High voltage supply current limit to not be hazardous live (IEC 61010-1)

I believe to achieve this, the capacitance on the supply output needs to be under 300nF (Figure 3, 6.3.1.c.1),

Agree. This allows a short duration higher current at contact, as the body takes time to realize it is being shocked. In effect this is a transient energy limit. Only valid if the caps are not refilled fast enough to violate the current limit condition.

and the current must be limited to 2mA (6.3.1.b.1) under normal conditions. Under single fault conditions these limits rise.

agree This is to do with muscle lock and startle responses.

For the current to only be 2mA through a 2k resistor, the voltage of the supply would be pulled down to 4V. This is well under the 60V limit for DC voltages. Does this mean that the 2mA limit no longer applies here?

No,  you are not looking at this bit the right way ;-)

potential victims of electrocution may be assumed to have any resistance higher than 2k - that lower value is  a large area wet skin contact.

Now if for example you had a 150V supply with 73k in series internally, the limits are met,  but note that the highest dissipating shock victim is not 2K but 73k also.

Clearly for the low resistance victim, the not quite short circuit 2mA limit dominates, and the voltage falls to a few volts, yes, BUT now let us assume the same victim makes a more casual contact, and his or her resistance is  a few tens of k, even so we do not want the voltage to rise to a level that will charge those caps to a dangerous level...

Think of the 2k test as you would a short circuit test.of the internal current limit, without the zap of screwdriver on capacitor terminals,  but a shock victim of any impedance must always be limited to 2mA or less with the caveat that  if you can be sure the open circuit voltage is always safely below 60V  then that current limit requirement does not apply.

It may be this is not the best way to clarify, I already know what it means - feel free to say so if it makes no sense !

The usual way to meet this sort of thing for e.g. an insulation tester, is with a proper current limiter circuit in the output to fold back the volts as the 2mA is approached. The most serious single fault is the failure of that current limit.

Mike.

Secondly, the capacitors on the output of the supply will discharge through the 2k resistor leading to an initial spike of current as they are pulled down to 4V. I can't see any way to avoid this, or any limitations placed on this within the standard.

You've already mentioned the limitation - limit the output capacitance to no more than 300nF.  The smaller the capacitance, the less of a jolt that a person will get if they stick their fingers into the output of the power supply.

Hi Mike,

Thanks very much for the response. So if I understand correctly, ultimately I have to limit the output current of the HV supply to be <2mA when pulled down to 4V. The output impedance of the supply must be high enough, or have some other current limiting circuitry.

This is a little tricky as the power dissipation of 1mA at 150V (my applications demand) is a lot more than the power dissipation of 2mA at 4V. I'm limiting the inductor current in the switching boost circuit, but that won't be enough as the same inductor current which provides 1mA at 150V, will provide a lot more than that at 4V. But if I use a 73k series resistor, the voltage drop across it whilst drawing 1mA for my application will be unacceptable. I think will need a sense resistor and feedback circuit to monitor the current and stop the boost converter if >2mA is ever detected at any voltage?

At the risk of pedantry, BS 7671 considers anything below 1500V DC (or 900V DC to earth) to be low voltage.

Right, thanks. Yes that was my suspicion.

yes - given your full load current is u to 50% of the limit, you will almost certainly need to limit, ether as a direct current limit at 150V, - some modern transistors are good enough for this, 

]

a couple of options of that kind of thing there any old 10V zener and HV transistor would do. can be installed in series hi-side or lo-side.

or by sensing and folding back the drive smartish,  if the transformer is double wound you can sense in the groundy side but watch what is referenced and decoupled  to what, and note that the feedback voltage goes more negative as load increases.

we use this latter for EHT supplies  100kV plus where there is no chance of a series limiter. When the sense voltage goes negative the HV has been pulled down to hard... There are a few extra components to protect the 'op-amp' as well.

Ideas only, I have no idea of your application or what is really sensible in your specific case.

If there are no other loads you could limit maybe just  on primary current.

Mike.

Agreed. It does not help when some manufacturers (Say bathroom lighting for example) state their 12V lamps as "Low Voltage" and infer mains voltage to be not low voltage. We all know that Low voltage is utilisation voltages and high voltage is transmission or distribution voltage but the man in the street thinks low voltage means arounabout car battery voltage

car battery voltage

We may need to revise our ideas of that too - given the introduction of EVs...

   - Andy.

Indeed 

Jeremy Leaf  Do we assume this product is to be used in dry condition only? Also, what is the assessed possibly contact area size?