Should You Replace All The Components That Come To Repair Power Boare

---

---

Once you've establish a component yous want to examination, or one that'due south patently diddled, you need to remove it from the lath. Dorsum when all components were mounted on leads pushed through holes in single- or double-sided circuit boards, removal was piece of cake. A little solder wick or a pump of the solder sucker, and the holes would articulate. After that, all yous had to do was pull.

Sometimes the process is still like that, only at present in that location's much more variety of component styles requiring unlike removal techniques, and multilayer boards accept complicated the state of affairs. Component removal ranges from piffling to maddening, and information technology'south like shooting fish in a barrel to destroy the circuit board when a recalcitrant part simply refuses to budge.

Unless both sides of the lath are accessible, y'all'll have to remove it from the unit of measurement before you tin can desolder annihilation with leads poking through the board. Either style, first make sure power is disconnected. I e'er look at the AC plug before beginning to unscrew a board or desolder components, only so I know the plug is definitely lying loose. Even if I remember having pulled information technology, I take another wait.

Through-Hole Parts

FIG. 1 Using solder wick.

Many larger components still use the sometime wire-through-the-hole mounting technique.

To remove ability transistors and other through-hole parts, the solder must be sucked out of the hole, or the lead has to be pulled out while the solder is molten. Clearing the hole is preferable. For pocket-sized joints, employ solder wick, as described in Section 6.

Place the end of the wick on the joint you want to desolder, and then press the iron'southward tip on the other side. Hold information technology there for about 20 seconds, and the solder should catamenia up into the wick. See FIG. 1.

This doesn't always work, though. Sometimes the solder won't flow well enough to clear out the hole. The usual reason is insufficient heat, but transferring the heat to the joint is an outcome too, equally is thermal absorption by large copper lands. If you tin't become a small land'south hole to clear, try adding some fresh solder, and then wick it out again.

Boards manufactured with lead-complimentary solder don't desolder well. Adding leaded solder to a atomic number 82-gratis joint lowers the existing solder's melting temperature, making removal easier.

The wick absorbs some oestrus too, so it takes a hotter iron to desolder a lead than it does to solder information technology. Plus, to remove a lead requires wicking out all of the solder in the hole. With thick or multilayer boards, some of it may be a millimeter or more away from the heat source, making the solder hard to cook.

Desoldering is complicated by the increased thickness of multilayer boards and their extra heatsinking result from internal foils contacting the copper coating inside the holes. Applying enough heat to wick the solder out can destroy the lath.

To remove a stubborn lead from a multilayer board, information technology'due south best to oestrus 1 side while pulling the lead out on the other, and then clear the hole after the pb is gone. Even and so, you may struggle with information technology and be tempted to reach for the big soldering gun.

That's too much estrus for small boards, and it can deform them and pause internal connections in multilayer boards, wrecking the device. See FIG. 2.

Large lands used for ability supply and basis buses create a heck of a heatsinking upshot. It can be quite frustrating trying to get them hot enough to melt and clear the solder. The big gun might exist called for here, but it'southward still possible to trash the board because in that location may exist other lines running over the big land inside the layers. Heating upward the large land can interruption them or short them to their adjacent layers.

If a part won't come out no matter how hard y'all endeavor, it's a lot safer to clip the leads and solder in the new office without immigration the holes. Prune the new part's leads shut and solder them to the residual solder in the holes. You lot should be able to heat a hole plenty to brand a good joint, even if the solder at the far end of the hole never melts. If at that place's room, you can go out a little of the old function's leads and solder to those.

Sometimes y'all tin can't get to the leads to clip them. On most electrolytic caps, the leads are under the parts, unreachable with any tool. The easiest mode out is just to chop off the component most its base with a pair of wire cutters. Then yous can clip or desolder the leads easily.

Bigger joints with lots of solder can overwhelm solder wick, saturating it before much solder is removed. To clean out an entire big joint might require a foot of wick, which isn't cheap. These are jobs for solder suckers. Subsequently applying the sucker a few times, you should be left with only a coating of solder on the articulation. A sucker won't remove that, so finish upward with wick.

Every bit mentioned in Department 6, avoid using a bound-loaded solder sucker on static-sensitive components like CMOS chips and MOSFET transistors. The rapid release of the plunger can generate static charges capable of damaging those parts.

