Friday, 9 April 2010

Repair to a USB socket

A common fault with USB sockets on a laptop computer, is mechanical damage, due to knocking whatever is plugged into it.
There is a central divide in the socket which supports the gold plated contacts inside. This can break off, leaving the contacts suspended in the middle.
The next time something is plugged in (if not noticed) is that the contacts will be bent over.
This can of course lead to short circuits and/or permanent damage to the circuitry within.
If you are lucky the electronics will survive and it only remains to repair the mechanical damage.

First, (with the laptop powered down and battery removed) carefully bend the contacts straight, using tweezers, watchmakers screw driver and very small pliers.

Now make a new central divider.
Easier said than done.
I found an ideal medium. I used an old credit card. The plastic is just the right thickness.
Using a modelling knife, I cut a small rectangle.
Trial and error to get just the right size. It has to allow space at the edges, for the metal shroud of the plug to pass into the socket and sit at the base of the socket without protruding.

I then cut some small slivers of plastic to act as guides for the contacts. These are almost square in cross section, cut five of them.

Stick them onto the prepared divider using super glue.
I put a blob of super glue on my chopping board and then picked up a drip using a pin.
Starting with the outer edges, run a smear of glue on the sides of the divider and place a guide onto it using tweezers.

By eye I guessed the centre line and glued a guide there.
Then, again by eye, I guessed the mid point between the outer and centre guide and glued the last two guides.
You should now have a new center divide with four channels to protect the contacts.

Trial fit the new center divide and make sure that the contacts fit in the grooves and that the finished divide does not protrude beyond the socket, trim as neccessary using the knife.
Make sure it is the right way around, by comparing with an undamaged socket and allow a space on the back side of the divide so that the plug shroud will slide down behind it. You can use an off cut of credit card to get the space right.

Once happy with the fit, smear the bottom end of the divider with glue (I used a fast acting, clear, epoxy resin, 'Araldite' for this) and slide the divider into place. Use the off cut of card to get the space right behind it and put aside to set.

For the first trial use, I used an old memory stick and an LED light incase of damage to the plug.
Gently try the fit. It may be tight at first. Beware that the contacts may get bent again unless perfectly snug in the new divider.

If all goes well you should have your socket back working again.

Friday, 28 November 2008

Mazda MX5 Seat Repair

The driver's seat of Sal's baby needed some TLC.

The seam on the side of the leather seat had come apart and started to expose the foam underneath.
Also the integral loudspeakers in the headrest, sounded awful. Apparently on cloth seats there is a zip around the top to allow access to the speakers. But Sal's car is a japanese import with leather and there is no zip. The only way to access the top is to remove the whole cover.

So it was time to remove the seat into the comfort of the lounge and see what I could do.

Unbolting the seat was faily straight forward. Just four bolts at each end of the seat rails, moving the seat back and forth on the runners to gain access.

First the seat back had to be removed from the seat base. Several bolts around the sides, through the plastic trim had to be undone.

I had a problem with the handle for the recliner. The plastic handle would not come off and the plastic trim wouldn't pass over it. The trim was loose and I managed just to swing it out of the way to gain access to the bolts underneath but leave it attached by the handle.

The parted stitching was on the seat back, where it hugs your side, and you slide in and out of the car.

I removed the whole of the covering.

This required the metal rings (Hog rings), that secure the leather to the seat frame, to be prised open using a screw driver.
There are several attachment points. Once the front is released it exposes the attachments (more hog rings) which hold the middle of the fabric to the frame.

It was just a matter of working round until they were all released and trying to remember where they all went, for re-assembly.
The cover then slides off, like skinning a rabbit.

Using strong thread I re-stitched the parted seam, using an overhand blanket stitch, from the underside of the leather.
While the cover was off I turned my attention to the faulty speakers.

The cause of the bad sound was immediately evident.
The paper cone had become completley detached from the frame, this allowed the speech coil to scrape along the magnet as it was no longer located centrally.
I looked online to see if I could replace them. My usual supplier of MX5 spares
had some equivalent replacements but they were the wrong shape and required some cutting of the foam to make them fit.

