Contents:
WARNING: In general, I DO NOT recommend making any sorts of measurements on the high voltage components of a live microwave oven. I only include this section for those who really want to know the details. You may be temped to break out your Radio Shack DMM and start poking away inside a live microwave oven. DON'T! This isn't like a CD player! Most of the time, no measurements of any kind on the oven while it is operating will be needed to identify and correct the problem. However, where this is not the case, here are some guidelines to a long life: WARNING: ALWAYS pull the plug and discharge the HV capacitor BEFORE doing anything inside! Never be tempted to make any changes of any kind while the oven is on - not even if your meter is being consumed by 5 foot flames! First, pull the plug and discharge the HV capacitor! * High voltage - DON'T even think about this unless you have a proper high voltage probe or meter, or a proper microwave oven tester - AND KNOW HOW TO USE IT SAFELY. Even professionals have been killed performing measurements of this type using proper equipment! Luckily, current measurements can provide enough information to help make a diagnosis. WARNING: The high voltage components inside a microwave oven are at a NEGATIVE potential with respect to the chassis. DO NOT be tempted to interchange the probe and ground wire if you are using a high voltage probe on a meter with a POSITIVE input (e.g., for testing CRT HV) and no polarity switch! The ground cable doesn't have anywhere near the required insulation. Get the proper equipment! * Magnetron current - Place a 10 ohm 10 watt resistor in series with the HV diode cathode and ground. Measure the voltage drop across this resistor. Sensitivity will be 10 V/A. Normal anode current is around 300 to 400 mA for a typical oven. This will be -3 to -4 VDC across the 10 ohm resistor with respect to chassis ground. SET EVERYTHING UP AND THEN STAND BACK and don't forget to DISCHARGE the HV capacitor after making the measurement: - If it is around this range, the magnetron is probably fine. - If it is very low or 0, magnetron is bad or HV is not working. Note that a shorted as well as open magnetron also results in no current. If the magnetron is shorted, it bypasses all current to ground. If the magnetron is open, the HV capacitor charges up and then there is no more current through the HV diode (but there will be an initial transient). - If it is much too high (whether fuse blows or not), capacitor is shorted.
WARNING: First, with power disconnected, discharge the high voltage capacitor. See the section "Safe discharging of the high voltage capacitor". Assuming the oven passes the above test for interlocks and door alignment, the triac (if used) may be defective. There could also be a wire shorting to the chassis. However, the most likely problems are in the microwave generator. An ohmmeter can be safely used to quickly determine if the capacitor, HV diode, or magnetron are a dead short (as well as for an open magnetron filament). Use an ohmmeter to test the diode and capacitor. While connected in circuit, the resistance in at least one direction should be several M ohms. (Try it in both directions, use the higher reading). Test the magnetron from the filament to chassis - it should be high in at least one direction. Test the filament for continuity - the resistance of a good filament is close to 0 (less than 1 ohm). Where the capacitor and diode are combined into one unit, it should be possible to test each component individually and replace only the one that is found to be defective if the entire assembly is excessively expensive or not available. These may be considered to fail/no conclusion tests - they can definitively identify parts that are bad but will not guarantee that they are good. Parts may test ok with no voltage applied but then fail once operated in-circuit. Connections may open up when they heat up. The magnetron may short out when full voltage is applied. Don't overlook the wiring as no heat or erratic operation can result from simple bad connections! An alternative way of determining if the problem is in the control circuits (triac, relay, wiring) or microwave generator (HV transformer, HV capacitor, HV diode, magnetron, wiring, etc.) is to connect the HV transformer primary directly to a line cord and plug. Tape the removed wire lugs to prevent shorts. Plug the transformer cord into a switched outlet strip which includes a fuse or circuit breaker. Put a cup of water into the oven cavity to act as a load. * Power the oven via its line cord. Initiate a cook cycle. It should go through the normal cycle (of course no heat) without blowing the fuse or any unusual sounds. If there is a problem in this case, something in the controller or its wiring is shorted. * Now, initiate a 1 minute cook cycle on HIGH and with the oven running, switch on the HV transformer. - If the transformer or other HV components are faulty, the outlet strip fuse will blow or circuit breaker will trip. Or, if a lamp is plugged into the outlet strip at the same time, it will likely dim significantly due to the heavy load before the fuse or breaker cuts out. - If the problem is with the triac or its drive, the oven will now heat normally. When the cook cycle is near its end, switch off the outlet strip. Check the water's temperature. More complete information on testing and replacing the individual components is provided in the next few sections.
WARNING: First, with power disconnected, discharge the high voltage capacitor. See the section "Safe discharging of the high voltage capacitor". The HV diode can fail shorted (most likely) or open. It is not likely for there to be anything in between as so much heat would result that the diode would not remain that way for long. * A shorted HV diode will likely result in a loud hum from the HV transformer when a cook cycle is initiated. The main fuse will probably not blow. * An open HV diode will result in AC instead of DC across the magnetron with a peak negative value (the only one that matters) about 1/2 of what it should be. The result will likely be little or no detectable heat but no other symptoms. The resistance measured across the leads of the HV diode should be greater than 10 M ohm in at least one direction when disconnected from the circuit. However, the HV diode is composed of multiple silicon diodes in series to get the voltage rating. Its forward voltage drop will therefore be too great (6 V or more) for a DMM to produce a definitive answer as to whether it actually works as a rectifier. The HV diode can be tested with a DC power supply (even a wall adapter of at least 12 or 15 V output), series resistor (to limit current), and your multimeter. This will determine proper behavior, at least at low voltages. The following is the schematic of a simple HV diode tester: 240 ohms, 1 W + o-----------/\/\---------+------------o + | __|__ HV Good: 6 to 10 V 15 VDC _\_/_ diode Shorted: 0 to 2 V | Open or reversed: 15 V | - o------------------------+------------o - The voltage drop in the forward direction should be at least 6 V with a few mA of current but may be somewhat higher (8 V or more) with a few hundred mA. If your DMM or VOM has a resistance scale operated off a battery of at least 6 V, you may get a reading in one direction (but only one) without the need for an external power supply. Or, assume for now that the diode is good if it is not shorted - which is likely.
WARNING: First, with power disconnected, discharge the high voltage capacitor. See the section "Safe discharging of the high voltage capacitor". Most HV diodes have press fit (Fast-On) or ring lugs so replacement is very straightforward. Discharge the high voltage capacitor. Make sure you get the polarity correct if your replacement can be installed either way. Putting the diode in backwards will result in positive instead of negative high voltage and, needless to say, no heat, but no other symptoms either. Note: the lugs on your new HV diode may just be crimped onto the wire leads and not welded or soldered. If this is the case, take care not to stress them excessively which might result in bad connections now or in the future. It may be a good idea to solder the lugs to the wires as well (though this may be overkill). Where the diode is part of the capacitor assembly, it should be possible to just replace the diode leaving the old one unconnected (at one end). This will probably be much much cheaper than replacing the entire assembly.
Most replacement microwave oven diodes are rated 12 to 15 KV PRV at .5 A. A PRV of around 8 KV is actually required even for a small oven. Here is why: Until the magnetron heats up and starts conducting in its forward direction, what you have is a half wave rectifier/filter formed by the HV transformer secondary, the HV diode, and the HV capacitor. The reverse voltage across the HV diode will be equal to: 2 * 1.414 * (VRMS of the HV transformer). This can easily be 6 or 7 KV or more! Once the magnetron start conducting, the reverse voltage goes down somewhat. HV diodes rated at .5 A are adequate for most domestic microwave ovens. For example, the largest of these will have a nameplate rating of around 1,800 W power line input and a HV transformer secondary of 2,500 VAC. While there are some losses in the HV transformer, and some power is used by the magnetron filament, controller, motors, and light, this still leaves, perhaps, 1,600 W into the HV generator. However, due to the design of the half wave doubler circuit, not all the power flows through the HV diode (as would be the case with a regular power supply. Thus, even though calculations using Ohms law (I = P/V = 1,600/2,500 or .64 A) would suggest that .5 A is not enough, closer to 1/2 of the total current actually flows through the HV diode. To be doubly sure that your new HV diode is happy, run the oven on full power (high) for 10 minutes with two quarts of water as a load (or a roast). Unplug the oven (while your spouse prepares the veggies), quickly DISCHARGE THE HV CAPACITOR, and then check the HV diode for overheating. It might be warm but should not be too hot to touch. Unless you have the largest oven on earth, this test is probably not needed.
