It was a beautiful spring morning and I was sitting outside the
workshop enjoying a cup of coffee and a chat with a friend. Suddenly, someone
shouted "FIRE!" and all hell broke loose. It was my neighbour’s factory unit and
it had smoke pouring out of the roof and from the gaps around the closed doors
I didn’t have my neighbour’s mobile number but someone else did
and called him. Naturally, the fire brigade was also called and while we waited,
a few of the other unit owners managed to organise some water hoses.
The problem was that the place was fully locked with the roller
door in the closed position, so all anybody could do was to cool the building
from the outside with the water hoses. It didn’t take long for the fire brigade
to arrive and they were about to break down the door with a battering ram when
the owner screeched to a halt outside. He got out and unlocked the door for them
and was immediately enveloped in thick acrid smoke.
The fire brigade took over from there and several men wearing
breathing apparatus entered the building dragging fire hoses. They quickly
extinguished the fire which was on the mezzanine floor and within about 10
minutes, the roller door was pulled up and the windows opened.
None of us in the now sizeable audience outside actually saw
any flames but it was fairly obvious that the fire was now out and it was just a
matter of clearing up and blowing out the smoke. After the firemen had finished,
the forensic police went in and then later on the insurance assessor
It wasn’t until four hours later that I was allowed in to see
the damage. Downstairs it looked as though nothing had happened but upstairs was
like Armageddon. Though no-one could actually swear to it, it looked as though a
mains plugpack or an extension lead had been the cause but there just wasn’t
enough evidence left to confirm this.
Immediately adjacent to the seat of the fire were the remains
of an elaborate computer system. I was asked to give an insurance estimate for
some of the items that were damaged and was also asked to see if I could recover
any of the data on some of the computers.
The first item was a 2Gb USB2 fast external hard drive in an
aluminium case. This had copped a lot of heat and the plastic had melted and
burnt on both ends of the metal case and had to be cut away. Eventually, I
extracted the IBM IDE hard drive and most of the controller card.
The USB socket was too badly fried but I managed to get a new
one and fit that. I also had to replace the red HDD activity LED, which had
After some cleaning, the controller card came up looking pretty
good. I then connected it to one of my computers with one of my hard drives and
was delighted to see that it worked OK.
Next, I turned my attention to the hard disk drive inside the
case. This was still wet and was covered with all sorts of powdered debris,
despite being deep inside the machine. I took it out and cleaned it up using
brushes and an air compressor. It came up looking quite good on the outside so I
connected it to the USB controller card and plugged it into my computer.
All I got were a few faint noises from the drive but no action.
I then decided to remove the drive’s PC board which was held in place by six
Torx screws. Once it was out, I was annoyed to see even more wet ash on the
inside. How on earth did this get there?
After delicately brushing out the debris and flushing it with
compressed air, I sprayed the corroded areas with a light penetrating oil and
then blew the excess off. I then reassembled the hard drive and reconnected
This time the noise was somewhat louder but still nowhere near
that of a working hard drive. I wasn’t game to remove the main cover on the hard
drive itself as I figured this was airtight. My feeling was that the motor
wasn’t spinning correctly and nothing more could be done by me. An expert would
have to do the rest.
The desktop computer was a Dell and had really copped a lot of
Inside, the motherboard didn’t look too bad. The DVD and floppy
drives had been destroyed but the hard drive looked OK. Connected to another
computer, it immediately spun up OK and the data was readable.
Unfortunately, the 20-inch Mitsubishi monitor had been
totalled, along with the keyboard and mouse , and was simply chucked into the
There were also a couple of notebooks in the room, one of which
was a write off. The other, a HP Pavilion dv9000, was borderline. It had been
saturated with fire retardant but because the lid was down, it looked as though
only the sides needed cleaning.
After a brief clean, it actually booted up and seemed to be
working fine. However, because it had been saturated with fire retardant, I was
asked to strip it down and clean it out.
That wasn’t as easy as it sounds. First, nobody ever tells you
about the lingering sweet, sickly smell that permeates everything touched by
smoke or fire. Second, I hate doing notebook computers at the best of times as
there are always concealed clips and screws, plus there are millions of
different screws which all look very similar but can only go back into their
Anyway, I downloaded the maintenance and service guide (some
300 pages) and printed out the relevant removal and replacement procedures.
Because the notebook was working, I decided to strip it down
just enough so that I could see inside it and see whether it really was damaged
or not. The really difficult part involves removing the keyboard and later, if
necessary, the display screen, system board and power supply.