FIG. 2 How non to do information technology! Excessive heat destroyed this multilayer lath.

Surface-Mount Components

Wicking surface-mountain parts is easier because all of the solder is touching the wick, and many of the lands are very small and readily heated. Most surface-mount pads will desolder without incident. If the solder on a pocket-size land won't flow into the wick, try the same fob I described above: add some fresh solder to the joints before trying to desolder them.

Big lands on power supply and ground buses may still be hard to heat, but a normal iron will have care of most of them. Using a big gun on a surface-mount component is asking to destroy the part and quite maybe the board. Tiny SMT (surface-mountain technology) resistors and capacitors have sputtered-on solder pads.

Too much heat can delaminate them, making reconnection to the parts' bodies impossible to achieve.

Nigh SMT components are glued in place before being machine-soldered at the manufacturing plant. Very ofttimes, desoldering the ends of a role volition break the mucilage and free the role, but not ever. If yous see a red shellac-like hulk around the edges of the component, information technology'southward glued on and may not budge later on desoldering. To motion information technology, wick both ends and then heat one stop while pushing on the component'due south body with a pocket-sized screwdriver.

The tiny role may popular off of a sudden and smash away into oblivion if you're non conscientious.

Somewhere in the universe there must be a room full of sad, homeless SMT parts that flew off circuit boards, never to be found. Plenty of 'em came from my workbench.

Choosing Components

Any time you need a new part, you but breeze on downwards to your local electronics supply store, buy the verbal replacement and pop information technology in. Um, right, sure you exercise. Ah, if just existent life could exist like that! We don't fifty-fifty have local parts stores anymore. And while lots of standardized components are bachelor via mail-gild, many newer consumer electronics products aren't fabricated from them. Instead, they're stuffed with all kinds of obscure and specialized components nobody but the manufacturer tin provide. Luckily, in most cases you take a few options.

Junk Box

If you've stockpiled components, see if what you have is a shut enough match. When using parts that have been sitting around for a long time, take some fine sandpaper to the leads to remove oxidation that will have congenital upward. Otherwise, soldering to those leads will be unsuccessful.

Salvaging for Parts

There'due south a reason I've encouraged you to save boards from dead machines. Those from the same manufacturer as the unit of measurement you're repairing might apply the same component, even if they're a different model. Manufacturers relieve costs by reusing parts of their designs and techniques in lots of models. If you tin't observe an exact replacement, you nevertheless might locate something close plenty to work. Check all your parts machines, even those fabricated by other companies. You're more likely to detect a compatible role from the same type of machine, since the function is similar. So, if you lot need a part for a camcorder, check boards from those; you probably won't find what you demand in a DVD player. If y'all locate something you can utilize, but the leads are as well brusque, solder on a piffling wire to extend them.

Substitutes

Substituting a part with something close but not an exact match requires consideration of how the office is being used, what parameters are critical, and what you tin get away with in a item application. The general idea is that a role with better specs tin can sub for a lesser one, but not the other manner around. Even then, there are exceptions.

Different component types accept varying requirements. Let's look at some common ones.

Capacitors

Almost of the capacitors you lot'll replace are electrolytics. Tantalum caps fail pretty often also, but they aren't used much anymore, and so yous may never see one. The major factors in an electrolytic are its size, its capacitance, its voltage rating and its temperature rating. Besides, switching power supplies and estimator motherboards oftentimes require caps with particularly low ESR, to smooth out the fast, precipitous pulses those circuits produce. Replacing such parts with standard electrolytics will cause malfunction.

The most important consideration afterwards size-it does, later on all, accept to fit on the board in the allotted infinite-is voltage rating. Electrolytics just won't stand voltages college than their ratings, at least not for long; they fail catastrophically by shorting.

Their life is reduced even past running them at voltages under but close to their ratings, nonetheless some manufacturers volition utilize a xv-volt part at xiii volts, leading to frequent failures.

In cases like that, a replacement with a higher voltage rating than the original part is not just okay, it'due south desirable.