The original speakers are oval and fit exactly into plastic housings. Originals were available but quite expensive.

I opted to try and repair them as they still worked electrically, all I needed to do was, to re-secure the cone to the frame and yet still allow the cone to move freely in and out. The original flexible scroll around the edge was the only bit that had perished.

I chose to use silicone sealant. This dries to a very tough, yet flexible, rubber.

I needed to apply a thin ribbon around the edge, thin enough to allow free movement but tough enough to secure the cone to the frame.
I used a small screw driver to spread it evenly.

It looked horrible, but no one was going to see it once covered. I cannot claim that it was Hi-Fi, but then again the orignal speakers weren't either.
Job done, I resecured the seat cover using the reclaimed hog rings, then bolted the seat back in the car and connected up the speaker wiring.

It looks and sounds a lot better.

Monday, 10 November 2008

Computer Crash

Fixed A Computer Crash

Asus A7V400-MX

Windows XP Home

My Brother in-law ‘Steve’ had a crashed computer he brought to me for fixing.

It would not start up. It would get as for a s the XP screen with the blue ‘scrolly’ then return to the Black and white options page.
It refused to start in Safe mode, or any of the other options.

At a guess, it would either be a virus or a hard drive crash.

I decided to remove the hard drive and inspect it fully.
I connected it to the USB on my computer using an IDE ATA caddy and fired it up.
It was very slow to react and I could hear the drive making some odd noises, making a group of four buzzes at regular intervals.
It eventually showed up on ‘my computer’ as an ‘F ‘ drive in this instance.
I could not open the contents of the drive and when I used a right click and selected ‘properties’, it showed up as empty, 0 bytes used. "Oh dear! Looks terminal", I thought.
I tried running my anti-virus (AVIRA AntVir premium) to check it but the program wouldn’t run and eventually crashed my computer.
Not giving up that easy. I decided to try some data recovery systems.
Being cheap, I wanted a free one, so Sal surfed the web looking for such things.
She came up with "PCInspector" a free download at this site
In German.
It downloaded and ran perfectly, with Teutonic efficiency, and found there was indeed data on the disk.
I managed to recover some data and stored it on my computer’s hard drive for later use.
It did have a struggle and found some sectors on the disk unreadable.
A bit more fiddling and playing around with the functions of the recovery program found some recoverable folders and saved these as well, on my host computer.
My computer eventually crashed when it got stuck trying to read the drive.
I re-booted my host computer leaving the USB caddy still connected. Then something amazing happened!…
As my computer re-booted and checked the hard drives, it found the errant ‘F’ drive and did a CHKDSK on it, finding all sorts of truncated files and lost links. It did it’s best to correct it.
When it had finished, I went to the ‘My Computer’ and looked at the properties of the ‘F’ drive. It now showed that it had Data and was almost completely full. WOW!
I was going to back up the whole drive (80Gb) using Nero ‘BackitUp’, but it required 9 DVD’s and was going to take ages. So I scrubbed that idea and went for a re-install in the original computer.
Plugged it back in and switched on.
It Worked! The original wallpaper and all the files came up with only a few errors. According to Steve these were errors that ‘normally’ happened. One was due to a wireless card that had no drivers.
Found the drivers on the net and transferred it to his comp using a CD-RW. Got it on-line using our wireless network and did a few updates and cleaning up. All seemed OK.

I reckoned that the hard drive was and is still, faulty and not to be trusted. It was interesting that the windows system picked up on the drive while it was plugged into my computer. Possibly only because it recognised that it was a bootable drive, with windows XP on it.
I installed a spare 10gig slave drive and upgraded the memory from my spares box, so that vital files could be moved across. It really needs an new main hard drive.
A future project.........................

Friday, 7 November 2008

LCD Monitor Repair

15" LCD Monitor repair

The power light came on but no Brightness at all on the screen.
If you looked closely when the computer was on, you could just make out a dull image. But obviously no good to use like this.

