1
\$\begingroup\$

I realize SMD Chip Resistors are not recommended for Mains Voltage and it's no wonder since their maximum voltage rating is lower than 100 V. I also understand that WireWound Resistors are recommended for Mains Voltage due to their higher Voltage ratings.

But lets consider we put 6 RC0402 Chip resistors (which have a voltage rating of 50V) in series (in a line, so that proper spacing between the pads is provided). Is it as safe as using a 300V rated WireWound resistor (because 6 * 50V = 300V) as far as breakdown is concerned?

(I'm not asking about the power they can dissipate, lets say the current flowing is merely 1 uA and that the resistors are the component controlling that current. I'm only concerned about they being able to handle the voltage without breaking down.)

\$\endgroup\$
2
  • \$\begingroup\$ You can get SMD resistors rated for much more than 100 V. Look at larger package sizes; Bourns' CR series, for instance, is rated to 200 V in any size larger than 1206, Stackpole's RMCF the same, and Vishay's CRCW series is rated for 500 V in 2512 size. If you look for specially-designed high-voltage resistors, Bourns' CHV series and KOA Speer's HV73 series are rated for 3000 V in 2512! And Ohmite even has an extra-large 4020 size entry in its HVC series that's rated for 6000 V. \$\endgroup\$
    – Hearth
    Commented May 22 at 14:45
  • \$\begingroup\$ Of course smaller packages will limit your voltage rating, simply due to arc potential. Even if everything inside the resistor is perfect and can withstand infinite voltage, you'll still get arcing through the air (or conformal coating, or potting compound) between the terminals eventually. \$\endgroup\$
    – Hearth
    Commented May 22 at 14:47

2 Answers 2

2
\$\begingroup\$

Chip resistors in offline power supplies is common -- do check that the datasheet provides nominal (operating) and surge/peak voltage ratings, however!

Then, use enough in series to meet rating even in case of a single device failure. That is, if any one resistor fails shorted or otherwise becomes bypassed, the voltage rating of the remaining string must be adequate. This is required for most safety standards e.g. IEC 60950-1.

Note that mains voltage is a type rating, where it's nominally e.g. "240VAC", but this implies a range of normal variation plus infrequent surges up to some amplitude, duration and (statistical) frequency (depending on CATegory). Use resistors rated for comparable statistics.

"Wirewound" here is a non-sequitur as no distinguishing or useful properties of a wirewound (high temperature; high peak power rating; perhaps relatively high inductance; etc.) are conferred by such arrangement. "High pulse" chip resistors are available, which may be useful for handling surge voltages, but you don't really gain much energy handling, a chip is a chip.

\$\endgroup\$
1
\$\begingroup\$

There are a few things you have to consider if you are trying to meet a safety standard. First, you already mentioned - don't exceed the max rating of any one resistor. Keep in mind, the voltage rating of the resistor is it's WORKING voltage. It will break down if subject to a hi-pot (which is required for compliance) at some voltage higher than its rating. The next thing is the creepage and clearance distances. In other words, the layout geometry matters. You need to consider the potentials at each junction and make sure there is enough spacing, both through air (clearance) and across surfaces (creepage). The last thing to consider is what the potential at those junctions look like under fail modes - what happens if the resistor fails shorted? How about open circuit? Do you have a cascade of component failures? Or does the design fail safe?

For a bench top experiment, your way is fine. But if I was doing this for real (aka commercial product) I would use parts rated for mains that will typically have the required markings (UL, CSA, CE ect. with HV rating). They are the cheapest components for a design. It's makes little sense to save a little cents on resistors that should be rated for current-sense.

\$\endgroup\$
1
  • \$\begingroup\$ Yes, I was looking at the WORKING Voltage, not the Overload Voltage. I understand the layout geometry matters, that is why I said I would place them all in a line to make sure the pads have proper distance between each other. \$\endgroup\$
    – Dinis
    Commented May 21 at 13:43

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.