WARNING: First, with power disconnected, discharge the high voltage capacitor. See the section "Safe discharging of the high voltage capacitor". * A shorted HV capacitor will blow the fuse instantly. * An open HV capacitor will result in no heat but no other symptoms. (The following assumes no internal rectifier or other circuitry except of a bleeder resistor. Adjust procedures accordingly if your oven is different.) The resistance measured across the terminals of the high voltage capacitor should be very high - several M ohms for bleeder resistor. If it is less than 1 M ohms, the capacitor is definitely shorted. Yes, if you measure 0.00 ohms across the terminals (and they are not bussed together on the case), then the capacitor is positively, without a shadow of a doubt, bad! A high resistance does not prove that the capacitor is actually functional, just not shorted with no voltage across it. If you have a capacitance meter, check it for proper value (should be printed on the case). Even this does not prove that it will not short when full voltage is applied. Substitution is the only sure test beyond this.
WARNING: First, with power disconnected, discharge the high voltage capacitor. See the section "Safe discharging of the high voltage capacitor". Make a diagram of the precise wiring as multiple connections are often made to the capacitor terminals. The capacitor is usually mounted with a clamp which is easily loosened. Sometimes, the capacitor is jammed into a location that requires moving some other components to extract it. Replace in reverse order. Tighten the clamp securely but not so much as to distort the case. Where the capacitor assembly also includes the HV diode, it is possible to just replace the capacitor if space permits leaving the old one unconnected (at one end). However, the cost of a generic replacement diode is small (around $3) so replacing both at the same time is usually best. However, you don't need to use the exact combined part - which may be very expensive or difficult to obtain. Just make sure the ratings of the capacitor and diode are correct (use a generic replacement microwave oven HV diode and a microwave HV capacitor with a uF rating within 10% or so of the old one and at least equal working voltage).
WARNING: First, with power disconnected, discharge the high voltage capacitor. See the section "Safe discharging of the high voltage capacitor". * A magnetron with an open filament will result in no heat but no other symptoms. The bad connection may be internal (in which case the magnetron will need to be replaced) or external at the filament terminals (which may be repairable). * A magnetron with with a short between the filament/cathode and anode will likely result in a loud hum from the HV transformer and/or magnetron when the cook cycle is initiated but the main fuse will probably not blow. * A magnetron with other faults may result in a variety of symptoms including erratic or low output power or intermittent operation. See the section: "Comprehensive list of magnetron failure modes". There is no totally definitive way to determine if a magnetron is good without actually powering it under operating conditions but the following tests will catch most problems: * Magnetron filament. The resistance should be infinite from the filament connections to the case and a fraction of an ohm between the filament terminals with the wiring disconnected from the magnetron. While measuring resistance from filament chassis, gently tap the magnetron to determine if there is an intermittent short. However, such problems may only show up once the filament heats up and parts expand. It may be possible to determine if the magnetron filament is actually working by connecting just the filament connections to a low voltage high current supply on a Variac (e.g., a microwave oven transformer but just the filament connections). The ceramic insulators are translucent and should show a glow with a working filament. The one at the antenna may be visible if the magnetron is removed from the oven or with a dental mirror looking into the waveguide. WARNING: Make sure you ONLY have the filament connected! * Evidence of arcing (visible blackening around ventilation holes in base or burnt odor) usually indicates a bad magnetron. * Melting or other damage to the antenna cover ('bull-nose' or 'bullet') may be the result of arcing due to problems in the oven cavity or waveguide (perhaps operating with nothing in the oven) or a defective magnetron. (This part is only visible with the magnetron removed from the oven). If a problem elsewhere has been corrected, the damaged antenna cover can be pulled off and replaced from a magnetron that died of other causes - try your local appliance repair shop. (The shape doesn't matter as long as it fits tightly - there are several diameters, however.) Your magnetron may still be good. Most common magnetron failure modes: * Filament could be shorted to case - check with ohmmeter. Anything less than infinity means the tube is bad though it could be charring due to arcing outside the vacuum in the box with the filament connections. Tap the tube while measuring to check for intermittents. * Filament could be shorted to itself - tough to test since it is such a low resistance to start. Compare with good magnetron. (Yeh, right. If you had one, this wouldn't be an issue!) Tap the tube while measuring to check for intermittents. This fault isn't really likely. * Filament could be open - check with ohmmeter. Tap the tube while measuring to check for intermittents. However, loose filament connectors (Fast-Ons) are more likely than a broken filament. Therefore, check directly at the magnetron terminals with both lugs pulled off. * Magnetron could be gassy (or up to air) and arcover internally once power is applied. The filament could expand, shift position, and short once heated. There is no easy way to test for these possibilities other than substituting a known good magnetron. * Internal or external arcing resulting in physical damage. External arcing could be at the antenna or inside the filament box. Internal arcing will not leave any visible evidence but the damage will result in the magnetron failing totally or running with reduced output. * Overheating might result from a broken or cracked magnet (reduced magentic field) or other internal problems. While there may be some output power, the thermal protector will shut down the oven prematurely.