LCD monitors work differently to the old, glass, cathode ray tubes.
CRT’s emit light from phosphor, just inside the front of the glass tube.
Liquid Crystal Displays have no light of their own and rely on miniature fluorescent tubes arranged inside, behind the flat panel, In the same way that LCD watches have a back light to see them at night.
It appeared that this monitor had lost it’s light source.

This monitor came apart by undoing three screws at the back and prising the bezel off from around the screen.
There were no obvious signs of burning on the circuit boards.
The miniature, cold cathode, fluorescent tubes need a high voltage to fire them up (about 1,500v). This is made from the 12 volt supply by a section on the board known as an inverter.
I could see the screened (pink) wires that go from the board and disappear round the edge of the screen. Following them back to the plugs on the board, I tested here for signs of a voltage.
If voltage was there, then it is likely that the tubes have both blown. If there was no voltage, then the tubes are likely to be OK and the inverter is faulty.
No Voltage! Dead! Not a sausage!. So the inverter is faulty.

I Googled the part number on the inverter/power supply board, to see if I could get a new one.
I got a couple of hits in the USA, but when I searched further, the pictures they showed, looked totally different shapes to the board I had. I couldn’t be sure that it would fit, despite doing the same job. Space is a bit tight behind the flat panel. I couldn’t find any circuit diagrams on the web.
The monitor is some obscure make from China and no one would own up to creating it. So I couldn’t attempt to repair the board. The components are surface mounted and a pig to change anyway, without causing collateral damage to the rest of it.
I found a company that actually makes inverter boards to order, and have several generic ones that will power almost any LCD. I just had to identify how many tubes I have and what size screen.
As there were 2 wires, I guess it only had 2 tubes and it was 15inch size.
I ordered a generic double LCD inverter from
Who are part of
It cost £14.90, but with post & packing and tax, came to £29.14.
A new cheap monitor would cost £50, but it would be good fun to try and reclaim it and keep it from the rubbish heap for another few years.
Picture 3. The Old board.

No wiring diagrams came with the board, but I got the general idea from their web site. The new board requires 12volts and has a ready made socket on the board to take a moulded plug. The outputs to the tubes fitted the plugs already on the wires, so no problem there. As I didn’t have a plug for the power supply I just soldered the wires on the pins of the socket to the main board, where I found 12volts coming in from the separate power supply. Pin 1 on the LCD board is 12v+ and pins 4&5 is the ground.
As a precaution, I removed some of the components from the board of the old inverter. Just in case it caused excessive drain or decided at a later date to fire up.
There was no where convenient to screw the new board down behind the screen, so I used self adhesive ‘Velcro’. Mostly because it would insulate it from the metal of the screen and also because I hadn’t got any low profile ‘stand offs’ in my spares box.
Wired up and switched on………………….NOTHING.
Hummmm…what’s wrong. Checked the output from the new board. Dead. 12 volts going in OK.
Checked back at the site and discovered that there is an On/Off pin on the board and it needs 5volts before it will fire up.
There is also a pin for changing the brightness, but as this requires detailed knowledge of the original circuit it can be left off and the tubes will remain at full brightness.

Probed around the board to find 5volts. Found one, but it was on all the time that the power supply is connected. I connected it to pin 3 on the inverter and … IT WORKS. I got the tubes to light and the screen showed the message that there was no input from the computer.

Picture 4. New board on the left
with wires attached.
Not Satisfied
For it to be really useful, I wanted the tubes to power down with the on/off button on the front panel and also to go into standby properly and save power by turning the tubes off.
I had to find a voltage that comes on initially and then goes off when no input is detected.
I probed around with my meter all over the old inverter board but could find no signs of life. Could this be the original fault?
I eventually found, on a plug that links the old inverter board to the video board, a line that had 3.5volts on switch on, returning to 0 after going into standby.
I tried that, and IT WORKS! Brilliant.

