I understand the user here so that fuses do not protect people from shocks. I am planning to put put fuses such as fast-blown fuses and PPTC fuses around all kind of positions such as computerUSB-microprocessor-link and wet-electrode-circuit link. I feel it is far better to blow up fuses than to blow up computer or get shocks. Now the user does not recommend any substitute to protect from shocks so how to protect from shocks? Not with fuses? What then? I operate with EM the level of coming from MBA.

Further reading

  1. EOS versus ESD -thread by TI

5 Answers 5


You question is really asking about a bunch of different failure scenarios and without more detail (really, more specification) it is difficult to answer in any detail:

Too much voltage (temporarily)

You need a "transient voltage suppressor". They come in many forms and scales from large consumer products to build-your-own with two diodes.

Too much voltage (constantly)

You need to reduce the voltage. There are basically two categories of solutions "regulation" and "drop/unregulated".

Regulators are available in everything from commercial switch-mode power supplies to individual IC's.

Unregulated voltage drop can be achieved with just a properly biased diode, a resistor, a step-down transformer, and many other approaches

Too much current (temporarily)

Current is a bit trickier since it is controlled by the load (and anything shorting it). Fuses and other energy threshold based devices will be too slow to react in the general case.

You will probably need to "slow" the current transient down (reduce edge-rate), by applying inductance in series, to buy time for whatever protective devices you employ to activate.

Current transience is best handled by active devices (sense-and-respond) -- just my opinion -- since you usually have a rather complex set of criteria to distinguish between the "good" and "bad" cases.

Too much current (constantly)


  • 14
    \$\begingroup\$ ... or his name wouldn't be Dr. Fried Parts. \$\endgroup\$
    – Kaz
    Apr 22, 2013 at 18:03
  • 1
    \$\begingroup\$ +1 thank you for your input, put a lot of random things together that I had not associated before -- just used them. \$\endgroup\$
    – hhh
    Apr 22, 2013 at 19:54

Overheating a component sufficiently to start a fire requires that either a moderate amount of excess power be delivered it for a comparatively long time (multiple seconds), a really huge amount of power be delivered (in which case the time may be shortened by a factor of a thousand). Fuses are generally intended to protect against fires, and so they will blow quickly when current massively exceeds specifications (e.g. by a factor of fifty), but will blow slowly when the excess is smaller (e.g. a factor of only two). If someone were to touch a circuit guarded by a 10mA fuse in such fashion as to allow 100mA to flow across the chest, the time required for that current to disrupt heart function might be less than the time required to blow the fuse.

Avoidance of electrical shock requires that there isn't any path by which any significant current could be pushed through a person, outside of fault conditions, and that any fault condition--even a minor one--must cause power to be disconnected quickly. A common arrangement is what's called (in the US, anyway) a Ground Fault Circuit Interrupter. I think some other countries use the term Residual Current Detector. Such a device measures the difference between the rate at which electrons flow to a piece of equipment via the supply wire and the rate at which they flow back via the return wire. A sufficient mismatch will cause current to be disconnected immediately. Unlike a fuse, which will only blow immediately under a severe fault condition (blowing slowly under lesser conditions), a GFCI will trip immediately under just about any fault condition which will cause it to trip at all.

  • \$\begingroup\$ eBay returned bulky things not things to use with microprocessor and super-sensitive electronics. Is the Ground Fault Circuit Interrupter meant to protect houses? Is there something like that used to protect sensitive circuits? I operate with EM level that is coming from my laptop, about 0.1-0.5A. \$\endgroup\$
    – hhh
    Apr 22, 2013 at 17:36
  • 1
    \$\begingroup\$ @hhh -- I'm guessing that you don't see small DC-GFCI because measuring small currents that accurately is difficult to do reliably and cheaply and installing sense elements in the paths results in supply isolation (ultimately resulting in noise). \$\endgroup\$ Apr 22, 2013 at 18:05

Hospital gear in the US is often tested to the NFPA 99 standard for health care facilities. This standard involves certain no-brainers like is the device double insulated, is the power cord good, etc., but it also tests devices under a variety of failure modes, often by pulling the ground pin, and looking at a variety of leakage currents. As the level of exposure goes up, the amount of leakage current with a ground pin goes down. For example, in ICU's where a patient my have a variety of catheters filled with conductive solutions going right into the heart, the limit for leakage current for devices with patient leads is tens of MICROamps!

Many situations are forced to test to different standards. IEC standards are fairly common. See IEC 60335 and some of the others on http://en.wikipedia.org/wiki/List_of_IEC_standards. If your plan allow your device to test to such standards, you're good to go. Unfortunately, getting many of these standards is a fairly expensive proposition, and often one standard points to many more, so the expense can snowball without a subscription service.


Not really a "complete answer" but helpful:

Two important terms to look up:

galvanic isolation - which is preventing current to flow by eliminating metal paths between places.

optical isolation - Using an *optoisolator *which is using essentially an LED and a light detector, separated from each other.

Both of these things/methods make it so voltage/current transients - even from things such as arcs, short-circuits or even lightning strikes cannot cross a given point. Such things/techniques are used to isolate both circuits from teach other - or people from circuits which could hurt them should something go wrong.

  • \$\begingroup\$ Not exactly sure - can you perhaps describe how/why you are trying to avoid the user from shocks - i.e. how would the user come into contact with your circuit. By touching the box, or are they "connected" to anything. What do you have on your high-voltage side? \$\endgroup\$
    – Brad
    Apr 30, 2013 at 11:57

I like DrFriedParts's points but one thing missing: physiological aspects.

Non-harmonic Frequency

The hearth rate is about 60-200 bpm so about 1-3 Hz, depending on your activity. The maximum hearth rate depends on a few things such as your age and your physical status. Now suppose a developer has HR about 1.5 Hz when he shorts AC. Now too high frequency let say 100Hz is less lethal than let say non-harmonic frequency -- let say 1.5Hz when HR is 1Hz -- ceteris paribus because the hearth pretty much stops with 100Hz while the non-harmonic frequency can do physical damage to the developer's body. So too high voltage or too high current may be less lethal than a EM of specific frequency.

Fuses do not stop the developer from getting the lethal frequency. You can find ways to protect against lethal frequencies here.

  • \$\begingroup\$ He was discussing methods and why fuses were not a great solution. There is still more to it then this. I think you have to accept at some point that you can just keep adding complexity. How long you are electrocuted for will cause breakdown and a change in current and increase in damage also... \$\endgroup\$
    – Kortuk
    Apr 22, 2013 at 19:12
  • \$\begingroup\$ @hhh You're asking another question in the last portion of your answer, which isn't appropriate as an answer to the original question. You should remove your question and post it as a separate question. \$\endgroup\$ Apr 22, 2013 at 20:08
  • \$\begingroup\$ @Madmanguruman Thank you, moved that question here. \$\endgroup\$
    – hhh
    Apr 22, 2013 at 20:11
  • \$\begingroup\$ If you think there is something wrong, say it. Question related to this was in an official medical exam audited by many medical professionals. I may use wrong terminology but it is should be perfectly clear what I mean here: selecting specific frequency can be much more lethal can some other factor let say voltage. \$\endgroup\$
    – hhh
    Apr 22, 2013 at 20:15
  • 2
    \$\begingroup\$ There is no lethal frequency, but watch out for the brown note. \$\endgroup\$
    – Kaz
    Apr 22, 2013 at 23:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.