Archived Forum

I have not tested the -000V, however I would assume that, if YHDC follow the same practice as at least one other manufacturer, the burden resistor is chosen on test to give an accurate calibration. The downside is, of course, that you lose the flexibility to choose your own burden resistor (within the VA rating of the transformer of course) and hence increase the sensitivity.

This website http://yhdc.en.alibaba.com/product/519312318-218696501/13mm_hole_Split_c... is not credible - there's an obvious mistake/inconsistency - and I can no longer find the YHDC website, that appears to have disappeared behind the Great Firewall of China, so I can't look up reliable data for the -000V.

Thank you for the insights Robert!

Good point about losing the flexibility of choosing your own value, I hadn't thought about that.  I have emailed YHDC requesting the datasheet for the SCT-013-000V and I will post it here once they send it.

Thanks

Hi Robert,

Saw your comment regarding the YHDC website. I had no trouble bringing it up a short while ago. (0322Z)

An ISP issue, perhaps?

regards,

Bill

The response I received from YHDC was that they don't have a separate datasheet for the SCT-013-000V, but "the internal resistor is about 20Ω".

This makes sense since they are claiming the output is 0 - 1V.  (2000:1 turn ratio on 0-100A is 0-50mA, and 0-50mA with a 20Ω burden resistor results in 0-1V output)

All this said, I'm going to take Robert's advice and stick to the SCT-013-000 model in case I ever have a need to change the resistor value in the future.

Thanks,

John

Hmm. The website has reappeared. Maybe Google trying to be too clever?

If you do buy the "V" version, you should find it easy enough to remove or change the internal burden. If you want an external burden, e.g. the one on the emonTx PCB, then it's important you replace the internal resistor with pair of nose-to-nose zener diodes. Any voltage above 5.6 V up to their 22 V should be OK, they are there to prevent painful if not dangerous voltages if the c.t gets unplugged whilst energised.

RW: Any voltage above 5.6 V up to their 22 V should be OK,

... for a person, yes, but I'm surprised that no processor of mine has ever become noticeably damaged by over-voltages.  Our 8kW shower must cause the CT's output voltage to stray well outside the processor's power rails.  My original Mk2 schematic includes a couple of protection diodes, but the standard circuit on the emonTx doesn't.

The internal protection of the Atmega 368P seems to be very robust.  I'm wondering whether anyone has ever managed to damage one of these analogue inputs by over-voltage from a CT?

Yes, but the Atmel processor's internal protection diodes can just about handle the current available from the c.t. when it is limited by the bias network (which is in series of course). If you were to (possibly misguidedly) increase the size of the capacitor, or derive the bias from an op-amp buffer itself fed from a higher voltage supply and with a high current output, I would suspect the processor might reach its end of life prematurely!

Good point about the bias network being in series with the CT.  When using the standard emonTx components, the "reference" end of the CT is connected to the power rails by 470K resistors, so there's nothing "firm" for any excess voltage to push against, except the 10uF capacitor.  Hence there's probably no need for any additional protection components (especially when using an low-valued burden resistor, such as 18R).

If the "reference" end of the CT is connected to a buffered supply (which is like having an infinitely large capacitor), the "processor" end will be able to deliver more current during overload conditions.  Additional diodes for tying the processor's inputs to the supply rails are therefore more relevant in this case.

Note to self: When using a burden resistor of 150R, and a buffered reference, don't forgot those protection diodes!