(Portions from: John Gallawa (mtek@pen.net)). Here is a list of typical magnetron failure modes. The percentage of each type of failure varies. Currently, internal shorts and loose filament connectors are probably at the top of the list. An internal plate-cathode short may only manifest itself under the stress of high voltage during operation. 1. Shorts. (a) Internal plate-cathode/filament short or (b) Internal arcing. Symptoms: No heat, loud hum when entering cook cycle, possible blown HV fuse (but will not likely blow the main fuse). In ovens equipped with fuses that monitor the high voltage system, such as some commercial Sharp models and most commercial and domestic Amana models, the high voltage fuse would probably blow. But, rarely will a shorted magnetron cause the main line fuse to blow. (I suppose the transformer absorbs most of the current surge.) In fact, with reference to the other symptoms below, there are almost no failures where the magnetron causes the line fuse to blow. 2. Loose filament connectors (these may be repairable). There will often also be visual symptoms at the magnetron: Signs of overheating, such as discoloration; and evidence of carbon tracks or pits on magnetron terminals when the connectors are removed. An intermittent filament (internal) is also possible (but not repairable). Symptoms: No heat or erratic heat. The slip-on connectors can loosen, overheat, build up resistance and eventually loose contact. If the the magnetron terminal(s) have not been burned too severely, the connection(s) can usually be repaired. We prefer cleaning up the terminal, then soldering the filament wires directly to the terminal. Note: when discharging HV capacitor, since there is no load, it may end up being charged to a much higher voltage than is normal. Be prepared for a larger spark if you use a screwdriver to discharge it! 3. Open filament. Symptoms: No heat. See note about HV capacitor in (2) above. 4. In the older glass-dome models, the vacuum envelope can rupture. Symptoms: No heat, loud buzz due to arcing when entering cook cycle, possible blown HV fuse. See comments about fuses in (1) above. 5. Filament breakdown. Usually occurs after a few minutes of normal operation, possible blown HV fuse. Symptoms: No heat, loud hum once it occurs. See comments about fuses in (1) above. 6. Low output. Occurs as cathode emission decreases from long use. Symptoms: Reduced cooking power. 7. Moding. Occurs when magnetron oscillates in one or more undesirable frequencies. Symptoms: (a) Reduced or no cooking power, (b) RF interference. However, some food products (with high water content) may cook normally, whereas the result with other foods is very unsatisfactory. RF interference is possible but usually only occurs if there is actual structural damage to either the magnetron, its RF gasket or waveguide flange, or its RF (feed-through) capacitors. 8. Off frequency. Physical characteristics can change and cause magnetron to oscillate at frequencies slightly higher or lower than 2.45 GHz. Same as (7a) above. 9. RF leakage. Structural failure can cause leakage from magnetron housing. Symptoms: Microwave leakage into electronics bay, erratic control panel behavior. It can be very frustrating because the symptoms disappear when the oven's outer cover is removed. With the cover in place, the escaping RF energy is confined, and eventually builds up around the control panel circuitry causing unusual symptoms. 10. Insulation breakdown of the internal leads or at magnetron insulators or antenna terminal. Symptoms: Arcing, burning smell from magnetron, loud hum, no heat. 11. Cracked magnet(s). Symptoms: Reduced or no cooking power, magnetron overheating, occasional 'snapping' sound.
