SMPS safety isolation taken a bit too far?

[zenerbjt]

We are running 270W from a 100V battery. We therefore need a transformer isolated SMPS to isolate the downstream 48V loads from the battery. But do you agree that we can in fact connect primary and secondary grounds since any disaster will not breach the transformer anyway?

But even then…..48V can kill just like 100V……so why do we even bother with safety isolation?

Does anyone have a video showing current in a 3ph BLDC?
I take it the attached is wrong?

 

[biswaIITH]

We are running 270W from a 100V battery. We therefore need a transformer isolated SMPS to isolate the downstream 48V loads from the battery. But do you agree that we can in fact connect primary and secondary grounds since any disaster will not breach the transformer anyway?

But even then…..48V can kill just like 100V……so why do we even bother with safety isolation?

Does anyone have a video showing current in a 3ph BLDC?
I take it the attached is wrong?

To answer your first question , it is mandated that if the source is more than 60Vdc ,you need to have isolation(mandatory)for electrical safety .
You have to use isolated DC-DC (with insulation resistance more than 100 ohm/V or 500ohm/V )to step down below 60V to use in your circuits.

You can refer to ECE R100(Protection against electric shock) for more details.

 

[KlausST]

Saftey isolation has nothing to do with 270W or power in general. (Only in some rare exceptions)
It has to do with voltage.
And it has to do with "not to harm humans"

And whether you need isolation or not does depend on your own decision.
You may build any system non_isolated as long as you protect persons otherwise.

48V DC is considered as not harmful. It won't kill a person.

Klaus

 

[zenerbjt]

48V DC is considered as not harmful. It won't kill a person.

Yes...you know i am constantly being asked...what is the voltage that will kill?........i mean, the limit is 60V....so touch 59v and we will be safe?.....touch 61v and we will die?

 

[KlausST]

Hi,

It's not the voltage that is dangerous, it's the current.
And there are many parameters that influence if it will kill.

100mA may be dangerous. When AC, when the current crosses the heart. It may cause the heart to stop or to cause ventricular fibrillation.
DC more causes the water in the blood to be hydrolysed.
HF causes the skin to burn.

One person may survive 500mA, another may die even with 20mA.

Klaus

 

[Akanimo]

Hi,

It is more of a consideration of magnitude of current and duration of time rather than voltage or current separately. So the factor 'time' brings the amount of available energy also into the picture.

The magnitude of current flow through the body is what causes the harm, however, voltage is the force that drives the current and available energy is what determines how long the voltage drives the current. The voltage may be very high, but if the available energy is negligible, then the effect is negligible as well because it results in the flow of only an insignificant amount of current (through a certain resistance, call it "R") and/or a too-short flow time. Another consideration: if the voltage is safely low, then it can only drive an insignificant amount of current through resistance R and time whether long or short would not matter anymore, so the available energy does not matter too.

If the voltage is significant and the amount of available energy is significant too, then a significant amount of current can flow for a significant amount of time.

Yes...you know i am constantly being asked...what is the voltage that will kill?........i mean, the limit is 60V....so touch 59v and we will be safe?.....touch 61v and we will die?

Relate that in making a safety related decision, a safety margin is left. 60V is just the set limit. 61V may not be different from 60V in terms of effect with regards to the magnitude of current that flows through any given R, but when you exceed it, you violate the rule.

 

[zenerbjt]

Thank.....regarding the transformer isolated SMPS between 100vin and the 48vout......and suppose its pri and sec grounds are connected...then i still dont see how the 100v can ever end up at the 48v output....i mean the transformer would have to break over its insulation...very unlikely.

Surely its ok to connect pri and sec grounds?

...i mean, if the fault is so drastic that 100v somehow gets to the output side....then even of the pri and sec are totally isolated, then the 100v and its ground could still end up at the output......so surely a non isolated transformer based smps is fine for safety?

 

[biswaIITH]

Thank.....regarding the transformer isolated SMPS between 100vin and the 48vout......and suppose its pri and sec grounds are connected...then i still dont see how the 100v can ever end up at the 48v output....i mean the transformer would have to break over its insulation...very unlikely.

Surely its ok to connect pri and sec grounds?

...i mean, if the fault is so drastic that 100v somehow gets to the output side....then even of the pri and sec are totally isolated, then the 100v and its ground could still end up at the output......so surely a non isolated transformer based smps is fine for safety?

Hii,

The whole point of adding a transformer in SMPS is to provide a galvanic isolation between high voltage circuit(DC>60 V ) as well as Mains supply & output/load.
It is not recommended to short both primary & secondary grounds. If you do that ,transient/surge etc will find a common ground i.e. low voltage ground to flow. In some application, they short the chassis with low voltage ground which case you may get a shock while touching the aluminum enclosure.

