That’s how it came about that I was into repairing a 2005 Panasonic 5.8GHz Digital Cordless Answering Phone System (model KX-TG5833ALM) – modestly called "GIGARANGE". Its problem seemed relatively simple and that was that it was unable to charge its battery from its main base station although it could with another charger unit.

It didn’t take long to disassemble the base station to reveal a complex motherboard with microprocessors and all sorts of other surface-mounted gizmos. I quickly realised that my enthusiasm was no substitute for a service manual. Fortunately, I was able to access one of these and had to laugh when I read at the top of the front page "5.8GHz Digital Answering System with Tree (sic) Handsets". I could see I was going to get my hands dirty!

The battery used in the handsets was a 3.6V NiMH 830mAh rechargeable (HHR-P104) and the AC adaptor for the system is a 9V 500mA plugpack (PQLV1AL). The battery was absolutely flat. When the handset is placed into its cradle in the base unit, the red charge LED comes on. The voltage on the "charge" contacts without the phone is a pulsating 9V.

What I didn’t realise was that the power and battery (6-hour) charge system is a sophisticated microprocessor controlled operation in both the base station and handset, where EEPROMs can also be addressed, reset and adjusted via software. Bearing in mind that I could still recharge the handset battery with another charger (KX-TCA1AL2), I felt the problem must lie within the main base station.

In order to measure the voltages when in the charge mode, I needed to disassemble the handset and then connect it to the base station via crocodile clips. It was then that I realised that the input voltage dropped from 9V to less than 2V, which was just too low to charge the 3.6V battery.

So where was the voltage going? Obviously there wasn’t a short in the handset or the base station because when disconnected it rose to 9V. And the act of the connections mating didn’t induce any kind of short.

The main voltage trail from the +9V AC adaptor to the positive charge connecting button goes through D301, D362 and L361. The return path from the negative button connector is via L371, Q371 and R371, R372 & R373 in parallel, ignoring all the peripheral sensor and control circuits also hanging off these.

While under load I measured the voltage drop across each major component until I got to D362 where I found there was nearly 9V across it. Using a DVM, I almost didn’t quite get the significance of this as the meter gives a 9V reading whichever way around you connect it. However, the crux of the reading was that it was +9V with the positive meter lead on the anode and the negative on the cathode; ie, there was 9V across the diode in the forward bias condition instead of only 0.6V.

Obtaining a replacement surface-
mounted diode (Part No. BOECK-000008) wasn’t straightforward, as the part number could not be found in the Panasonic database. However, with persistence, I eventually managed to get a substitute which fixed the problem.

Resurrecting a revered Revox

In keeping with unfamiliar stuff, a dead 1967 Revox A77 MkIII Dolby reel-to-reel tape recorder came to me recently. This Swiss-designed and made classic deck is beautifully constructed despite its 40 years. It was one of the first fully solid-state decks with three motors, plug-in modules and solenoid controls for just about everything. It also had an electronic speed governor.

Unfortunately, this particular unit had had a hard life and was pretty rough on its exterior. I trawled the internet for a service manual where I began to learn just how highly regarded this piece of technology was as the standard in semi-professional reel-to-reel tape recorders.

The mains fuse had blown on this set due to capacitor C115 (0.47μF) having gone short circuit and taking series resistor R123 (4.7Ω) with it. Replacing these got things going but the set really needed an overhaul. All the PC boards, including the 21V B+ control P106 board, and all the module edge connectors needed cleaning. In addition, all the grey round hard plastic electrolytic capacitors needed replacing, especially C210, C211, C405, C425, C507, C509 & C813. The 24V auto-shutoff lamp had also failed and though I was tempted to replace it with a LED, I made the effort to get an original.

I am sure that enthusiasts would have many other tips available about this highly regarded deck but the work I did was enough to get this old bird back to work.

Screen burn

Last month, I touched on some unusual faults I have had with rear projection TVs. I had another interesting case recently, concerning a very popular manufacturer which has subsequently removed the product from the market entirely. This is about a 60-inch LCD 16:9 rear projection monitor (manufactured in about 2000) which had a problem due to the fact that it had been connected to a digital set top box (STB).

The set came in with two slightly green stripes running up each side of the screen where the colour changed hue. When a pure colour was displayed (eg, red, green or blue) or even sometimes white (depending on its saturation), it would show no problems. However, on grey or a detailed picture, the two bands would show quite markedly. These two bands were like a CRT screen burn but lacking slightly in contrast.

