3. Eight Decade Resistance Box (2012)
I had long contemplated buying a decade resistor box for my projects but I didn't want a plastic Chinese one that would probably blow in no time, and even on eBay used commercial ones were not cheap, probably because they were higher precision than I needed, but with lower power handling as a result.
When I first looked for low cost high voltage power supplies for the LHC project, I found electrophoresis PSUs were usually priced lower than conventional bench PSUs, and I bought a couple.
Electrophoresis PSUs are designed to separate chemical strands in liquid gels containing DNA, see [Adventures: Montenegro Scientific Expedition, Day 8]. A lot of these PSUs, and certainly the Bio-Rad and Life Systems ones I bought, have safety circuitry built in that prevents the PSU from turning on if it cannot detect the resistance of a gel between its output terminals. I was unaware of this before I bought them (another annoying feature is they also have electronic dials that don't remember their settings once power is lost).
Lacking a schematic for the PSUs and observing their complexity, I abandoned the idea of modifying them to disable the interlock, and decided instead to opt for the brute force of a power resistor across the terminals to mimic the gel resistance, although without a specification this is somewhat hit and miss.
I had few high voltage power resistors in stock and they have never been particularly cheap. This renewed my interest in a decade resistance box because by stringing lots of resistors in series the overall voltage rating increases by the number added, but now I was looking for something with a little more power handling than a precision box.
There are dedicated high power decade boxes, typically the Clarostat 240 2% series specified 'up to' 225W (it appears they utilise a resistor network to reduce the number of resistors), see datasheet [I45] below. They are specified for 1kVdc or 600Vac max. Winslow (Tele-Tronics) is a cheaper competitor with identical looking boxes (e.g. Model 336) but both are bulky and heavy and although useful, absolute overkill for my needs.
Unable to find something both affordable and practical, I decided to build my own resistance box with enough features to meet my current and future needs although the initial use was to determine the resistance needed to get the electrophoresis supplies to work, albeit at low power. I based its design on this task as it is more demanding than most other uses I'd have for it, and decided 2W 1% resistors were an affordable compromise; assuming a resistance up to 1MΩ and applying [Q1] Ohm's Law P=V²/R:
2W allows 1MΩ up to 1400V (1.4kV²/2W = 980kΩ), and down to 125kΩ on 500V (500V²/2W = 125kΩ), the higher voltages of course dependent upon me building up the value across several switches, although of course the voltage is not infinite; I would not want to exceed 2kV [E36] Later note: I should really determine this using my Clare HiPOT [I20].
I could then scale the results and buy the correct high power resistors for the PSUs.
Resistance boxes are typically 6 decades but I made mine 8 decades spanning 0.1Ω to 1MΩ per position with an extra '10' position on each dial, providing 0 to 11,111,111.1Ω. Each is a 1% 350V 2W resistor. I used a shielded metal box, useful for analogue circuitry and phono, 4mm banana and 4mm turret connectors for both terminals, and a separate terminal for the 0.1Ω range.
The total cost was around £60 and it took a couple of weeks' evenings after work to build it:
