|
The following series of posts to the BoatAnchors list consider the problem of protecting equipment from current and voltage transients that can occur when the power is first turned on. Unfortunately in the course of saving these and putting them together, the headers were lost, along with the names of some of the posters. Thus, I've generally omitted the identification of the poster. This also leaves me free to correct the spelling and delete the repetitive parts. But, if you recognize your stuff, and would like your name reattached, just ask! As always, these posts are offered without any guarantees!!
Post 1: Selecting an Inrush Current Limiter Some time ago I discovered a neat little device that solves the inrush problem and, as a side benefit, reduces high line voltage. All this for a bit over $2! The device is an Inrush Current Limiter made by Keystone Carbon Co. The beasties look like ceramic disk capacitors with a black vitreous coat. The limiter is a Positive Temperature Coefficient thermistor which is designed to handle current. When cold (room temp -25C) they exhibit some resistance. As current passes through them and they warm up, this resistance drops by a factor of about 100. The limiters are rated by current handling capability (1.1 to 16 Amps) and cold resistance (0.7 to 120 Ohms). Not all possible combinations of resistance and current are available but at last look there were about 20 different types. You use the limiter by installing it in series with the line cord (preferably the hot lead) input to your BA. This can be done in a fashion that is totally esthetically pleasing (read "out of sight") and completely reversible. IMPORTANT: Since the device is a resistor (and a HOT one at that) you must mount it away from heat sensitive components. I have mounted them under chassis without trouble but keep 'em away from just about everything. Don't attempt to heat sink it - that ruins the operation! Pick the right value by first measuring the steady state current of your BA. That is, after it is fully warmed up and all accessories are turned on. While you're at it , also read your line voltage. Pick a unit that has a MAX steady state current of 120 - 130% greater than the current you measured and has the HIGHEST no-load or cold resistance. Example: You measure 2.5 Amps (a moderately hungry BA!) and the line voltage is 123V. The KC008L is rated for 3.0 Amps with a cold resistance of 47 Ohms - a nice fit. Benefits: A BA drawing 2.5 amps probably has a transformer with a primary DC resistance of about 3 Ohms. Inrush, at the peak of the AC sine wave, could be as high as 40 Amps but probably not less than about 20 Amps. With the limiter installed, the inrush will not exceed about 2.6 Amps at 123 line Volts. After the limiter warms up it will have about a .49 Ohm resistance (actually a bit higher because we're not drawing the full 3 Amps.). This means that the line voltage across the transformer will be about 122 Volts (also a bit lower because of the higher resistance). This example came from real life and my actual results showed that the line voltage was reduced to 118 Volts (the BA was rated for 117) which means that the limiter was adding about 2 Ohms. Negatives: If your area suffers from brownouts, the limiter will exaggerate the effect. If voltage drops, current drops. The limiter will cool a bit, its resistance will rise, and the voltage your BA sees will drop more than the line voltage. This is a very minor problem for me but I feel bound to mention it. Post 2: Experience with a 51S-1 Receiver In the past few years, a new kind of thermistors has become available for limiting start-up surge currents in electronic instruments. They differ from conventional thermistors in having a negative temperature coefficient (resistance decreases with increasing temperature), and this property gives them a useful self-regulating characteristic. Placed in the ac line of an instrument, they initially have a high resistance, which limits the inrush current through the instrument. Upon application of power, the current through the thermistor causes self-heating, which lowers the device's resistance. At some point the resistance stabilizes to a value that depends on the equilibrium temperature of the device. The equilibrium temperature is determined by the steady-state current drain of the instrument and the ambient air temperature surrounding the thermistor. Current-inrush thermistors are inexpensive and provide an effective way to protect power supply components in vacuum tube receivers, particularly those that use solid-state rectifiers. Note that you should not use current-inrush thermistors to protect transmitters or amplifiers; they are only suitable for instruments that