For you maybe it's a hassle, but for those of us in North America/Europe who don't mind through-hole parts, I recommend Vishay HVR25 (1/4-W) or HVR37 (1/2-W) resistors- easily available from Digikey and rated to take decent mains spikes.
If you must use a string of regular resistors- 1% metal film types or (better but less accurate) Metal Oxide Film power types are going to be less prone to setting off impressive fireworks compared to anything made with carbon (film or composition) as well as more accurate.
In my younger days I have done quite a few "experiments" that some could characterize as application for the Darwin Award. Anyway I would never let anyone else (including Internet) to use my experiments.
For a one off device, is ten resistors series a problem ?
For one off products, just put ten in series, since cost is no issue.
Just to clarify, carbon (graphite) is a semimetal; it has negative tempco of resistance at low (cryogenic) temperatures, positive tempco at high temperature. It's good for resistors because room temperature is near the stationary point (low tempco).
Normal carbon composition (solid, not film) resistors have voltage ratings of 350V for 1/2W and 500V for 1W body sizes. This is an rms value. Pulsed instantaneous voltage ratings are double this.
Solid carbon and even solid film parts have a surge capability that exceeds that of spiral-cut film constructions.
Normally you'd derate any resistor to 70% voltage rating for 50% stress levels, so you're looking at two parts in series, regardless.
There are still one or two mfrs claiming to supply these parts, and you can still order them through DigiKey and other reputable distributors but you'd be better off to find something else, unless you actually need to use them, or it's a restoration/repair.
The only way to know if you can get away with it is to try it. It doesnt se eem to be something people here have tried. I'd expect real life voltage ca pability to degrade badly if condensation occurs. You can always include 1.
25" fine copper wire links as backup fuses, strung across air so they dont set fire to anything.
5mm creepage is plenty. Phenolic FR2 should be fine, FR4 just gives better strength & track adhesion, and makes PTH possible. Most 1 & 2 layer things are fine on FR2. I dont know how RH affects FR2 V rating though.
I put 400v or 500 on a .3w carbon film once. The result was instant carbon arc lighting. I think the idea was to try using it as a fuse, it wasn't eff ective, arced far too long before going oc.
Why do you need to push it with a single R? 0.25w Rs are surely very cheap there.
A 70% voltage derating in a resistor produces a 50% power derating.
Who is folk, and where is this recommendation made?
Optimum part choices may depend on application, availability, volume of intended market, location of fab, material and tooling cost targets and possibly a host of other factors.
In this specific case it's a recommendation for use in a single test instrument or probe, with some pre-owned parts being available. The OP questions their suitability for the basic function.
Although I don't want to damp anyones spirit of adventure, my own habit is usually to overbuild, where feasible, anything constructed by hand on my own bench, even if it is intended to be used only once.
Determining suitability of a part for applications outside of it's original intended role requires both common sense and solid experimental technique. There are good reasons why fuses rated at varying interrupt capacities and voltages of AC or DC are identified as being constructed and employed safely.
If you want a 300mW axial film resistor to serve as a fuse, you'd be restricted (in my working experience) to resistance values below 18R, as even 'carbon film' series parts need to revert to a nickel metal film in order to cover this lower impedance range. Specific sources of supply must be tested and vetted for this specific feature.
These will fuse fairly predictably between 80V and 240VAC, if suitably contained to prevent explosive ejection of their component parts. This results in a single-fault abnormal behavior that is considered to be acceptible by most NRTLs. They are not and must not be refered to as fuses in your documentation - and they may not serve as such, where rated fusing is called for.
Fusible or flame-proof resistors are also manufactured over a wide range of sizes and ratings, requiring only that you source them and check their suitability in your application.
Maker's max voltage rating is independent of the power dissipation rating.
The dissipation limit for a 1Mohm, 0.5 watt resistor resistor implies a max continuous voltage of 707V while the makers maximum is far less at 350V or 250V.
For low value resistors, the rating sets the max pulse voltage allowed.
I was explaining reasoning behind the relationship between these two common deratings. In any application, or stress table, there is likely to only be one that dominates - a 50% voltage derating would be unnecessarily punitive, in an application where power stress is not a factor. Not the only consideration, by any means, but a start.
Guidelines of derating recommendations for reliability are available, though possibly dated, in TM 5-698-3
seeem to be something people here have tried. I'd expect real life voltage capability to degrade badly if condensation occurs. You can always include 1.25" fine copper wire links as backup fuses, strung across air so they do nt set fire to anything.
er strength & track adhesion, and makes PTH possible. Most 1 & 2 layer thin gs are fine on FR2. I dont know how RH affects FR2 V rating though.
on arc lighting. I think the idea was to try using it as a fuse, it wasn't effective, arced far too long before going oc.
ap there.
The OP is in the 3rd world. Overengineering there is not normally a good id ea
NRTLs are a very long way from normal 3rd world engineering practice. A lit tle time in such places can sure bring some perspective.
Those parts are inexpensive, e.g., 12 to 15 cents for the HVR25, qty 100, and are pretty well stocked. They have a DC voltage rating of 1.6 or 3.5 kV. What's not to like?
It's not the price. Mail order of electronic parts never took off here in India and, with the global change in the retail market, has not improved in the internet age. Besides, I don't just live in India, I live in the most isolated state in the least developed region of the country.
On the other hand, brick-and-mortar shops still thrive in the larger cities. In the nation's capital Delhi, for example, literally hundreds of parts shops are crammed into an area roughly 250x100m. When I visited the place last year, one significant change I noticed from earlier times was that most shops had set minimum quantities for each item. I found a few still willing to sell in small quantities but those don't ship out in any quantity.
Physical electronics parts stores have disappeared from US cities. There used to be zillions of radio parts stores and surplus stores. No more; everything is shipped. I can order something this afternoon from Digikey or Amazon and, for a little money, have it tomorrow morning. That's good, because parking to shop is impossible.
Silicon Valley still has a few surplus stores, but the cost of real estate has killed most of them, and they were getting full of computer junk anyhow. ebay + UPS has killed surplus stores and electronic flea markets.
Sounds like India really needs an equivalent of UPS.
--
John Larkin Highland Technology, Inc
picosecond timing laser drivers and controllers
jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.