SERVICE TIPS
For your convenience, I'm starting to compile a list of faults for electronic things that I have repaired. Many of the items listed have been repaired without any circuit diagrams available. A point that may be of use is that when servicing stuff that uses IC's and you have no circuit, get the IC data off the internet. Generally, the circuits around the IC will be very close, if not identical, to the examples provided in the data.

HP 1702 LCD monitor stays on for a few secs after power up then goes blank. LED stays green but no image. Shine a torch into the screen to verify it's just the backlight gone off. Replace the 100uF 400V filter capacitor fed by the mains rectifier. Cap tests OK in terms of capacitance, but ESR is too high causing supply voltage to drop under load (i.e. when the backlight comes on).

On a separate issue, why do manufacturers continue to use .47uF and 1uF electros? These are notoriously unreliable, especially in switchmode power supplies. An MKT type has lower ESR, temperature stability and fits in the same space.

Dreambox 7025 satellite receiver. Power supply: 12V standby suppy working but main supply starts up for a fraction of a second. Replace all the electros on the primary side. Because of heat, they have all gone low in value. Fortunately, this doesn't seem to be a design that self destructs when this happens. Use an MKT for the 1uF replacement. To activate main supply, connect first two pins of multipin plug together with 10K resistor. 1st pin is 12V all the time (standby) & 2nd pin is the power control. Electros near heatsinks are a recipe for problems. I recommend a fan be installed for anything that has a switchmode supply that runs hot.

HP D530S desktop PC. Major production fault in the power supply. If the computer is unplugged from the mains for a while (~few weeks), the PSU will blow up when next plugged in. The problem is the brown glue around the main switching FET becomes conductive and absorbs moisture when the power supply has cooled down (i.e. when not plugged in). Being a high impedance device, the FET  doesn't need much leakage between its gate and drain terminals to switch on, which it happily does so upon next power up, discharging the filter electro through itself and the control IC. The 22R also blows.
If you have one of these computers, remove all the brown glue on the power supply PCB while you still have a working PSU. Either that, or never turn off at the mains.

LG RH1777 DVD/HDD Recorder. Dead power supply. 1) STR-W6251 s/c between pin 4 & 5. Replacement obtained from Statewide Appliance Spares. This company also supplies the remote control. 2) ZD103 s/c. This is a 12V zener. 3) optocoupler LED o/c.
4) IC103 TL431 s/c between adjust & ground pins. Power supply will not run if there is no feedback from the optocoupler. Normally, this would result in the switchmode IC running full blast and then self destructing, but the STR-W625 is well protected by not allowing it to run in a fault condition. There is an intial burst on power up, the time of which is determined by the capacitor on pin 7. This burst is enough for the voltage reference IC on the secondary side to cause the optocoupler LED to power up, resulting in pin 6 of the STR-W6251 being controlled. Data for this IC was a pain to find, but does have useful test circuits.

Theben time switch. This DIN rail mounted 240V time switch had no display. The supply for the relay is about 35V DC which is obtained from a .15uF dropper capacitor and bridge rectifier. From this 35V supply, a supercap is charged to about 3.7V to drive the clock. No 35V supply was present, but everything worked when external DC was fed into the filter cap on the DC side of the bridge rectifier. There is a .1uF surface mount cap on the AC input of the bridge. This was s/c.

Dick Smith and Jaycar Inverters.  DSE M5110 Inverter output too high. R26 1.3M gone o/c in regulator feedback circuit. M5300 blows fuse when turned on. One set of output FET's shorted. I used IRF840's as replacements as they are higher rated and cheaper than the IRF730's.
These inverters are widely available under differing brands, apart from DSE & Jaycar's. All operate on the same principle, and I've repaired quite a number now with essentially the same faults.
The 12VDC supply is stepped up to about 340VDC via a ferrite cored transformer. A set of high current FET's drives the transformer at a supersonic frequency. Output is rectified by high speed diodes and filtered by an electrolytic capacitor of around 100uF. The high voltage DC is then converted to AC by means of high voltage FET's in an H bridge configuration, the output of which feeds the 240VAC socket. The four FETS are driven by either a free running or crystal controlled 50c/s oscillator. The pulse width is set so that the output is 240Vrms. The driving waveforms are opto coupled into the FET's to ensure isolation, or the oscillator circuit is run off its own low voltage supply from an extra winding on the ferrite cored transformer. Usual fault is blown output FET's. Remove all four of them and do ohmmeter tests; the faulty ones will be all s/c. Check the driver transistors if used (MPSA44 and the like). Often these survive. Run the inverter with no FET's and see the main filter electro has around 300-340V across its terminals. This will indicate the DC-DC converter is OK. I have had one where the voltage was too high and the capacitor had exuded electrolyte. The waveform being fed into the transformer was full of spurious oscillations. One of the low voltage FET's was crook. Once correct DC-DC converter operation has been checked, also check the low value current sense resistor. This may have been blown open when the 100uF discharged into the shorted FET's. Replace the FET's and run the inverter off a current limited power supply; limit to about 1.5A. Check the output waveform. One inverter I worked on was actually giving out DC which did not show up with an incandescent bulb as a load. Larger inverters, like the 800W ones use the same principle but more FET's in parallel.

