverse voltage, or use an anti-parallel dual LED.

[mikeselectricstuff]

Do not really need the bridge rectifier except for efficiency.

But you do at least need an inverse-parallel diode across the LED.
You don't need the zener - the high impedance of the dropper cap makes it effectively a current-source

[paul_g_787]

Do not really need the bridge rectifier except for efficiency.  Keep the anti-parallel diode to protect the LED from reverse voltage, or use an anti-parallel dual LED.

Initially I tried it with an inverse diode and also an inverse LED. However the flickering was very obvious when using both of these methods.

Considering the 1N4007s were less than £0.01 each and the LED has no noticeable flicker with it, I think the full bridge rectifier is perfectly justified.

[paul_g_787]

1: why the 2 V Zener? (where on earth did you find that one?). LEDs are not voltage controlled, but current controlled.

All the store-bought LED light bulbs I dismantled use this same basic layout:

Capacitive drop with bridge rectifier and smoothing capacitor. and a Zener diode in parallel with the LED(s). The only thing they had which mine didn't is a fusible resistor which I need to add before the capacitor.

[Zero999]

Orange LEDs have poor efficiency. Use a phosphor converted LED, which consists of a blue/violet/UV LED with a fluorescent material, to produce orange light. Unfortunately, the only ones I can find are these, which can only be bought from Rapid Electronics, in the UK. TME also sell them, but they no longer sell to individuals in the UK.

http://www.optosupply.com/newproduct/detail.asp?id=1898

The orange peach colour should be fairly close to a neon lamp, especially if it has a filter.

https://www.tme.eu/en/details/oscg4l5111a/tht-leds-5mm/optosupply/
https://www.rapidonline.com/Catalogue/Product/55-9177

The LEDs linked to above should give a nice bright light, at nor much more current, than a neon lamp. Use an SMT bridge rectifier and two 68k current limiting resistors in series to drive it.

[analogMensch]

Make sure your C1 is a X2 type, so it can handle pulses on mains.

I build my mains driven LED status lights also like this with a FBR, but I use a 5.1V zener, 25uF smoothing cap and a 2k or 4.7k resistor in series with the LED. Always works like charm!

[Benta]

@Zero999, has your account been hacked?

[paul_g_787]

Make sure your C1 is a X2 type, so it can handle pulses on mains.

I build my mains driven LED status lights also like this with a FBR, but I use a 5.1V zener, 25uF smoothing cap and a 2k or 4.7k resistor in series with the LED. Always works like charm!

Yes it is an X2 type 

Have you got a schematic for one of your ones please so I can see?

[james_s]

@Zero999, has your account been hacked?

Why do you keep responding to threads like this? It's starting to sound like paranoid delusions. What makes you think his account was hacked? Because he provides some links to the places where he bought some of the LEDs he is describing? There is no rule against doing that, it's not like Rapid Electronics is some random sketchy Chinese company.

[Terry Bites]

Yes the 100uF is a bit ott. Make sure the 100n is class X (don't say no one told you) and R2 is a fusible resistor. The diode bridge is also overkill. A reverse polarity diode acros the LED is sufficient. Better still buy a new neon or a ready rolled mains rated LED panel indicator. Trawling your options is highly educational.

[floobydust]

OP your circuit is very close to LED nightlights. Actually difficult to design.
The input needs a say 100-220R flameproof, fusible resistor that tends to fail open if you power up at a mains peak, the inrush is quite high.
The dropper capacitor is generally huge for ones rated for the application.
It's better to use the zener at say 9V or 12V and then a resistor to drop that down for the LED. You can use a much smaller cap and get no flicker, compared to being right across the LED. The electrolytic cap's ripple current is high and can shorten life.
Without any clamp zener, I had an LED fail open-circuit and the capacitor exploded 

What I do to replace neons, if flicker is not a problem, is a resistor and LED with reverse-diode across it. NOT SERIES diode, that fails fast.
For no flicker, a bridge rectifier is good enough for 100/120Hz harder to notice, and just a resistor for the LED. Even 1mA can be quite bright with high efficicency LEDs.

[james_s]

On top of that, I believe this forum has a ban on cap-drop mains powering.

Since when?

Maybe you're thinking of one of those weenie forums where they don't even let people talk about anything mains powered.

People have discussed far more dangerous things than transformerless power supplies here.

[tooki]

But I'm not a friend of cap-drop powering of anything. The only isolation you have in this design is the resin of the LED.
On top of that, I believe this forum has a ban on cap-drop mains powering.

Just because people often reply that capacitive droppers are a bad idea (or outright refuse to help someone who asks how to do it) doesn’t mean the topic is banned. The reason people often refuse to provide the information is that the people asking about them are generally obviously not skilled enough to do it safely. (Like how it’s invariably absolute newbies who want to look at mains with their new oscilloscope.)

In terms of content, the forum really only has two or three rules, which is that it has to be about electronics and may not be political. Blatant piracy is also not tolerated, though that’s really a gray area if you consider how much discussion there is about unlocking device features. :p

[Zero999]

@Zero999, has your account been hacked?

Why do you keep responding to threads like this? It's starting to sound like paranoid delusions. What makes you think his account was hacked? Because he provides some links to the places where he bought some of the LEDs he is describing? There is no rule against doing that, it's not like Rapid Electronics is some random sketchy Chinese company.

Exactly.

I haven't bought those exact LEDs, but have purchased similar items. It's true, phosphor converted LEDs are often more efficient, than those which directly produce the required wavelength for colours in the middle of the visible spectrum: yellowish green, to orange. This is more likely applications, where the higher volt drop is a non-issue and using a lower volt drop device would just result in more heat in the resistor/linear regulator.

[floobydust]

I appreciate new LED colours- no need to give someone flak for that. Kingbright offers some as well.

Overall, I find mains capacitive droppers simply don't last. For LED nightlights, you can do LTSpice sims and see the peak currents are high which shortens the life of almost all the  components including the LED. I've taken apart quite a few (that have regulatory approvals). You need a fuse, flameproof resistor, dropper-rated capacitor and then a scheme for low flicker. The one where the electrolytic cap exploded stunk up the house.
I've also put three LED's in series for higher brightness.

[David Hess]

The well designed ones I have encountered lasted longer than the neon bulbs they replaced, but I prefer the look of a neon bulb.

A zener shunt can be used to limit the voltage across an LED and current limiting resistor.  The demands on the voltage dropping capacitor are severe so a high quality part is required for reliability.