It is a regulatory safety compliance for the user/customer if you want to sell the product in market. It also mandates the isolation required between high voltage circuit & low voltage auxiliary circuits(12V,15V,24V etc) , communication circuits such as CAN as well. I suggest you refer to IEC 61851 standard for EV charging (Just to understand the requirement).

You may design your own product in lab( just for testing )without following the above regulation of course, take necessary precautions for your own safety!!!

 

[FvM]

Thank.....regarding the transformer isolated SMPS between 100vin and the 48vout......and suppose its pri and sec grounds are connected...then i still dont see how the 100v can ever end up at the 48v output....i mean the transformer would have to break over its insulation...very unlikely.

Surely its ok to connect pri and sec grounds?

...i mean, if the fault is so drastic that 100v somehow gets to the output side....then even of the pri and sec are totally isolated, then the 100v and its ground could still end up at the output......so surely a non isolated transformer based smps is fine for safety?

The answer isn't that simple. You need to analyze possible faults, e.g. ground short of the +100V line or interruption of ground connection. Can you guarantee by design, that possible faults can't create a hazardous contact voltage at accessible parts? Depending on the applicable safety standard, the accepted contact voltage in single fault condition can be higher than in regular operation, e.g. doubled value. On the other hand it's halved for equipment intended to be used in wet locations.

All in all you may get away with non-isolated DC/DC for 100V battery input under circumstances.

 

[zenerbjt]

Hi,
We are doing another power supply which runs off a 14S LiPO (42-59V).
Vout is 24V and 130W.
This is under the SELV rating so i assume that we are not required to have transformer isolation?

However, is it true that one other reason for isolation is so as to prevent damage to the load in the case of a shorted FET?. For example, if a primary side transistor in a transformer isolated SMPS fails short, than the Vin doesnt get impressed across the load....whereas eg if a high side buck fet fails short, then vin goes across the load, potentially destroying it. So is this the "other" reason for isolation?.....ie other than safety?

 

[Easy peasy]

Isolation is also not required in situations where trained personnel are handling the circuits - for joe public - ouput must be SELV - this includes two ckts to prevent the o/p going high (UL - for North America )

 

[c_mitra]

Yes...you know i am constantly being asked...what is the voltage that will kill?........i mean, the limit is 60V....so touch 59v and we will be safe?.....touch 61v and we will die?

Good questions do not have good answers.

Just for basic information, even 12V can kill. Medical instruments are designed with extra precautions. Say an ECG or EEG machine.

The question is how does it work? What is the actual process responsible for the death?

If the voltage is rather high, say 1000V or so, the current will be so high that the body is going to get burnt. Dry skin is a reasonable insulator but not a great one.

If the voltage is low, say 10-50V, the skin can block current. But this is good only if the skin is dry. You can take a 9V battery and touch the two ends to your tongue; you can feel the current. But you will not feel the same sensation with a 1.5V cell.

If the skin is broken (say via an injection needle), current will find a path via the blood stream. Blood is a decent conductor and connects well to the heart.

The beating of the heart is a (sort of) relaxation oscillator. Another external signal can confuse the oscillator and may even force it to stop. Recall that for a sustained oscillation you will need 180 out of phase feedback.

The electrical pulses that make the heart beat are less than 1V and use currents that are less than 1mA.

So the answer is that the guidelines are for general use; your body should be dry. You should take all precautions.

 

[Easy peasy]

the limit is 60V....so touch 59v and we will be safe?.....touch 61v and we will die?

this is a great question - 60V evolved from the -48V telephone system - i.e. widely regarded as being safe for human hands ( although you can get a tingle with wet hands ) as with anything human related as you get further from the centre of the bell curve there are people who are unusually sensitive and are greatly affected by 60VDC - so the 60V is a compromise, some people can hold 230Vac with their very dry and thick skin - but I don't recommend trying it yourself ....

 

[zenerbjt]

Thanks for these great replies.
And may we confirm that there are three reasons for transformer isolation of an SMPS?......

1....Safety
2.....Load protection (FET failure in non-isolated supply could result in Vin being impressed across the load and damaging it.
3.....Electrical noise.......separation of grounds by a transformer can reduce noise pulse currents travelling through a sensitive control ground.

Possible reason number 4 seems to be that the most efficient resonant converters seem easier to do with a transformer in there......eg PSFB and LLC converters. Have you ever seen a control chip for a resonant Buck or Boost converter?....No?........

....we went through a resonant Buck in College (not quasi resonant type ), but i never saw it after that day, anywhere.

 

[Easy peasy]

generally low voltage buck & boost don't need to be resonant - as at low volts the difference in losses is minimal, - at higher powers and higher conversion ratios, transformer based designs work better - these are often resonant to allow high freq operation with lower switching losses ....