The capacitance rating is non as disquisitional as you lot might suppose. About electrolytics have rather broad tolerances, in the range of -20 to +80 per centum. If the cap is beingness used to couple signals from one phase to some other, the capacitance value is more important than it's when the function is a bypass or filtering cap. You might find a few electrolytic coupling caps in audio and video gear. In audio amplifier stages that use caps of a few microfarads from an emitter to footing, it pays to keep the value shut to the original, because a higher value might increase low-frequency response, upsetting the audio quality.

You will see tons of electrolytics in power supplies and for bypassing and filtering in all kinds of products, from unproblematic analog devices to today's most complex digital gear. Those are the parts that usually need replacement. If your available replacement's value is no more than 50 percent higher, go ahead and employ it. A footling actress filtering never injure anything, and +50 percentage is probable inside the stated tolerance of the original office anyway. To be sure the new part isn't at its maximum tolerance value of, say, 80 percent over the stated value, measure it with a capacitance meter. Despite their wide stated tolerances, most electrolytics I've measured take been within ±20 percent or so of their printed values.

Combining capacitors to become about the needed value is fine in most applications. I don't recommend it, though, for the big storage cap at the input of a switching ability supply (near the chopper), or in other high-voltage circuits. Putting caps in parallel adds their values, and putting them in series drops the final value according to the following formula:

Capacitors combine exactly opposite to how resistors exercise. For more info, see the section "Resistors" a bit later in the Section.

When putting polarized capacitors in series, be sure they connect + to -, and so you wind upwardly with one + and ane - at the ends of the string. When you parallel them, connect all the + terminals to each other and all the - terminals to each other. In either instance, be certain each capacitor'south voltage rating is equal to the entire applied voltage. When in series, the individual caps won't really exist subjected to the full voltage during normal operation, but a large voltage spike tin occur when ability is first applied, so it's a smart safety move to be sure every one of them tin can handle it.

Especially in power supply applications, the cap'southward temperature rating matters.

Electrolytics that get charged and discharged very fast, as they do in a switcher, can become enough warm from the power dissipation of their internal resistance. Standard electrolytics are rated to operate at 85° C, with college-temperature caps rated every bit loftier as 150° C. Manufacturers hate paying for things they don't need, so respect the temperature ratings if yous want the repair to last. For quick testing purposes while troubleshooting, you can disregard the ratings considering the part won't be running long enough to fail from overheating.

Tantalum capacitors should always be replaced with the same type. They accept lower impedance at high frequencies than do standard electrolytics, and are used only where that matters. Replacing a tantalum with a garden-multifariousness electrolytic will result in performance degradation or circuit failure. The capacitance tolerance of tantalums is much tighter than that of standard electrolytics, and then use a part with the same value.

An increased voltage rating is fine, withal.

Diodes

Diodes and rectifiers accept 4 primary characteristics: forrard voltage, reverse voltage, current and speed.

The forward voltage spec tells you how much voltage tin can exist beyond the part in its conducting direction. You won't often see this specified, considering in an AC circuit the forward and contrary voltages are usually the same. The opposite voltage is specified as PIV, for peak inverse voltage, and information technology tells you lot how much voltage the diode can withstand in its nonconducting direction. Exceed the PIV, and the part will arc over inside and be destroyed.

The current rating indicates how much current tin pass in the conducting direction without overheating the part and burning it out. No current should flow in the reverse direction, of course.

The speed of a diode is very important in some small-scale-signal applications similar radio signal detection. Information technology's also critical on the low-voltage side of a switching power supply, where the part volition be rectifying the fast pulses from the conversion transformer.

In the sections of power supplies operating at the low frequency of the AC line, any rectifier is more than fast enough. The span or individual rectifiers at the Air conditioning cord side of a switcher are not high-speed devices; nor are the rectifiers on the low-voltage side of a linear supply.

When subbing a normal, low-speed rectifier or bridge, pay attention to the PIV and the electric current rating. As long as those are equal to or college than the original part'southward ratings, the new part should work fine.

Wait up loftier-speed rectifiers in a commutation guide or online. Replace them with parts of equal or better PIV, electric current and speed. Never replace a high-speed rectifier with a low-speed part, fifty-fifty if the PIV and current specs are fine. It merely won't piece of work.

Some products use lots of glass small-scale-bespeak diodes. Wait for numbers like 1N914 and 1N4148. They're interchangeable. Even if you see no number on the diode, either of those numbers should work fine. Just exist sure the diode y'all're replacing is in fact a simple diode and not a zener. There are another special-purpose diodes, too, including germanium diodes (also drinking glass just noticeably larger than normal silicon diodes), gallium-arsenide diodes, tunnel diodes, and varactors. They're constitute in receiver front ends and other weak-signal, exotic applications. You won't run across them very often, but they must be replaced with diodes of the same types.

Resistors

Many resistors are carbon composition types and easy to sub. What matter most are the resistance value and the power dissipation capability. It's fine to use a 1 percent precision resistor in place of a standard v percentage one, and it doesn't hurt if the replacement is rated to handle more power.

If the original resistor was a special blazon, such as a wire-wound or depression-noise part, it's important to replace it with the same type. Those kinds of parts are used but in special applications. You won't notice them very often in consumer electronics gear, simply they show upwards now and so in switching power supplies, preamps and stages handling particularly pocket-sized signals, like receiver forepart ends.

If y'all can't notice the exact value you need, consider the original function's tolerance (encounter Section 7), and try to combine a few other resistors to become to a value well within the original function'due south specs. For instance, if yous demand a iii.3 k-ohm resistor, yous could put a 2.2 thou-ohm and a 1 k-ohm in serial. Resistor values in series add together, so that'd go y'all to 3.2 k-ohm. If the original resistor had a 5 percent tolerance, as well-nigh do, information technology could vary by ±165 ohms and still exist okay. So, three.2 one thousand-ohm would be fine as long every bit the combined resistors' ain tolerances didn't push their total value outside the tolerance range of the original part. Bank check the real value of the combination with your DMM to exist sure.

Resistors in parallel combine contrary to how capacitors practise. The resistance value goes down co-ordinate to the formula shown hither:

Two resistors of the same value will produce one-half the resistance. The larger the resistance of the second resistor, compared to the start, the less effect it has on it.

Play around with a few resistors by combining them in parallel and measuring them, and yous'll get the hang of it.

Transistors

Transistors are the almost complicated parts to substitute. Major semiconductor manufacturers used to give away big transistor substitution books filled with hundreds of pages of transistor types and their brands' appropriate cantankerous-referenced substitute office numbers. Because many types of transistors accept similar characteristics, a few hundred parts can sub for thousands of parts.

These days, yous can look up this stuff online, but you may see numbers for which you can't find a cross, or there might be a valid sub but you tin't get ane.

Alas, some parts are still fabricated of unobtainium. Even when a substitute component is available, you may adopt to speed up the repair process by using a function yous already have.

To choose your own substitute requires some understanding of the part'south application and how a change in characteristics would affect circuit performance. Some functions, like simple switching of voltage to direct it to diverse circuit stages or turn an indicator on and off, volition piece of work with only nigh any transistor of the same bones structure (bipolar or FET) and polarity. Others, such as high-frequency indicate processing or electric current amplification in complementary output stages, often crave stringent adherence to the original function's specs.

All this assumes you know the old office'south number. Usually you will, but at times you might take to fly blind. If the original transistor literally blew apart, which occasionally happens when a heck of a lot of current has been pulled through ane, there may non exist a number to read! I've seen SMT output transistors in LCD backlight inverters blow so difficult that at that place was little left betwixt the solder pads. Even when the number is visible, it could be a proprietary business firm number, with no cantankerous-reference to a sub. And some transistors, especially tiny SMTs, show no numbers in the first identify.

If you're lucky, the board will be marked with ECB or GDS, showing what terminal goes to which pad. ECB indicates emitter, collector and base of operations, thus a bipolar transistor.

GDS means gate, drain and source, the terminals of a FET. Those markings besides requite you strong clues to the part's polarity. If C goes to the positive side of things, it's an NPN. If E does, it'southward PNP. With a FET, if D is positive, it'due south an N-aqueduct part. If S is positive, it's P-aqueduct.

Without board markings or a role number, the transistor is a full mystery. Use your scope and understanding of bones transistor operation to deduce the role'south polarity and layout of connections. Start past looking for the power supply voltage feeding it. If it's positive and fed through a resistor or a transformer, yous've probably found the collector of an NPN transistor or the drain of an Northward-channel FET. Detect the stage'due south input by looking for whatever signal operates the transistor. If information technology's a continuous signal, you lot should run into it. If it's something that happens only when yous press a switch or some other operation signals that area of the excursion, create those weather and find the signal. When yous notice it, you've establish the base of operations or the gate. Whatever's left will be the emitter or the source.