Sanyo VTC9300 Betacord VCR. 1)Sluggish tape loading. Rollers supporting loading ring sticking. Clean their support posts, and carefully lubricate. One drop of very fine oil will do. Applicable also to U-Matic VCR's. 2)No play, but rewind and FF ok. Pinch roller is not up against the capstan because the pinch roller post has fallen through the pinch roller arm and is catching on the loading ring. Remove pinch roller arm. As it's made of muck metal, it may have swollen slightly and be tight on its post. Remove carefully. A drop of CRC may help. Remove pinch roller and tighten post. I used pliers to burr it slightly and pushed back into the arm. You could use a drop of superglue. The bottom of the post must be flush with the arm. If the pinch roller arm moves tightly on its support post, gently use a round file to enlarge the hole. 3)Unlocked colour (rainbow effect) on play. VR214 on board W2 (the big board underneath) is the colour AFC. Gently tweak it until colour always locks.

JVC HR3660  VHS VCR. Intermittent tape loading, FF & rewind, even though heads spin etc. Gradually gets worse. The two main flat drive belts are slipping. The loading up functions and reel drive are done by the motor just to the right of the cassette housing. If you see the pulley spinning but the belt not moving, that's the problem. This belt drives another pulley which reemerges on the underside of the chassis to drive a second flat belt. This may also be slipping. Generally, mechanical things not happening are belt related in this model. Once or twice belts have come off due to a jammed tape. The reliability of these machines is incredible. These are almost 30 years old and out of both of mine, have only had to replace a few belts (once), a cassette lamp, and an RF amplifier IC after a storm.

Early VHS and U-Matic VCR's, Various Models. No tape functions. Check the cassette lamp. If this goes open circuit, the microprocessor shuts down disabling mechanical functions. The cassette lamp is used to sense the end of tape (transparent leaders). Later VCR's use IR leds instead of an incandescent bulb.
If the machine seems to randomly shut off when servicing, see that any room lighting or sunlight is not striking the optical sensors. I recall one JVC CR6060 U-Matic that would shut off in the morning when looking for a totally unrelated fault. The sun was coming through the window at a particular time, striking the opto sensor which made the machine think it had got to the end of the tape. Betamax machines detect metallic leader tape and do not exhibit this problem.

Crompton PIR sensor switch.
Fault was that although the light could be switched on by a rapid turn off and turn on of the supply, in the usual way, the PIR part was non functional. I took the LDR out of circuit so that it would think it was night and thus function with ordinary room lighting whilst working on it. Tried changing the LM324 first. Touching the back of the PCB actually made it work, so it was obviously a resistor gone high. The 100K resistor next to VR1 had risen to 245K.

Fisher & Paykel GW508 washing machine motor controller (phase 4).
The 15.5V power supply had blown up; more specifically, the TOP224Y switchmode IC, the P6K200E transient suppressor diode, and what appear to be the zener diodes for the error control. Of course, the 4A fuse had blown also. For those not au fait with these machines, F&P use a three phase stepper motor rather than a conventional induction motor and gearbox. This allows an extremely high spin speed, etc. The stepper motor is switched by MOSFETS fed from 340V DC (rectified and filtered 240V AC). The 340V DC also feeds the switchmode power supply that provides 15.5V for the control circuit and solenoid valves. The pump is conventional and is switched by a Triac from the 240V AC supply. So, it can be imagined that any spikes on the mains can easily destroy the motor controller. A novel feature is that the MOSFET heatsink is water cooled. Incoming water to fill the tub is circulated through a length of aluminium tubing to which the power semiconductors are clamped.
First point of call was to feed 15.5V from a bench supply into the controller - it actually looked very promising with LED's lighting up and the buzzer doing its start up sound. None of the motor switching MOSFET's tested short circuit. Given the time taken to get the parts to repair the 15.5V supply, and that it would still be a typical vulnerable SMPS, the decision was to make an external linear supply to provide 15V. The damaged parts were removed and also the switchmode transformer. A 15V 1.2A transformer, rectifier, fuse, 3300uF cap and 7815 regulator assembled in a plastic box did the trick. Wires were run out of the motor controller to the new black box secured to the back of the machine. Complete success.
Note that the 15.5V circuit is all live at mains potential. Therefore the external power supply must be completely insulated and connecting wires rated for 240V AC.
I replaced the 4A PCB mount fuse with an M205 fuse. It just so happens the PCB allows the use of an M205 fuseholder. Also note the PCB is covered in a plastic layer - I used 3M clear protective coating to restore this. With 340V DC circulating, it won't take much leakage to destroy something.
While these F&P machines give an excellent wash and have a fantastic fault diagnosis system, they are unreliable and the internet has many stories of problems. This particular machine has over the last few years had the pump, lid switch, hot solenoid valve, and balance switch replaced. Luckily these parts are cheap on ebay. It was only a matter of time before something went wrong with the yellow box...
I would advise these machines to be unplugged from the mains when not in use - the SMPS operates and 340V DC is present while ever the mains supply is on. A surge protected power board is also a wise idea.

Home

cablehack at yahoo dot com