shep@avsdS.UUCP (10/26/83)
While "hot-roding" some thirty phase linear 700's years ago, I too had to confront the problem of surge suppression. The modified amplifier drew less than an amp quiescently and up to 15 amps full tilt off of the bipolar 100 volt rails. The power supply was simply a massive transformer feeding an fwb rectifier feeding equally massive filter caps. The original phase linear power transformer was left alone; but a new bridge and caps were fitted. As I recall, about half a farad was used on each rail (~50 joules/rail!). The in-rush current, as you might expect, was very large and a "relay charge `em up" scheme was unacceptable, mostly because of chassis space limitations. Enter NTC (negative temperature coefficient) resistors, little widgets whose resistance decreases as the temperature increases. Some firm (the name escapes me) offered them to us in a variety of flavors. A network was built that would handle the power required, and it worked fine. When the amplifier was off, the devices were cool and at at their greatest resistance. At turn on, the caps looked like a short, the resistors got warm, the caps charged. Once charged, the bias on the many transistors kept the quiescent draw on the rails high, and thus the ntc resistors warm and at a low resistance. The low, yet finite, resistance was never a problem since the caps were so large. Shep Siegel, Ampex Corp. s:shep