Safe? Mains to a CFL fitting and bulb

[davelx]

One thing you can guarantee is that whatever is inside one of those CFLs will be the cheapest nastiest set of components they can get away with...

Interestingly, the price of all the 1N400x range at Farnell is the same - but I guess if you are buying 1000000's of them there may be a saving.

ericmark, what brand was the one you showed? I'm tempted now to open one up and 'scope it

 

[JohnW2]

Interestingly, the price of all the 1N400x range at Farnell is the same - but I guess if you are buying 1000000's of them there may be a saving.

I hadn't looked, but that doesn't surprise me. Indeed, for the ones that come in identical packages, as with an awful lot of electronic components, what end up as specifically-rated components often all come off the same production line and then are 'graded' and the appropriately labelled according to their characteristics on test - hence no true reason for price differentials. The other factor of course, is that with components as cheap as IN400x diodes, the materials and manufacturing costs are probably an almost negligible part of the total price - so, again, no reason for appreciable price differentials.

Interestinglyericmark, what brand was the one you showed? I'm tempted now to open one up and 'scope it

I've taken a few apart in my time, and all seem virtually identical - albeit the electronic components are sometimes not (meaningfully or at all) labelled.

Kind Regards, John.

 

[Gnomer]

Thanks everyone.

I think I've got more of an idea what's going on.

I have changed a bulb before but I haven't ever made my own lamp, straight from a wall socket. It just feels to me a bit awkward, I guess.

This, and your first message, are not particularly clear about what you are trying to do.

If you feed a light from a wall socket you need to check first that the socket is capable of supplying another feed hardwired into it. Be aware of adding to spurs and sockets and junction boxes already supplying a spur on ring circuits.

From the socket wiring usually 2.5 mm2 cable will supply a fused connection unit, typically with a 3 or 5 amp fuse. This can be switched or unswitched. From here 1.0 or 1.5 mm2 cable to the light, and light switch as required. All bare earth wires sleeved with green and yellow sleeving, and all connected together.

Or are you intending on just plugging the light into the wall socket??

Apologies if I have misunderstood your posts - they don't make a lot of sense.

Yeh it's a 11W? Or something, Tesco CFL bayonet with a fitting, I am plugging into a three pin BS 1363 standard wall socket, using a male 13A fused plug, using wire which has blue, brown and green/yellow wire in it.

Untitled

• Gnomer
• 26 Jun 2011
• 1

 

[EFLImpudence]

At the plug -

Brown to Live
Blue to Neutral
Green/yellow to Earth


At the lamp -

Brown & Blue to bayonet fitting. It doesn't matter which way round.
Green/yellow to Earth if there is one (metal fitting). If there isn't one (plastic fitting) put connector on the end and tuck out of the way.


Edit - Didn't see your drawing - that's right but use 3A fuse.

 

[davelx]

Interestinglyericmark, what brand was the one you showed? I'm tempted now to open one up and 'scope it

I've taken a few apart in my time, and all seem virtually identical - albeit the electronic components are sometimes not (meaningfully or at all) labelled.

Kind Regards, John.

Just under 350V peak on the non-conducting 1/2 cycle. Mains voltage here is 245V so that's about what I'd expect.

- will post a photo of the waveform when it gets dark. It's a Philips 11W lamp and uses 1N400? bridge with a 2.8uf/400V electrolytic on the output. I can't see which 1N400? part it is as the vital number is against the PCB on all four diodes.

 

[JohnW2]

Just under 350V peak on the non-conducting 1/2 cycle. Mains voltage here is 245V so that's about what I'd expect.

Thanks. They surely can't be trying to get away with less that 400V PIV, then; I'd personally probably use 1000V, or at least 600V.

- will post a photo of the waveform when it gets dark. It's a Philips 11W lamp and uses 1N400? bridge with a 2.8uf/400V electrolytic on the output. I can't see which 1N400? part it is as the vital number is against the PCB on all four diodes.

I don't suppose you have a high enough AC voltage to hand to measure the PIV, do you?

Kind Regards, John.

 

[wcavanagh]

I took one apart once for a project, and it had HER208G diodes in, so 1000V

 

[davelx]

- will post a photo of the waveform when it gets dark. It's a Philips 11W lamp and uses 1N400? bridge with a 2.8uf/400V electrolytic on the output. I can't see which 1N400? part it is as the vital number is against the PCB on all four diodes.

I don't suppose you have a high enough AC voltage to hand to measure the PIV, do you?

Kind Regards, John.

I have a huge power transformer from an old Cossar valve 'scope which I've been using for a battery charger ( it can deliver about 20A from the filament windings) which has some decent HT windings, but I lack a variac to control the input in order to ramp up the applied voltage and plot the breakdown curve.

I'll remove one of the diodes to see what type it is.

I've got a photo of the 'scope trace, but it's exactly what you'd expect from a normal full-wave bridge so not a lot of point posting it.

 

[JohnW2]

I have a huge power transformer from an old Cossar valve 'scope which I've been using for a battery charger ( it can deliver about 20A from the filament windings) which has some decent HT windings, but I lack a variac to control the input in order to ramp up the applied voltage and plot the breakdown curve.

You don't really need to do that. If you have a transformer with an HT winding (or combination of HT windings) which has a peak voltage greater than the diode's PIV, just (carefully!) connect the diode in series with a high value resistor across that source and look at the voltage across the diode on your scope - the reverse breakdown voltage should then be obvious.

Kind Regards, John.

 

[davelx]

I have a huge power transformer from an old Cossar valve 'scope which I've been using for a battery charger ( it can deliver about 20A from the filament windings) which has some decent HT windings, but I lack a variac to control the input in order to ramp up the applied voltage and plot the breakdown curve.

You don't really need to do that. If you have a transformer with an HT winding (or combination of HT windings) which has a peak voltage greater than the diode's PIV, just (carefully!) connect the diode in series with a high value resistor across that source and look at the voltage across the diode on your scope - the reverse breakdown voltage should then be obvious.

Kind Regards, John.

Ah yes. that would work, of course, so long as I choose the resistor to limit the current to avoid letting the magic smoke out. I've got a few multi-megohm resistors and I think I've got some with a high voltage rating.

Will see what I can do.

 

[JohnW2]

Ah yes. that would work, of course, so long as I choose the resistor to limit the current to avoid letting the magic smoke out. I've got a few multi-megohm resistors and I think I've got some with a high voltage rating. Will see what I can do.

The maximum (Full cycle) reverse current on IN400x is about 30 μA - so something like 13MΩ (probably 10MΩ or 15MΩ in practice) for 400V or 33MΩ for 1000V ought to be fine (given that you'll probably be talking about far less than a full cycle, if you choose a reasonable voltage), and power rating of the resistor is not going to be an issue.

Kind Regards, John.

 

[davelx]

Can't seem to find my high value R's. But I popped one of the diodes out and they are 1N4007 (1000V PIV), so Philips at least are using an appropriate component.

 

[JohnW2]

Can't seem to find my high value R's. But I popped one of the diodes out and they are 1N4007 (1000V PIV), so Philips at least are using an appropriate component.

Thanks. Ah well, no surprises, then; as I said, 1000V is probably what I would have chosen.

Kind Regards, John.