Indeed, this was in fact the LCD version of a CRT burn. The set had been run in a prolonged 4:3 format and the two stripes were equivalent to the black border between the 4:3 and 16:9 formats.

Unlike CRTs, LCD (and plasma) elements get hottest when the picture is black because they have to cut off and absorb the light from the lamp. This light is extremely bright and very hot. In fact, the instruction book actually advises the owner not to use it in the 4:3 mode for precisely this reason.

Of course, there was nothing that could be done. The optical engine had been damaged and that was that.

Two tricky Philips sets

We had two Philips sets come in recently with one thing in common – apart from not working properly that is! One was a 36PW9525/79R and the other a 29PT9418/79T but both used the MG3.1 chassis. The latter was a 100Hz TV with twin tuners and had no colour, while the former was stuck in the Service Access Menu mode.

I chose to tackle the set with the Service Access Menu (SAM) problem first. This set had good pictures and sound as well as being almost fully functional with the remote but the SAM was permanently superimposed on top. I could navigate and change the functions within the menu but whenever I tried to exit the menu it just kept coming back. Normally, the SAM is exited via the "MENU" command or by switching off and on with the main switch but this just wasn’t happening – even though it was obviously trying to.

There were no error codes in the buffer and the set did its functional test perfectly. The Option Codes were all correct too. It was as though someone was constantly trying to put it into the SAM mode. This can be done with the remote and also by shorting pins 1 & 2 of connection O356 on the Small Signal Panel (SSP). I disconnected the remote control receiver via plug O341 (pin 3) which made no difference, so I concentrated on the SSP service pins.

Pin 2 is ground and pin 1 has a 6.5V zener diode 6012 connecting it to ground. It then goes via a 470Ω resistor to pin 119 of IC7003, the OTC microprocessor. I measured the voltage on this line to find it low at 0.9V, as opposed to its cousin Service Default Mode pin 3 which was at 3.83V.

This suggested to me that this rail was being held low – but by what? The ohmmeter gave no clue as it showed both rails to have the same impedance. I unsoldered the surface-mounted zener diode but that too made no difference.

The only other clue I had was that this set was normally kept near the beach and the SSP was slightly corroded. So I removed and washed the board thoroughly before examining it carefully under a maggie-lamp. Using a continuity tester, I could find nothing to cause this problem. Besides which, if any track had gone open, surely it would remove the load on this line, not actually load it down.

Next, I tried using a link between the two service pins to see if pin 1 would rise to the same voltage as the other. It wouldn’t so I concluded it was more to do with a load problem rather than a resistor strapping it to a line that had gone open circuit – if indeed there was one (I couldn’t find one on the circuit diagram).

In the end, I finally assumed that there was a problem with the OTC microprocessor (IC7003) but I wasn’t about to replace a 120-pin high-density surface-mounted IC. An exchange board, even if available, would be expensive and in any case, a corroded one is unlikely to be accepted as an exchange. And so, with the owner’s consent, the set was written off.

I now moved on to the second set which had no colour. I checked it out with a signal to each of its AV inputs to find there was no colour except with the DVD component inputs which don’t go near the set’s colour decoder. What I did realise was that there were actually bands of colour. This told me that the colour signal was there but the burst or reference oscillator was either not working or not locking.

This set does not have a conventional colour decoder as of old but rather a digital one with phase locked loop control of the crystals. Interestingly, the PIP (picture-in-picture) had full colour but lost it when swapped with the main. The digital decoder consists of a large microprocessor (IC7501) called HIP (for High-end Input Processor) which does the luminance and chroma processing.

Using my wet finger technique, I found that when I touched the area around the 4.43MHz crystal, I could lock in the colour. When the colour wasn’t locked, the picture tended to have a bit of jitter at the bottom but when locked it was perfect.

This told me that the frequency was out of range to lock to the burst signal, so I started by changing C2525, an 18pF capacitor in series with the PAL 4.43MHz reference oscillator crystal (X1525). I fitted a ceramic capacitor in place of the surface-mounted chip but this made no difference, so I directed my gaze to pin 52 of the processor which is the colour PLL.

This pin has C2522, a 3.3nF capacitor, connected to ground which is known to give a buzz in the sound. This is in parallel with resistor R3521 (100kΩ) and capacitor C2521 (100nF). Replacing all these components made no difference but I did notice a small black mark on the board near the IC. I checked the continuity of the copper track and found that is was open circuit between C2522 and pin 52. I couldn’t see where the break was but fitting a wire link between the surface-mounted components fixed the problem.