Most bipolar transistors are NPN. If the connection to the positive supply line is direct, without a resistor, or there is a resistor but it'due south of very low value, the transistor could be PNP, and that connectedness would be its emitter. Find what looks like the base past scoping for signals. See if there's a resistor from the base to the transistor final closest to the supply. PNPs are used to turn on and pass current from the supply to another circuit when the input bespeak goes low, toward ground. The resistor going upwards toward the supply keeps the base high and the transistor turned off until the input signal pulls it low. You'll discover PNP circuits of that sort in power switching sections of battery-operated products.

Assume the function is an NPN bipolar transistor, and you'll be right about of the time.

If your replacement turns out to exist the incorrect polarity, the circuit won't work, but it shouldn't do any harm.

All bets are off if the transistor is part of a complementary button-pull amplifier.

They employ NPNs and PNPs in more than complicated, hard-to-deduce ways. And if the original role was a FET, the issues of enhancement and depletion mode, and JFET versus MOSFET, make the whole thing very tough to fathom. Getting the identification correct requires your understanding how those parts work and looking at the bias on the gate final to infer what the output should do as the input changes.

Don't effort to sub chopper transistors in switching power supplies without knowing the correct role number and finding a legitimate sub from a cantankerous-reference. Most choppers are power MOSFETs with specs that must exist closely matched for reliable operation. Even if a sort-of-close sub works, it probably won't run for long before declining. Sometimes even a legit sub will dice in a hurry, and the only part that will work is an exact replacement of the original part number. The aforementioned is true of horizontal output transistors in CRT TVs, some other awarding involving fast pulses at fairly high voltages and currents.

In some cases, the original and replacement transistors are electrically compatible but their organization of leads, called pin basing, is different. Near small-indicate American transistors are EBC, left to correct, while Japanese parts are usually ECB. You tin can supersede one layout with the other equally long every bit you switch the two leads, being careful not to let them touch as they rising from the board toward the transistor. Small FETs are usually SGD. Power transistors are usually BCE, with C connected to the metallic tab (if at that place is 1), just cheque to make certain. Ability FETs typically use GDS, with D continued to the tab.

Once you've figured out what should go where, whether from the original part or from scoping and deducing, y'all can go on with trying out a new part. The master characteristics to be concerned with are proceeds, loftier-frequency cutoff point and, with larger parts, ability dissipation adequacy. Secondary characteristics, just notwithstanding very of import, are the maximum voltages permitted from base to emitter and from collector to emitter.

Very oft yous'll find a transistor that's pretty close but has a little more or less proceeds. Depending on the application, that might work. If the circuit is linear, producing output proportional to the input, the transistor isn't normally saturated (fully turned on), and so a slight proceeds difference may not cause a problem. In switching circuits like backlight inverters, though, inadequate gain can event in lots of oestrus from the transistor'due south not-fully-turned-on resistance, burning out the part in a hurry. Too much proceeds in a linear circuit may cause distortion, increased output or spurious signals. Not enough commonly just results in a bit less output.

The high-frequency cutoff bespeak specifies at what frequency the transistor's gain will have decreased to one. In other words, it won't be amplifying at or to a higher place that frequency. In low-frequency applications such as audio, any transistor will be more than fast plenty. At radio frequencies, the situation can exist quite different, requiring a transistor whose cutoff frequency is approximately equal to the original role'due south spec.

As well little might outcome in low or no output, while too much could upshot in unwanted harmonics or spurious signals riding on the desired one. When in doubt, go for too much, every bit long equally the difference isn't excessive; at least the thing volition try to work.

Power dissipation is very important. The new office should be able to dissipate at to the lowest degree as much power as the old ane. A ameliorate dissipation spec is fine.

Maximum permissible inter-electrode voltages must be respected. Exceed them and the transistor might emit some of that magic smoke. Near transistors' collector to-emitter specs are well beyond what a small-scale-signal circuit produces. The excursion'due south base of operations-to-emitter voltage, however, could exceed the capabilities of some replacement parts, so keep an eye on that. Big parts used in output stages tin have pretty high voltages practical from collector to emitter, then don't accept that spec for granted.

