Resistors - not quite ten a penny!

[Terry]

The subject of the voltage rating of resistors seems to confound many people including some who have posted here, based on the contents of their posts ...

Mark is the only one who has addressed this topic objectively ...

When this subject was discussed on another forum a couple of years or so ago on another forum, I created a chart in Excel to show the maximum voltage across resistors of various values which caused them to meet their maximum power rating - the answers might surprise you!

The values are down the left hand side and the power ratings are along the top. The remaining values are the maximum voltage permissible for each value within its maximum power rating.

   

The original table was for the full E12 series but has been edited to show only E6 values in order to keep the length of the table manageably short.

Why I didn't calculate the figures for the lowest values of 1W and 2W resistors when I produced the table three years ago, I have no idea!

The values, for anybody who is interested will, of course, be double the figures in the ¼ & ½W columns respectively.

[Nick]

What formulae was used to calculate this table? A back-calculation from the maximum power? i.e sqrt(rated power * Rvalue)

The maximum continuous working voltage for a resistor is the minimum of that quoted by the manufacturer for the class of resistor (Limiting Element Voltage or LEV) and the back calculation from maximum power above.

I'm not sure how you can have a table of maximum continuous working voltage that holds for all resistors of all types from any manufacturer when the LEV is independent any calculation - you've just used typical LEVs for a given power rating but the actual LEV is that stated by the manufacturer for that series of resistors and may or may not be the "common" value... e.g. tje Welwyn RC55 250mW series have a LEV of 350V, not 250V and when SMDs are considered, all bets are off...

[pwdrive]

http://www.vishay.com/docs/49873/49873_sg2113.pdf

Lawrence.

[Yorkie]

This has turned out to be quite an interesting and instructive thread - well to me, anyway!

Almost everything I build in terms of electronic projects rather than radio restoration are simple projects (small things amuse small minds!), which aren't safety critical and use low voltages. However, occasionally I have built projects which use high voltages and in an earlier post I was trying to say (not very well perhaps), that just because a resistor is connected to a high voltage point - say 230V - doesn't mean that the PD across it will be 230V or anything like that, but as Terry points out in his table,  it isn't simply a case of 'oh well, these are rated at 250V so they'll be fine in all circumstances'.

Just to give a little example, some years ago I built a couple of nixie clocks from a design on the excellent 'Mike's Electric Stuff' website. The design goes back to 1999, and was initially for use in America, with a lower mains voltage, but the circuit and PCB allows for use on UK mains. Nick is familiar with the design, and expressed some disquiet at it being run straight from the mains with no isolating transformer, and Mike does show two back-to-back transformers to achieve that objective.  From a safety perspective, the two clocks I made (one for me, one for my son) are in non-metallic enclosures have no external metalwork and the time is set externally using a small button magnet to operate two reed relays internally

Setting those considerations aside, and concentrating on the voltage rating of the 47k resistors used to drop the voltage for each cathode on each tube, they have to cope not with 230V, but with the voltage dropped across them and the current drawn. (Obviously). The strike voltage of the cathodes is 170V, so the  voltage dropped across each resistor is 230 - 170 =  60V, so the current drawn is  60/47K = 0.00128A.  Hence, 60V x 0.00128A = 0.077W.  I appreciate of course that there will be lots of situations where a higher voltage drop will be called for, at a higher current rating,  but in the case I've mentioned,  even 1/8 Watt resistors would cope.

Not to take the thread even further off topic, (not that it matters here in happy valley!), here's the nixie clock design:

http://www.electricstuff.co.uk/nixclock.html

Thanks for the erudite input to the thread, and the table from Terry and  the Vishay  link by Lawrence.

[Terry]

I'm not sure how you can have a table of maximum continuous working voltage that holds for all resistors of all types from any manufacturer when the LEV is independent any calculation - you've just used typical LEVs for a given power rating but the actual LEV is that stated by the manufacturer for that series of resistors and may or may not be the "common"  value... e.g. tje Welwyn RC55 250mW series have a LEV of 350V, not 250V and when SMDs are considered, all bets are off...

I doubt you'd be buying Welwyn RC55 resistors at ten for a penny, Nick! That is, after all, what we are discussing here.

My table was - and still is - intended as a generic guide, the maximum voltages being taken from a manufacturer's data sheet for typical run of the mill resistors.

It's a damn site more intuitive than putting 450V across a resistor to see what happens!

I would expect anybody with special requirements to research manufacturers' data sheets for the correct component.

The formula I used, as you surmised, is V = √PR where P is in watts and R is in ohms.

[ppppenguin]

The table also illustrates that for higher value resistors you reach the voltage limit before the power disspation limit. Exactly where that happens depends on the voltage and power rating of the resistor. Given by inverting Terry's formula, R = (V^2)/P

Terry's formula above is a touch ambiguous since it's not clear that the square root applies to the product of power and resistance rather than power alone. I appreciate that it's not possible in ordinary characters to extend the top bar of the root symbol to make this clear. V = √(PR) might be better.

[Terry]

Thank you Jeffrey, I must admit that I did consider using brackets but thought it clear enough as it was.

At least I managed to find the square root symbol!

[ppppenguin]

Being too lazy to seek out these odd symbols I usually use the sort of notation usually used in computer languages. Hence I would write V = SQRT(PR) or even V = SQRT(P*R). Not elegant but effective. Likewise I rarely bother with superscripts hence X^2 for X squared and 4.5E-2 for 4.5 times 10 to the power of -2.

I could even use X^0.5 for square root but that's probably going a bit too far for a non-mathematical forum.

[Joe]

You're not kidding.

— Joe