To get to the keyboard, you first have to remove all the drives
and accessories and then remove the switch cover on the top at the rear. This
assembly is held on by six screws in the battery compartment and then by a lot
Once the cover was off, I could access the keyboard mounting
screws and remove it. I could then see about 50% of the computer and it didn’t
look too bad. However, the USB socket and other sockets on the side of the case
were already beginning to corrode.
What is it with this fire retardant foam that the fire brigade
use? It is unbelievably corrosive and conductive too. It may put out the fire
but everything else that it touches is ruined.
Once again, I used brushes and an air-gun to clean out the gunk
and afterwards applied a minute film of oil over the contacts and metal
brackets. I then reassembled the computer and replaced the battery. And that’s
where my worst nightmare started. As soon as the battery was replaced, smoke
started to pour out the lefthand side of the computer near the VGA socket!
Unfortunately, the battery fitted to this notebook is not one
you can unplug and remove easily, as there is no finger grip available. Instead,
you have to release the battery lock, hold the computer horizontally and
literally let the battery drop out.
Once the new fire had subsided, I examined the relevant area
very carefully but just could not find the source of the smoke. I then gingerly
connected the external DC power supply and noticed smoke begin to come from
under the VGA socket before I unplugged it.
At this stage, I really had no idea what was causing this
problem except that it had to be in the power supply on the system board. So
there was nothing for it but to disassemble the whole thing.
Referring to the guide, there were four major assemblies to
remove. In addition, I had to unplug lots of extra module accessories. I started
at section 5.4 and worked through to section 5.25. Each time I took out a part,
I placed it in a box together with a bag holding the appropriate screws.
As I peeled off the layers and got deeper and deeper into the
works, I began to see all the areas I had missed when cleaning, especially
around the edges. Eventually, I got the system board out and removed the
heatsinks and then the insulating tape covering the power supply next to the VGA
Well, I looked and looked but I couldn’t see any sign of a hot
spot or fire. In fact, I had to power the system board up again to finally
locate the source of the trouble. It was a tiny dag on the VGA support bracket
that was just touching a PC board track.
Access was very difficult but I clean-ed the area up
meticulously and angle ground the metal dags off the bracket so it didn’t touch.
This stopped the pyrotechnics in their tracks – but had the computer’s
electronics been damaged? I wouldn’t be able to tell until after another two to
three hours of re-assembly.
Next I paid attention to the other areas of corrosion. The
audio board was badly corroded in and around its connection plug and one of the
tracks had actually been "etched" open circuit by the foam. I bridged this gap
and cleaned it thoroughly.
Similarly, the power input socket and some of the extra sockets
and assemblies all around the edge of the case required attention. That done, I
then started the tedious task of re-assembly.
However, despite all the precautions of saving the screws with
their assemblies, I found there were discrepancies between the service manual
and reality. Also many of the screws looked very similar but weren’t. Then I had
the hassle of reconnecting all the miniature plugs and sockets. Some of the
ribbon cables were very fiddly to plug in.
Finally, when it was all back together, I had one screw and one
socket left. The only trouble was the screw didn’t match the socket! Anyway, the
notebook was solid – nothing rattled – it would just have to work without
And fortunately it did – all the functions seemed to test
It was only now that I could prepare an estimate of the cost. I
also realised that I couldn’t possibly guarantee that the chemical reaction from
the foam would stop right there and then. In fact, it was quite likely that this
computer, now working, would be unreliable, with repetitive failures due to
In the end, I decided that the only way out would be to replace
both the system and audio boards. However, this would be more expensive than a
replacement notebook. All that hard work for nothing!
LCD TVs & monitors
Repair work on LCD TVs and monitors is now increasing, with
most repairs involving the backlight inverter power supplies. These power
supplies provide the backlights with high voltage and also crudely control their
brightness levels. As yet, I haven’t actually had any backlights fail unless
they have been smashed due to some sort of accident.
Recently, I had a Grundig Vision II LXW 68-9620 Dolby using a
G1 chassis come in with "no-picture" symptoms. However, if you shone a torch at
the right angle to the screen, you could just discern an image.
This set has quite a large inverter board as it had a lot of
backlights. Unfortunately though, the service manual shows no part of this
circuit board as it is part and parcel of the display panel. In other words, if
you can’t fix this board, you will have to replace the panel, which is
Fortunately, I could not detect +24V on the BUI-27 connector to
the inverters. This was due to a dry joint on one side of the diode that
supplies this rail on the main board and resoldering this fixed the problem.