Panasonic DVD recorder

The 2004 Panasonic DMR-E55 GN is a DVD recorder and analog tuner. It will only record and playback DVDs and won’t record in any CD format – not even for music. It will do DVD-RAM up to 16 hours on a 9.4GB double-sided disc in EP mode and has "Time Slip" where you can record and playback at the same time.

This set doesn’t give much trouble except sometimes in the power supply. If the unit stays in self-check mode or shuts down, then you can suspect IC1 (STRG6353). The easiest way to confirm this is by measuring the 12V rail with a CRO (a DVM is too slow). If it is too low or varies, then IC1 is the prime candidate.

Following some recent storms, we have been doing a roaring trade in these regulator ICs. I was given one of these models after one of the opposition decided he had gone as far as he economically could. The set had been hit in a storm and IC1 plus a few components around it had blown up and had now been replaced but the set still kept cutting off after the self-check had completed.

The voltages out of the secondary of chopper transformer T001 were all correct at +13V, +9V, +6V and +7V. However, the other rails derived from these weren’t all kosher, especially the +12V rail which was low and varying, despite IC1 having been replaced.

Next, I took the board out of the set to get better access. In order to start it, I had to turn it on by shorting Q111’s collector to its emitter. This in turn switches on Q110 which applies +13.5V to the DC/DC circuit.

I also switched on Q114 by shorting transistor Q107’s collector to its emitter, which supplies +13.5V to a capacitor within the DC/DC circuit. That done, all the power rails came on correctly and gave me the chance to check all the voltages within the DC/DC circuit. Everything was pretty good except for the output (pin 1) of IC102 (LM2904DR) which was down from 4.3V to 2.5V despite both inputs (pins 2 & 3) being correct at 1.2V.

I thought that the surface-mounted IC might be faulty but fortunately when examining the component side minutely, I found that R115 (3.5kW) was cracked. It measured open circuit and a new one fixed the problem. However, what caused this surface-mounted resistor’s early demise? I thought it highly unlikely it was due to storm damage as it was so far away from the damaged areas. In the end, I concluded that it was a man-made error inflicted during the earlier repair of the switching IC (IC1).

After the repair I downloaded the relevant software, made a CD and installed a firmware upgrade to update the machine from 0900EK to 0920EK.

Panasonic TV

A common fault in Panasonic TVs has been dry joints to the vertical output IC (IC451). Quite often, these dry joints are invisible to the eye but re-sweating the joints with fresh solder nearly always fixes the fault.

I had an exception the other day, though. This was with a 1996 TC-25V50A employing an MX-2A chassis. This particular set was really past its use-by date and was somewhat corroded from being near the beach.

It came in with intermittent vertical scan, another common problem. Often, a good swipe on the side of the set with your hand will bring back the vertical deflection temporarily but this set did not seem so inclined. In fact, the scan was more directly proportional to the brightness of the otherwise very dull picture. This could have meant that the tube was pretty "flat" too.

Despite this contradictory evidence, I chose to take the problem on as I was sure it was only dry joints causing the problem. I still didn’t twig when I discovered that the original soldering was in fact excellent. I resoldered it anyway but to no avail.

OK, occasionally the LA7833 IC fails too, so in went a new one but there was still no joy. The vertical deflection was as intermittent as ever and no matter how much I bashed it, twisted it, froze it or heated it, it stubbornly remained that way.

But I wasn’t going to be beaten by this old monster and so I decided to look at the other symptom – ie, the lack of contrast. This had to be around the ABL (Automatic Brightness Limiter) circuit which comes off the tail-end of the EHT overwind on the flyback transformer, in this case pin 3. This is decoupled to ground via capacitor C512 (0.47μF 180V).

On old Philips sets, losing capacitance in this component would give low contrast but this one measured spot-on out of circuit. I replaced it all the same and then, based on past experience, looked around for a high-value resistor nearby that might have gone open circuit. R529 (330kΩ) looked a likely candidate but it too was spot on. I then found R525 which was a weird-looking carbon resistor colour coded brown, red, violet, orange, brown. This equates to 127kΩ 1% but it measured open circuit. Bingo!

I don’t stock 127kΩ resistors as they aren’t in huge demand, so I made one up using 100kΩ and 27kΩ resistors in series. This fixed the contrast problem immediately and demonstrated that the picture tube was still in excellent condition. As a bonus, it also fixed the intermittent vertical scan problem as it is indirectly coupled to the vertical output IC.

Anyway, I was lucky and it put me on my guard against rash generalisations when diagnosing faults.