draw a relatively constant current from the line. (See later post) Here are the details for protecting a typical boatanchor receiver, in this case a Collins 51S-1. The steady-state current drain for my 51S-1 is about 0.8 Amps at 120VAC. To measure the inrush current, I temporarily removed the 1.5 ampere slow-blow fuse and jumpered a 1 ohm resistor across the fuse terminals. By measuring the voltage developed across the resistor with a scope, I determined the peak inrush current to be slightly more than 7 amperes! The equivalent load resistance presented by the 51S-1 at turn-on is thus (120 VAC/7 Amperes) = 17.1 ohms. As the filter capacitors charge and the tube filaments warm up, this load resistance increases to a steady-state value of (120 VAC/0.8 Amperes) = 150 ohms. A 7 ampere inrush current is very hard on the power switch, and isn't so great on the power transformer, rectifier diodes, and filter capacitors. The most suitable inrush thermistor I could find was Digikey (1-800-DIGIKEY) part number KC014L-ND, at a price of $2.13. This thermistor is specified at 50 ohms resistance at room temperature (54 ohms measured), and dropping to 0.89 ohms at 1.1A load. I measured the resistance at 1.1 ohms at the current drain of the 51S-1. To install the thermistor, I clipped the wire to the fuse socket of the 51S-1 and relocated it to an unused lug on a nearby turret. I then soldered one lead of the thermistor (which physically resembles a small disk capacitor) to the same lug and the other to the recently vacated lug on the fuse socket. I used a bit of Teflon tubing on the leads, and kept the leads long so I could suspend the thermistor in free space away from other components. The thermistor dissipates about a watt of heat and runs rather hot. After installing the thermistor, I replaced the fuse with a 1.5 Amp fast-blow type. I then remeasured the peak inrush current and found it now to be only about 1.8 Amperes, which is consistent with the theoretically expected value of 120 VAC/(54ohms+17ohms) =1.69Amperes. The peak inrush current is now only slightly greater than the steady-state current drain and should thus pose no problem for any of the power supply components. Note that this particular thermistor is appropriate for almost any boatanchor receiver that draws 75-150 Watts from the power line. Concern is often voiced about a related turn-on problem (actually, a turn-OFF problem), namely the inductive voltage spike caused by the power transformer inductance when the power is switched off. This spike is reputed to cause sparking and welding of contacts in hard-to-replace power switches, particularly in rigs like the KWM-2 and S-line. I checked on this problem with my 51S-1, but measuring the peak voltage developed across the power switch when the rig was shut off. (My Fluke 87 DMM has a peak-reading feature which can capture voltage transients as short as 1 msec.) To my surprise, I found that the inductive voltage kick was only about 5 volts higher than the line voltage, and was no cause for alarm. I had thought about using an MOV surge suppressor across the switch contacts, but decided it wasn't necessary. This is not to say, of course, that the problem isn't greater in some other rigs, but 51S-1 owners need not worry. Post 3: Inrush Protection for Transmitters Comment: >>Note that you should not use current-inrush thermistors to protect transmitters or amplifiers; they are only suitable for instruments that draw a relatively constant current. Response: Au contraire. Inrush current limiters work nicely in transmitters and transceivers and probably in amplifiers as well, although I've not tried that. The only stipulation is that the device must be selected to allow the maximum current needed by the transmitter. The resistance of the thermistor after the initial surge is very small, a fraction of an ohm, and less than the resistance provided by the typical AC mains. Consequently, its effect upon the load regulation of the transmitter is negligible. I used an inrush limiter a while back in an Eico 753/751 transceiver supply with excellent results. Prior to using the inrush current limiter, the power on surge produced an unnervingly loud KWUMMP! After installing the inrush current limiter, powering up the unit produced no audible effects at all. I don't happen to remember the voltage drop across the inrush limiter when just the receiver was operating, but I did measure it and found it be negligible; on the order of only a volt or two. And Someone Else Added: Of course, on larger transmitters one has to use thermistors on each element. Generally the current draw is too large to protect the entire transmitter. The filaments transformers, the plate transformer, low voltage transformers