Advice on mains current sensing, safely

[GMB]

I have now managed to find my power meter and so can say that the consumption does indeed vary considerably with picture content - and doesn't come close to the rating plate figure of 352W. I guess it might get up to that if we turned the brightness to maximum and plugged in a load of hard drives into the USB ports.

I also found that the power factor is terrible on standby and not great at full power.

So it looks like the current tends to be in the 0.25-0.5A region most of the time, so not too bad. But the threat is there of much higher surge loads I guess. Of course I haven't measured it driving its loudspeakers.

It also seems that 0.16A of that is used by the CPU because when you turn it off the current drops to 0.16A for quite a while - doing an orderly shutdown I guess - and then drops to 40mA. The TV pretends to be "off" while this is going on.

[Ambientnoise]

Yes, that’s around what mine does but it’s likely to rise whilst updating channels, schedules, software etc even though it’s apparently OFF, usually at 2am or so. In my case, it takes just about the same as ON with a dark screen during these times. Hence the problem of discrimination using current.

[GMB]

Ours doesn't seem to automatically do stuff as it sometimes mentions the channels are all wrong but never tries to sort it (unlike the recorder box).

It also doesn't get to use the Internet as I do not trust modern things not to be evil in their spare time (or get hacked).

[Radio Wrangler]

Going back a bit, a current transformer would not be a huge risk. In the event of a short circuit and a huge fault current, the core will saturate. Transformer action will hit a ceiling. The primary inductance will vanish to negligible and the voltage across the primary will essentially be due to the copper resistance. Passing one mains core through one of those 'lollipop' wound cores for SMPS sensing ought to be safe. A pair of beefy diodes shorting the secondary each way as a clamp will protect stuff downstream.

This will tell you if the telly is active. Add a timer function to mask out any overnight updates. You won't want the sound on for them.

Alternatively, just use receiver to tell when the set is radiating EMC

David

[GrimJosef]

Originally Posted by Radio Wrangler

... A pair of beefy diodes shorting the secondary each way as a clamp will protect stuff downstream ...

If we moved the pair, or maybe more than one pair, over to the primary then we'd have protection for that as well. I can't think of a good reason not to.

Any timer would have to run for minutes, I'd have thought, if it was going to block overnight updates. If it were my telly I think I'd want the sound to come on more quickly than that when I was sitting down for Strictly/MOTD/The Queen's Xmas Message .

Cheers,

GJ

[GMB]

I am not seeing a problem with the set powering up for its own private reasons overnight, because presumably there would be no audio being generated? There is certainly no audio after "switching off" when from the power consumption I can see the set is clearly still computing away.

[Bookman]

Originally Posted by kalee20

Current transformers don't mind a short on the secondary, but they object to being open-circuited.

I'd suggest a CT with permanent load consisting of a bridge rectifier feeding an electrolytic capacitor and your 24V relay coil.

If your TV takes 400W, that's about 1.6A, and the relay needs 50mA, then you need a current ratio of 32:1.

So, I'd suggest a 240V:9V mains transformer, (ratio 26.7), rated at about 20VA. Connect the bridge rectifier to the 240V winding. Connect the 9V winding in series with your TV. It's now working as a current transformer. And the insertion drop will be around 24V / 26.7 which is less than a volt.

And of course, you get your isolation. Twice, once in the transformer, once in the relay.

Kalee is correct in that you cant really short out a current transformer. You could however short out turns but its only impact would be its effect upon primary to secondary turns ratio and its corresponding impact upon what is being measured.
As stated the problem you would encounter are open circuits where high Voltages can be measured across the terminals. However, such situations apply more to high impedance secondary circuits used on advanced protection circuits costing many thousands of pounds.
Even then there is a simple arrangement to protect against excessive Volts and that is to use a metrosil. In this one would calculate the appropriate Voltage requirement which I dare say in your circuit would probably be lucky to attain double digits compared to say 1000 Volts in a protection circuit.

[G6Tanuki]

I still think a reed-relay with a few turns of wire wrapped round it would be the easy way. Alter the # of turns to get the required sensitivity. Fine-tune it by using an external bar-magnet to 'bias' the reed.

[Something similar - with 2 opposing coils - was used on vintage Volvos for the bulb-failure-warning circuits]

[GMB]

Originally Posted by G6Tanuki

I still think a reed-relay with a few turns of wire wrapped round it would be the easy way. Alter the # of turns to get the required sensitivity. Fine-tune it by using an external bar-magnet to 'bias' the reed.

[Something similar - with 2 opposing coils - was used on vintage Volvos for the bulb-failure-warning circuits]

If it really works with just a few turns?

I thought the amp-turn ratings were in the 10-30 range so that's more than a few turns - it will be in the hundreds (All the relay coils I have seen for reeds are not so small).

[G6Tanuki]

The ones I remember [from Volvi] had about 30 turns per winding - they were feeding 12V 21-Watt bulbs - something like 1.5A - and wound with quite thick wire [22-gauge?] - you don't want significant voltage-drop when you only start off with 12V.

The reeds were about 5mm diameter and 20mm long.

[GMB]

about 30 turns per winding - they were feeding 12V 21-Watt bulbs - something like 1.5A

Quite. So this equates to about 300 turns in my situation. But still needs thick wire. No so few.

[G6Tanuki]

Why does it need thick wire if it's only passing a couple of amps max?

[ Think of thinner wire in this application as representing an 'embedded fuse' ]

[GMB]

By "thick wire" I mean thick enough to carry a couple of amps inded.

I do not want an embedded fuse! I want it not to catch fire in a fault situation.

[crackle]

Turn it all of at the mains.
Mike

[McMurdo]

I used a reed switch with ecw wound around it to sense the lamp current in an antibacterial UV lamp based water purifier supplied to the OEM. Obviously the turns depends on the sensitivity of the reed switch and the number of turns you give it. You can get reed switch formers just for this application but I just wound the wire round the glass in this low voltage application. For high current applications you use thicker wire with fewer turns and for low current you can use thinner wire with more turns, which is convenient!

I'd have thought a modern smpsu in a telly would give a kick of current at start-up, settling down to a lower level, ideal to bring a reed switch in and hold it. Then on turn off the low standby current mandated by EU efficentcy directives would hopefully be less than the hold-in current of such a relay.

What you do with the reed relay contacts is up to you!