If all this seems overwhelming, stick to replacement office numbers from a cross reference guide or online source, and yous'll be fine. Even with expertise, matching up transistors is very much a roll of the die. Run into, I told you it could become complicated!

Zeners

The purpose of a zener diode is to break downwardly nondestructively in the opposite direction and comport when the part'south contrary voltage spec, or zener voltage, is reached. The important specs are the zener voltage and the power dissipation. Dissimilar normal diodes, zeners' dissipation limits are specified in watts, not amps. Always supersede a zener with one of the same zener voltage and at least as much dissipation capability. A higher dissipation spec is fine.

Yous tin put zeners in series to add their voltages, but don't parallel them to increment dissipation adequacy; even zeners with the same zener voltage won't start conducting at exactly the same voltage, so one will always take more current than the other, resulting in its premature failure. When combining them in series, be sure that the wattage of each zener is at least as high every bit the original role's rating, and watch the polarity. Each zener should feed the next one cathode to anode, and then you air current up with one anode and 1 cathode at the ends of the string.

Installing the New Parts

Once you've procured or substituted components, it's fourth dimension to put them in! Proper installation is crucial for successful, long-term repair. Permit'due south look at some issues specific to diverse kinds of parts.

Through-Hole

Replacing a through-hole component is pretty easy, requiring nothing more than pushing the leads through the holes, bending the ends a little and so the role doesn't autumn out, soldering the leads and and so clipping off the excess.

If the part is attached to a heatsink, it'southward a trivial more complicated, but non much.

For a free-floating heatsink bolted or clamped to the elevation of the component, install the heatsink earlier soldering the part to the board. When the part mounts on a fixed heatsink, put the leads through the board's holes without soldering them, and then spiral or prune the component to the heatsink. Solder the leads only after the mounting procedure is complete.

If the original function used heatsink grease, you need to do the same with the new one. The grease used is a special silicone compound formulated for maximum estrus transfer. You tin can become information technology from online parts houses, and estimator supply shops that behave CPU upgrades and bare motherboards likewise carry it. Most heatsinks, including those with mica or sparse plastic insulators, do require the grease. Those with safe separators usually don't, though. FIG. 3 shows typical insulator setups requiring thermal grease.

FIG. three Transistor mounting hardware with screw sleeves

A sparse smear of the special grease on ane of the mating surfaces helps oestrus transfer across the less-than-perfect contact area, filling in tiny gaps and increasing effective surface area. Likewise much grease can separate and insulate the surfaces, reducing heat flow, so don't overdo it. Smear on the grease with a swab, and exist careful not to put angle pressure level on the insulator or it may break. Mica insulators are especially brittle, and even a single fissure can lead to a short later on on. To avoid bending it, place the insulator on your workbench before applying the grease.

The insulator's task is to isolate electric contact between the component and the heatsink while facilitating heat transfer. If there's no insulator, either the part has no contact point on its case, as with an all-plastic transistor, or it'south okay for it to be continued to the heatsink. Voltage regulators sometimes have their ground connections on the metal tab, and then contact with a grounded heatsink is a adept thing.

Many power transistors, though, have their collectors or drains at the tab. Those are ordinarily connected to voltage sources, and contact with ground would be a short. Insulators are used to avoid the connection.

When there is an insulator, and the component has a metallic tab, the mounting screw volition pass through a plastic washer with a sleeve. Exist certain to use information technology, and watch its orientation. The sleeve should fit into the pigsty on the transistor'due south tab, preventing the spiral from touching the inside of the hole.

Tighten the mounting spiral more than than you would a screw holding a lath down or a case together. Yous desire skilful heat transfer, and that takes some force per unit area. Don't overdo it to the point of breaking the insulator or stripping the screw, of course.

Occasionally, you will discover a thermistor (a heat-sensitive resistor) glued to the instance of a ability transistor, especially in the output stage of a push button-pull sound amplifier.

Thermistors are used to accommodate the bias of bipolar power transistors as the parts heat upwards, because their gain and optimum bias point drift with temperature. If you can become the thermistor off without destroying it, glue it with epoxy to the new role. If yous can't remove information technology, you lot'll need a new thermistor. Await up its part number and order one just similar information technology.