Depending on the age of your oven the magnetron may still be under warranty. Check the original paperwork that came with the oven - either the users manual or a separate warranty document. Contact the manufacturer if specific instructions on how to file claims are not provided. Full coverage on the magnetron of several years is common. If you have not sent in the warranty registration card (right, who actually does this?!), a copy of the sales receipt or other proof of date of purchase may be required. Both original and generic replacement magnetrons are available. Going direct to the oven manufacturer will guarantee a compatible magnetron but is by far the most expensive option. For a typical oven, one without the gold-plated trim :-), such a replacement may be more than half the cost of a similar new oven. In some cases (like Sears), you may need to convince their service department that you are qualified to be poking around inside one of *their* appliances before they will consider selling one to you (too many lawyers). In some cases, original magnetrons may also be available from parts suppliers like MCM Electronics - at somewhat less rediculous prices. They will be identified as 'original' or 'genuine' along with the manufacturer and their part number. Generic replacement magnetrons are available for the majority of microwave ovens. These will almost certainly be much less expensive than original parts. Essentially, there is only one type 'tube' (at least for any similar power range). The differences are mostly mechanical. However, quality may vary. In some cases, the generic variety may actually be better than the original. See the section: "Comments on replacement magnetron quality" for some recommendations.
(From John Gallawa (mtek@pen.net)). In my experience, mags purchased from after-market suppliers may or may not be OEM parts (there are not that many manufacturers of magnetrons in the world). Here's the interesting thing, though: In many cases, these after-market tubes are actually higher in quality than the original tube, as in the case of the OEM Sanyo magnetrons, which tend to fail prematurely. Of course, the opposite can also be true, depending on the after-market supplier. Some manufacturers, such as Toshiba and Hitachi, produce both high and low end magnetrons. They sell these under a variety of specialty names, as well as under manufacturer brand names. I have seen the low-end tubes in many brand-new microwave ovens. When buying magnetrons from other than the manufacturer, I have found it best to go to a supplier who specializes in microwave oven parts (i.e. AMI, Global Micro-parts, QB products). These sales people are usually more knowledgeable about the magnetrons they sell, and they can help you with proper choice and application.
WARNING: First, with power disconnected, discharge the high voltage capacitor. See the section "Safe discharging of the high voltage capacitor". When you receive the replacement, compare it with the original. It is critical that the replacement magnetron be mechanically identical: this means that the mounting configuration (studs or holes and their location), waveguide seating surface, and the orientation of the filament connections and cooling fins are the same. The studs may be removable so that the same assembly can be used with or without them. The cooling fins are particularly important as there must be adequate airflow from the fan for removal of the substantial waste heat - up to half of the input power to the magnetron ends up as heat. The shape of the antenna terminal - cone, bull nose, or square - doesn't matter. Magnetron replacement is generally straightforward but other assemblies like the cooling fan may need to be removed to gain access. Make careful notes of both the wiring and mechanical relationships. Usually, the magnetron is fastened to the waveguide with 4 nuts on studs. When removing it from its mounting, do not lose the RF gasket - a metal mesh ring which seals the connection against microwave leakage. Reuse it unless your replacement magnetron comes with a new one. Transfer any thermal protector to the new unit. Replace other components in reverse order and then reattach the filament and HV wires. Although the magnetron is a vacuum tube, there is probably no glass in yours (unless it is quite old) so it isn't really very fragile. However, a sharp blow or fall (during shipping as well if not properly packed) could shatter the filament. Do keep it (the magnets) away from your diskettes unless you want them bulk erased! As for the old one, see the section: "The magnets in dead magnetrons" :-).
