[Amps] Resistor voltage ratings

[Richard]

>Can any one tell me what is the origin of voltage ratings on ordinary 
>resistors (i.e. film and composition types.) In every case where I had to 
>design a high voltage divider, the dominant factor was always the power 
>rating of the resistors. This is intuitively a limiting factor, since 
>thermal dissipation requires a certain amount of surface area and/or volume 
to 
>conduct and convect away the heat.
>
>But I don't really see how any serious voltage stress is present within a 
>conductive medium, whose conductivity has been specially selected to produce 
>as much loss as possible within a small but finite length.
>
[]  The problem is that the conductive medium begins to arc internally.  
This can be seen by observing the current-versus-voltage relationship.  
For example, a 10M, 2W carbon-comp resistor dissipates 0.1W at 1000V, so 
all should be well.  The current at 1000V should be 100uA, however, the 
current measures substantially higher because the carbon is arcing 
internally at 1000v.  The rating of a 2w carbon comp resistor is 500v, so 
a 10M, "2W" resistor is really and truly  a 250mW-max resistor. 

>Some high-voltage resistors are simply long film resistors wound into a 
>spiral on a cylindrical substrate. For this topology, the inter-turn spacing 
>would become a factor to consider. However, I am not inquiring about 
>resistors of special HV design.
>
>For instance, when you look at a 1W ceramic body DC resistor, the end caps 
>alone look like they could easily withstand many kilovolts. So, I am really 
>baffled as to the origin of this 'voltage rating'. Manufacturers have some 
>interesting and unusual ways of establishing specs on their products, and I 
>have never personally had a need to be involved in this one. But listening 
>to 
>the comments here has piqued my interest to the extent that I would like to 
>know if there is any factual basis for it.
>
>By the way, you can fabricate very nice HV meter multipliers using 1206 or 
>2012 SMT resitors. I have made up little "sticks" of them that end up 
>looking 
>like ordinary fuses. I keep the voltage across each resistor down to  2 or 3 
>hundred volts. I think they would handle 500 easily.
>
>Eric von Valtier K8LV

-  R. L. Measures, a.k.a. Rich..., 805.386.3734,AG6K, 
www.vcnet.com/measures.  
end