SMT

Putting in a new SMT function is a scrap tougher than installing a through-hole component, thank you to the size scale. How practise you hold it in place long enough for soldering? Gluing is non recommended. Sure, the manufacturers practice it, but they take special glue made for the purpose, and we don't. More than likely, some other line runs underneath the component, and a later endeavour to remove the gum volition tear the copper off the board.

Also, the electric properties of the glue yous might utilise are a wildcard; you have no idea how its presence might affect circuit performance. Information technology could exhibit capacitance or even conduct current.

To get an SMT in place, commencement employ wick to make clean the board's solder pads so that there are no raised bumps of solder on them. You lot want the SMT to prevarication flat. Cook a little solder onto your iron's tip. Now place the part on the lath and line it up carefully with a tiny screwdriver. Center it betwixt the pads so it tin't create a short across two lands. Concord down the body of the SMT with the screwdriver while touching the iron's tip to one finish of the part. The solder on the tip should menses onto the board, making a joint at that terminate. Don't worry almost getting a good joint; all you lot want to do is prevent the function from moving.

Once the component is held in place past the solder on one finish, solder the other finish properly. Then become back and redo the messy end. Take a adept, close look with your magnifier to be certain y'all haven't created whatever solder bridges to adjacent pads or parts.

Finding Parts

Proprietary components have to be procured from the manufacturer (unlikely these days, simply worth a try), the component maker who supplied them (possible) or a parts unit. For popular gadgets, finding a parts unit may be the easiest way to become. Check eBay.

Standard components are widely bachelor through online post-guild, just many parts houses have minimums, so y'all might have to spend a lot more than the part is worth. Oh well, you lot can e'er stock your components supply with other goodies you lot might utilise later. Or, you can salve up your parts needs until you have a big enough order. That'll delay your repair work a long time, though.

Here are some places to await for components:

RadioShack (radioshack.com) • This seller's parts diverseness is small, but the visitor offers a few transistors and chips, along with standard five percent resistor values and some electrolytics.

DigiKey (digikey.com) • This mail service-order parts business firm has just near everything y'all could ever desire. Its catalog is overwhelming, and you can download information technology equally a PDF.

Mouser Electronics (mouser.com) • Another powerhouse, Mouser has a wide variety of components, including many used in consumer electronics devices.

All Electronics (allelectronics.com) • This is a surplus house with lots of interesting cloth at bargain prices. It has inexpensive, generic backlight inverter boards that can exist retrofitted to LCD monitors, though the boards lack terminals for brightness control.

Do an online search and you'll turn up dozens more sources for both prime and surplus components.

Saving Damaged Boards

When you lot desolder a through-hole component, ane unfortunate issue of failing to get the hole hot plenty is that its copper lining comes out with the atomic number 82. If you see what looks similar a sleeve around the lead, y'all've torn out the copper. On a double-sided board, information technology's not a ending. When you replace the function, be sure to solder both the top and lesser contact points, and all volition be well. Y'all might take to scrape some of the green solder mask coating off the pinnacle area to get contact betwixt the lead and the foil. That's best done with the tip of an Ten-Acto knife.

Pulling the sleeve out of a multilayer lath tin destroy it because you have no mode to reconnect with interior foil layers that were in contact with the sleeve. If yous're lucky, that particular pigsty might not have had inner contacts, and soldering to the superlative and bottom may save the 24-hour interval, so it's worth a attempt. Don't exist surprised, though, if the circuit no longer works.

If y'all tin can effigy out where they get, cleaved connections tin exist jumped with wire.

On double-sided boards, it'due south not besides hard to trace the lines visually, though you may have to flip the board over a few times every bit you follow the path. When you observe where a broken trace went, verify continuity with your DMM, from the end back to the suspension.

Don't forget to scrape off the solder mask where you want to contact the broken line.

Wire jumping can help save boards with bad conductive glue interconnects, too.

