Archive of the former Yahoo!Groups mailing list: MOTM
Subject: Overvoltage Protection
From: mbedtom@...
Date: 2002-04-26
Shemp posed a couple of questions to which I will attempt a response:
Power line filters... Power line filters are usually an integral part of the power entry module on a power supply. They serve two main functions: 1) Power line filters help to oppose the entry of radio frequency energy that might have some negative impact on the instrument operation. 2) Power line filters help to contain any generated radio frequency energy that might "leak" out and interfere with the operation of other instruments in close proximity.
Modular analog synthesizers don't generate a whole lot of radio frequency energy. But a synth is composed of many circuits connected in series that amplify signals. To achieve good fidelity, many of these circuits are quite happy to amplify radio frequencies. If unwanted radio signals enter through the power cord (because it acts like an antenna), those signals can be coupled into the various amplifiers used in a synth. And anytime you have a junction of dissimilar metals, you have formed a crude diode detector. A diode can demodulate AM radio signals, leaving the raw audio to blend into your synth audio. That isn't a good thing! I can look out my den window and see the antenna of a 5 kilowatt AM transmitter. My telephone wiring comprises a wonderful antenna for receiving this radio station. Sometimes this unwanted audio is so loud it drowns out the caller's audio. Have you guessed yet that there is a bad-ass filter on my synth's power supply? You might not live in as hostile an area as I do, but if you gig with your gear, can you guarantee that you'll never be near a transmitter that can cause interference? A filter can be helpful.
Overvoltage protection... This too, is a required item in my opinion. The basic operation of an overvoltage protection circuit is pretty straightforward: A small circuit sits between the +15 and -15 output rails of the regulated supply. The overvoltage protection circuit contains a voltage sensor and a high-current "electric switch" (an SCR, actually). When the applied voltage rises above a preset level, the "electric switch" is closed and shorts out the power supply output pins... on purpose! The rationale is this: if the voltage rises beyond the preset threshold, something has gone disastrously wrong in the power supply itself. This extra voltage could easily damage the circuits connected to it. So it shorts out the power supply which will cause a heavy current to flow in both the primary and secondary side. In a moment or two, the fuse on the primary of the power supply will blow as a consequence of this extra current draw. When the fuse pops, the supply goes dead and the possibility of damage to other circuits is minimized. (You don't use a larger fuse than you're supposed to... right!?)
The reason a synth power supply can produce a voltage dangerous enough to damage circuits, is related to how the power supply regulates voltage in the first place. Most power supplies are rated to provide the specified output voltage with as little as 105 volts AC input (in the United States). In order to regulate the output to +/- 15V, the input voltage to the regulator circuit must be higher than the desired output by about 2 volts. When the input AC mains voltage is 105 volts, the input to the regulator circuit (on either polarity) will be at least 17 volts (probably more). But these supplies are also required to deliver the same regulated voltage at 130 volts on the AC mains. That means the input voltage to the regulator circuit will rise to over 21 volts (again, probably more). The main power handling component of the regulator is a pass-transistor. If this pass transistor were to develop a collector-to-emitter short, the full unregulated voltage will be sent to the output. Therein lies the problem. A single component failure (the pass-transistor) could cause your expensive synth modules to be damaged. (Or they can be stressed to the point that their expected life is much shorter than if they weren't stressed.) It's one thing to zap a 60 cent opamp, but quite another to blowout a CEM chip like that used on the MOTM-110.
Did that answer your question or did I just make your eyes glaze over? (I do that quite often it seems!)
Cheers!
Tom Farrand