should all be individually "thermistorized". In mine, I find a volt or two drop at the thermistors is just what the doctor ordered as the line is slightly high. Post 4: Mounting Caution Don't solder them into your circuit unless you want trouble. They do get hot in operation and repeated heating and cooling of a solder joint will cause it to crystallize and eventually fail. This was a common failure in televisions with thermistors used in the degaussing circuits, and even with some of those cement-block power resistors on circuit boards. Put in a small screw terminal strip to mount the ICL. Crimp terminal lugs on the ICL and then attach it to the strip with the screws. In the long run, this will save lots of grief and it also makes installing and insulating the ICL a snap. Post 1: MOVs Turning off a rig can cause a big voltage spike across the transformer primary and the AC line. Usually it just burns out or welds your switch, as R-390A users often learn. A back issue of The Collins Journal suggested getting 240-volt MOVs and wiring them across your primaries to absorb the transient. Note that if the MOV fails (shorted) it will suck lots of current, but you have a fuse in the line, right? These will protect your switch, and apparently your transformers could use it too. I doubt the big toggle snappers in a Viking need it as much as the wimpy switches in an R-390A or KWM-2, but your transformers may last longer this way. And you'll get protection from nasty things that come in through your power line, and your gear won't put glitches back out there when you turn it off. Post 2: Selecting MOVs Query: >> There have been a number of posts touting the use of varistors to protect against voltage surges. Question is: How to decide what specs when buying these little doo-dahs? Answer: My background is in Mechanical Engineering, so take what I am about to say with a grain of salt. When I have picked MOVs (metal oxide varistors) in the past, say to protect stuff on the AC line against spikes, there are two things I have been concerned about. First is the clamping voltage. These little do-dads work by turning from a non-conductor to a conductor at the clamping voltage. The other rating is the amount of current they can handle. Usually this is broken into two numbers, a surge number with a time (like 7000 amps for a microsecond) and a steady state value if I remember correctly. So when I picked one to make into a AC surge suppressor, I picked a clamping voltage of about 150 volts with the highest current capacity I could afford. Post 1: Theory On Fri, 23 Aug 1996, Jan Skirrow, VE7DJX, asked me some excellent questions about thermistors, varistors, and such. I hope he does not mind me posting his questions or my reply to the group. Thermistors are not often seen in boatanchors (or in a lot of modern semiconductor stuff for that matter). I know a little about them because of their use in temperature measurement and instrumentation. >>First, I conclude that NTC thermistors would be placed in series with, for example, a transformer and would thus limit in-rush current because their resistance is inversely related to temperature, which would rapidly increase on start-up. Exactly. They are particularly beneficial with power supplies having capacitor input filters. Look at the special devices sold as Inrush Current Limiters, not conventional thermistors. Keystone is probably the most common NTC Inrush Limiter manufacturer. The typical resistance ratios of common NTC thermistors (for other than Inrush Current Limiting operations) is generally between 5 to 10 for 0 C to 50 C temperature changes. Plugging these numbers into the typical resistance relationship R = Ro * exp(B/T) R, Ro in ohms, T in Kelvin gives a Beta in the approximate range of 2800 to 4000. Using a value of 3400 as an average gives an Ro value of 0.0011 ohms (the resistance at absolute zero). So at 50 C, the resistance should be around 41 ohms (and at 0 C, the resistance is 284 ohms and the ratio is: {ta-da...} 6.9). In true Inrush Current Limiters, the Beta value is MUCH higher. If, for example, B is 10,000, the 0 to 50 C ratio is 290. This is such that a few ohms cold becomes very low resistance when hot. I don't really know what the Beta number is for these devices but I might be able to estimate it from the specs knowing the dissipation of the hot device and estimating some heat transfer conditions. It is not necessary to know it for picking an Inrush Current Limiter for your operation. In any event, a typical Inrush Current Limiter might have the following specifications (actually those for a Keystone CL-110): Resistance at 25 C: 10 ohms +/- 25% Maximum Steady State Current: 3.2 amps Approx. resistance @ maximum steady