On a double-sided lath, you tin can scrape out the glue and run a strand of bare wire through the pigsty, soldering information technology to either side. Forget about trying this on a multilayer board, notwithstanding; you'll probably trash information technology while trying to clean the hole. On those, it'south best to run an insulated wire around the board from one side to the other. That adds actress length to the conductive path, which could cause problems in some critical circuits, peculiarly those operating at loftier frequencies. At audio frequencies, information technology should be fine. If some interior layers are no longer making contact with the glue, this won't work. Almost conductive mucilage boards I've seen have been double-sided, making them suitable for wire jumping.

If the lath is cracked from, say, having taken a fall, scrape the ends of the copper lines at the crevice. Information technology's possible to simply solder over them, bridging the crack, but that technique tends to exist less permanent than placing very fine wire over the break and soldering on either side. To become wire fine enough, look through your stash of parts machines for some pocket-size-estimate stranded wire. Pare it, untwist it and remove a single strand.

Sometimes in that location are multiple broken lines likewise close to each other for soldering without creating shorts between them. To salvage boards like that, scrape the solder mask off close to the cleft on every other line. Then scrape the in-between lines farther away from the cleft. Use the bare wire strands to fix the shut set, and use wire-wrap wire (very thin, unmarried-strand, insulated wire used with wire-wrap guns for prototyping experimental circuitry) or enamel-insulated "magnet wire" to spring the farther set. Wire-wrap wire is specially good for this kind of piece of work because its insulation doesn't melt very easily, so it won't crawl upwards the wire when you solder shut to information technology, exposing blank wire that could short to the repaired lines nearby. Plus, it's sparse enough to fit in pretty small spaces. For fifty-fifty tighter environments, utilise the magnet wire. Simply be sure to tin the ends of the wire to remove the enamel, and so you'll become a adept connection.

It's possible to repair broken ribbon cables in stationary applications (the ribbon doesn't move or flex), if they are the copper-conductor type of ribbons, non the very sparse, printed way. Fixing cracks with wire is a slow, time-consuming technique, but information technology works. Accomplishing it without causing shorts takes practice and isn't e'er possible with very small, dense boards and ribbons.

On multilayer boards, cracks and torn sleeves are extremely difficult to bypass. If you have a schematic, you may be able to find the path and jump with wire. Without one, it'due south pretty much impossible when the tracks are inside the board.

LSI

Back in Section 7, I promised to describe a trick for resoldering big ICs with very close pb spacing. Those large-scale integrated (LSI) chips with 100 leads are in just about everything these days. It's not probable you could notice a replacement fleck, and then why would y'all want to resolder one?

With so many leads, an intermittent connectedness to an LSI is not uncommon. SMT boards are mill-assembled with reflow soldering, in which solder is applied to the pads so reflowed onto the component leads with hot air, infrared lamps or in a special oven. Reflow soldering relies on low-temperature solder that can break after the numerous heating and cooling cycles encountered in a product's normal apply. Now and so one connection out of an LSI's long row of them will go flaky. The leads are so close together that there's no manner to apply solder to one without causing a brusque to the adjacent leads.

Here's the fob: get alee and short them! With all power removed, of course, solder away and let as many leads get shorted together as you want. In one case you accept good solder on the problem lead, lay solder wick across the leads where the excess solder is shorting them. Estrus information technology up and wick off the excess, but don't wait until the leads are os-dry. Pull the wick off a little sooner. If you lot become the timing right, you'll be left with a perfectly soldered row, with no shorts. The wick soaks up the solder in between leads faster than what's underneath them (where you want it to stay).

If you lot look as well long and current of air up removing and then much solder that the connections to the lath aren't solid anymore, resolder the area and do the wick trick again. After you endeavor this procedure a few times, you'll get the hang of how long to await before pulling the wick. I've had tremendous success with this arroyo. The 1 caveat is that information technology's hard to wick out solder if it gets under the edges of the fleck. To avert that problem, solder as far from the body of the IC as possible. That helps forbid overheating the fleck, too.

When yous're all washed, employ your magnifier to verify that the contact points with the board are soldered, and that no bridges exist between leads. Honest, information technology really does work! I've fifty-fifty replaced a few LSIs this way, using chips from parts units. Getting those babies lined upwards accurately on all four sides ... well, that's another story.

Source: http://www.industrial-electronics.com/Diagnose-Repair-Elec_12.html

Posted by: hartmuns1953.blogspot.com

Share this post