I had similar symptoms on an HP f1703 (P9620A, probably made by
Liteon) LCD computer monitor, except that the picture was intermittent. It would
come on for 15-30 seconds and then go off.
This monitor used an external 12V power supply which was OK.
When I stripped it down, there were only two boards inside for the inverters and
the video processing and scalar drivers.
I immediately went for the inverter board, looking for dry
joints on the coils, but could find nothing. I then searched the internet for a
circuit diagram or service manual but again found nothing. However, I did find a
lot of hits for this very problem from other owners.
Next, I tried to contact HP but only got through to an Indian
call centre. They knew nothing about this problem and could only suggest I check
my screen saver. They did however "cherish" the opportunity to help and wished
me a "blessed" day!
Almost all the hardware fixes on the net involved completely
resoldering the four copper coils, which I did by first removing them from their
silicone rubber bed and then resoldering them back in position. This didn’t work
and without access to another f1703 monitor to compare voltages, I can only
conclude that it is probably something to do with the processor board switching
the inverters off.
I don’t have a circuit either and so have been left in a rather
frustrating cleft stick – at least for the time being.
Similarly, Apple has had problems with their Studio Display
17-inch monitors (M7649). The symptoms with these are uneven dull pictures and
the power light flashing.
Once again it is the inverter board that is the problem but the
modified replacement (V041063.00M3S by MoniServ) costs US$110 plus freight.
Why can’t these companies offer free circuit diagrams with
detailed information on how to repair or modify these products once they are out
of warranty? If an owner doesn’t have the technical ability to do the repairs,
at least he can supply the information to someone who is competent to do the
work. Wishing people a blessed day doesn’t quite cut it.
Grundig TV set
I was asked to do a service call on a Grundig ST84-796-9 TOP
LOG (CUC 6380 chassis) TV set. I don’t normally do service calls but I made an
exception in this case. This large 80cm CRT TV was located in a home unit and
was simply too heavy to lift and take to the workshop.
I was asked to phone before I arrived at the block of units
because of their secure internal parking arrangements. When I arrived, it wasn’t
hard to see why – the location was right in the middle of a large shopping
centre with no on-street parking.
Once inside the unit, I was surprised at how quiet and secluded
it was despite its location. I was also filled with dismay when I realised the
set was situated in one of those much-dreaded entertainment cabinets, with very
The problem was no sound which is an unusual fault these days.
Fortunately, I could see the cause as soon as I got close to the TV.
What happened was that the mains power on/off switch had been
giving trouble and wouldn’t latch in the on position. So, to keep the set on,
the owner had jammed a match splinter in the side of the on/off button.
Unfortunately, he didn’t realise that this switch also has a
momentary switch built in. This normally re-
sets the microprocessor and
mutes the sound to prevent unnecessary noises when switching the set on or
I proved the point by removing the match and pushing the switch
in far enough for the set to come on then releasing it slightly so that the
momentary switch was no longer on. This brought the sound on with the
I told the client that I would have to order the switch in from
Grundig and it would take time. In the event, the switch (part no. 297032917200)
was no longer available but as luck would have it, I did have a 297032917204
switch in stock.
Back at the customer’s unit, I managed to rotate the set in the
entertainment cabinet and remove the back and the main chassis. I then had to
remove the control and display chassis to gain access to the switch. Finally, I
removed the whole assembly and examined it carefully.
It was indeed a very special switch with a solenoid built in.
It had three functions – On, Standby and Full Off – which could be controlled by
the remote. You only had to push the switch in to go into either the full On or
Standby modes which are controlled by the remote. If you pressed the remote’s
Off button once, it would go to Standby. If you pressed it twice, it would pop
the switch out to Off using the solenoid.
The original switch looked identical to the one I had (thank
goodness) but it also looked as though there was nothing wrong with it as it was
now latching properly every time outside the cabinet.
What was crook was the power on/off button itself which had
cracked and so was not sitting on the switch properly. Once again I was left
with a dilemma. The faulty knob would probably work perfectly well super-glued
to its original switch. However I could only go down that path once. The knob
was no longer available and once superglued on, you would never be able to
remove it if the switch failed.
In the end, I decided to replace the switch just to make sure
and super-glued the knob onto it. This fully restored the sound and it was a
win-win situation all around.