state current: 0.18 ohms >>I assume your reference to older metal oxide devices doesn't refer to metal oxide varistors - which seem to be a transient suppressor that functions by clamping the voltage across itself to some fixed level. NTC thermistors are generally made from oxides of manganese, nickel, cobalt, copper and iron. Metal oxide varistors for transient voltage suppression are generally variations on zinc oxides. Older thyristors were generally silicon carbide. It is interesting that while quite different in operation, the thermistors and varistors obey similar exponential relationships. The simple thermistor relationship is shown above. The current through a varistor follows a similar one: I = Io * exp(a*V) I, Io in amps, V in volts If you look at more exact relationships with both temperature and voltage dependency included, the equations start looking VERY much alike. Basically a varistor draws very little current at low voltages, but as the voltage increases, the current increases very rapidly. >>So these would be used by placing them across (for example) switch or relay contacts that switch an inductive load, and would prevent the voltage across the contacts from going too high due to transients, thus arcing and damaging the contacts. That is one use, although in snubbing an inductive load, the presence of a diode in a DC circuit or a varistor is an AC circuit will slow down the response of the relay. You really need something that will absorb the energy stored in the magnetic field. The more common use of a varistor is across the AC line as a transient suppressor. The voltage rating is chosen such that the device does not conduct much at normal voltages, but conducts heavily during a voltage transient. >>So, comprehensive protection for, say an R-390A, would be an NTC thermistor in series with the power transformer and a varistor that clamped at something over normal line voltage (perhaps 150v rms?) across the troublesome main power switch. Sort of! An Inrush Current Limiter in series with the transformer primary would reduce the current surge during turn-on. A varistor across the main power switch might help a LITTLE but what you really need here is a snubber network of a resistor in series with a small capacitor. Typical values might be 10 to 100 ohms in series with a 0.01 to 0.05 uF capacitor (rated at 1 KV minimum). A better approach would be to use a better switch! A 130 volt varistor, like a V130LA20, would be a good choice to add after the power filter network across the line. It would protect against line voltage transients. However, it won't protect the filter here. You should probably use a proper transient protected multiple outlet strip to power the radio anyway. The best ones will have 3 varistors inside. One from line to neutral, and one each from line and neutral to ground. Inrush Current Limiters and Transient Voltage Suppressors are quite inexpensive today. Small and unobtrusive, they can often be tucked inside your Boatanchor giving you some added protection. Post 2: Additional Comments >>I seem to remember horror stories about some so-called transient protected outlets that worked once, and then provided no protection as the varistors went south. All of my outlets are so protected, and I hope they all work! This is important too. In transient suppression, you want to have as much impedance between the source of the transient and the device you want to protect as you can get. Thus for best protection, a staged approach is a good one. At the service entrance to your house, you should have one of the lightning arrestor/transient suppressor blocks made for this purpose. These cost $15 to $30 at a commercial electrical supply house. The only problem is that with installation at the service entrance, you usually have to pull your power meter. Between the service entrance and the wall outlet, your house wiring provides some distributed capacitance and inductance. A 3-MOV protector at the outlet is a good idea here. Checking them is a problem as there is no simple way to do this. If your circuit breaker or fuse blows upstream of the protector for no apparent reason or during a thunderstorm, you can probably assume the protector "went south" and needs to be replaced. Finally at your equipment, its line cord and RFI filters provide even more impedance. A transient protection MOV inside the rig provides the final stage of protection. It can be smaller in its ratings since the earlier protectors should have already taken most of the energy away from the transient. Nothing protects against a direct-hit of lightning though. But I would still rather have a few MOVs explode, and maybe a line-filter or two, than the entire rig to replace!
May 13, 2003
|