WARNING: First, with power disconnected, discharge the high voltage capacitor. See the section "Safe discharging of the high voltage capacitor". * A shorted winding or short between a winding and the core/chassis in the HV transformer may result in a blown fuse, loud hum, overheating, audible arcing, a burnt aroma, or simply no heat. * An open winding will likely result in no heat but no other symptoms. Disconnect terminals as required to make the following tests: * The resistances of the primary should be .1 to .5 ohms (.2 ohms typical). * The resistance of the filament winding will likely be so low as not to be detectable with your multimeter. The only measurement easily made would be that there is no short to the chassis. * Typical resistance readings for the transformer HV secondary are in the 25 to 150 ohms range (depending on the power rating of the oven) from HV connection to chassis. A typical midsize might be 65 ohms. An open would be an obvious failure. However, based on the way these are wound, a winding-to-winding short would not cause enough of a resistance change to be detected with an ohmmeter unless you could compare with an identical model transformer from the same lot number. Testing the high voltage transformer more fully is difficult without fancy equipment. Only major short circuits can be identified in the transformer with an ohmmeter since the nominal resistance of the windings is unknown. However, open windings (not very likely) can be located and other faults can be identified by the process of elimination. Note: in the discussion below, it is assumed that the fuse is blowing due to a possible short in the HV transformer. Alternatively, there may be a loud hum as the HV transformer struggles due to a fault in the HV transformer or a shorted HV diode, magnetron, or a short in the HV wiring. Also note that depending on the severity of the fault, the fuse may not actually blow (at least not immediately) but there will likely be a loud hum when the HV transformer is powered. * Disconnect the primary of the HV transformer and initiate a cook cycle. If the fuse still blows, you have a problem elsewhere such as a defective interlock or shorted wire. * Assuming the fuse does not blow, unplug the oven and reconnect the primary of the HV transformer. * If the other components - HV diode, HV capacitor, magnetron - test out, remove the high voltage and filament connections to the transformer, power up the oven, and initiate a cook cycle. If the fuse does not blow, the transformer is likely good and there are still problems in the high voltage components. Possibly something is failing only when full voltage is applied. * If the fuse still blows, then the problem is likely with the triac (if used), a shorted wire, or shorted transformer. * If the fuse does not blow with the secondary isolated, reconnect only the magnetron filament (not the HV) to the transformer and power it up again. If the fuse now blows, then it is possible that the magnetron filament is shorted. * If your oven uses a triac, remove and bypass it. Now, if the fuse still blows when the oven is plugged in (door closed to enable the interlocks), the problem is likely with the transformer. Unplug the oven, discharge the HV capacitor. * Check for damaged wires that may be shorting to the chassis. Repair or replace these as necessary.
WARNING: First, with power disconnected, discharge the high voltage capacitor. See the section "Safe discharging of the high voltage capacitor". Replacement of a HV transformer is straightforward but other assemblies may be using the transformer bolts for their mounting and/or may block your way. Label the wires before pulling off the Fast-Ons if there is any doubt as to where they go. If the replacement transformer is not mechanically identical, you may need to use some creativity in anchoring it and any structures that are attached to its frame. However, the transformer must be secure - don't just sit it in place. Try not to drop either the old or new transformer on your foot!
WARNING: First, with power disconnected, discharge the high voltage capacitor. See the section "Safe discharging of the high voltage capacitor". Inspect the wiring - especially between the magnetron, HV transformer, and other components of the high voltage circuits for signs of arcing and excessive heating or burning. Arcing may be the result of the wire scraping against a sharp sheet metal edge due to poor placement and or vibration. A bit of electrical tape may be all that is needed. Since the magnetron filament in particular uses high current, any resistance at the press (Fast-On) connections will result in heating, weakening of the lug, more heating, and eventual failure or erratic operation. Try to pull off each of the lugs. They should not be loose - you should have to work at removing them. However, note that some lugs are of the locking variety and require that you push a little tab to release them. Check for loose, burnt, or deteriorated lugs in the filament circuit (not just the magnetron). If you find evidence of this: * Remove the lugs and clean the terminals with fine sandpaper or a file. If they are not too badly deteriorated, they will still work even if they are somewhat ugly. * If the lugs and their wire connections appear to be in good condition but come off their terminals easily, try squeezing them a little tighter with a pair of pliers and reinstall. Otherwise, cut off the old ones and replace them. * If any connections between the lug and the wire or HV diode are at all loose, solder it with a high wattage soldering iron or soldering gun. * Alternatively, use a drill to make a hole in each terminal, and then fasten the (tinned) wire directly (or better yet) a new ring lug to the terminal with a machine screw, nut, and lockwasher. Soldering is also possible. These approaches will work as long as there is enough metal remaining for a solid connection and may permit you to salvage a magnetron or HV transformer that would otherwise need to be replaced. Also check for bad solder connections between the terminals on the high voltage transformer and the enameled wire used for its windings. If you find anything suspect, scrape away the enamel and surface corrosion and resolder with a high wattage soldering iron or soldering gun.Go to [Next] segment
Go to [Table 'O Contents]