Re assembly
After putting the casing all back together and switching on..NOTHING.
The light on the separate power supply was out. Removed the power plug from the monitor and it came back on. Uh Ohh! A short.
Disassembled the screen again and checked all the wiring. All OK.
Connected it to my laptop and switched on with it apart. Working again!
Put it back together with it working.
I was just clipping the bezel back on when I heard fizzling and a funny smell…. OOOpps!
Pulled the plug and disassembled again.
This monitor actually has little speakers in it and one of them was touching the high voltage output from the inverter, as shown by a bit of blackening on one of the pins.
Luckily the inverter board is ‘Velcro’d’ in place. I detached it and moved it further up and away from the loudspeaker attached to the back shell, adding extra insulation to make sure the underside of the inverter is away from the metal back plane.
Reassembled and thoroughly tested in all modes. Brilliant! Job done.

Thursday, 6 November 2008

Engine Overhaul

Repairing Daniel's Vauxhall Astra

A 1993 with C14SE engine

2nd to 4th June 2008

Rough running, very noisy valve gear. Low compression on No 2. Emulsified oil in valve cover.

Top end overhaul and inspection of valve gear.

First day,
strip down.
Removed O/S inner sill cover, 3 torx screws, revealed fuel pump relay inside 'A' pillar base and removed it. Turned engine over, fired then stopped. Fuel system now de-pressurized. Disconnected battery.
Removed air filter housing. Removed auxillary belt and tucked it out the way. (captive by engine mounting). Removed timing belt top cover (3 x 10mm bolts). Jacked up front O/S and removed wheel to gain access to lower side of engine. Removed wheel well access cover, (2 x self tapping bolts and plastic rivet). Turned engine to No1 TDC, to align timing marks, (previously high lit with white paint, suggesting previous work)
Attempted to remove crank pulley bolt by putting in gear and with wheel on the road, no go.
Had to weld adjustment to my previously made chain wrench, wrap the crank pulley with cycle inner tube and used that successfully to remove bolt. (18mm ring spanner)
Removed lower belt cover, 3 x 10mm bolts.
Inspection of timing belt, OK. The adjuster indicator, showed that it was slack. Removed timing belt, holding the adjuster back with a drill in the locating hole. Tried to undo the back timing belt cover. (4 x Torx30 head screws) 1 very tight and broke one of my torx set, left it. Tried to loosen the water pump securing screws (3 x 5mm allen). 1 very tight and tool in risk of chewing it, left it.
Drained coolant by siphon from expansion chamber.
Removed throttle cable, injector electrics and various pipes and hoses from the inlet manifold.
Removed Cam sprocket. Had to make up a tool to hold the cam sprocket still.
Tried to remove exhaust down pipe from the manifold. One bolt and spring came off OK but the other had a chewed head and couldn't get a spanner to fit. Ended up angle grinding the head off.
Removed the cylinder head bolts in reverse tightening sequence. (standard hex head bolts).
Lifted the cam carrier off, pushing the plastic belt cover aside to clear the sprocket end.
Removed the rockers, slippers and hydraulic tappets, keeping them in order in a sectioned box.
The damage to No 2 cylinder valve gear was immediately apparent. The cam lobes were horribly re-profiled and the rockers worn badly. Most of the other cam lobes were OK but showed deep scratches and wear. (See picture at the top)

Removed the cylinder head complete with manifolds.
Dismantled manifolds from head on the bench. Most of the exhaust studs came out with the nuts.
Tried to remove the bolt remains from the exhaust manifold. It was only moving a bit but showed signs of cross threading. Used a stilson wrench and broke it off leaving the remains in the hole. Gave up for the day.

Second day,

Clean up and rebuild.

Sourced a secondhand cam complete with housing, rockers and slippers, just at the end of the road at Sadds yard.
Cleaned all the old gasket from the cylinder head and engine block. Looked in very good condition, very little carbon deposits.
Siphoned more coolant from the block. Took the Exhaust manifold to Towler engineering, Oxford road, for them to drill out and tap the hole. Collected the new cylinder head bolts from CMD and Exhaust studs, bolts, springs and sealant from Vauxhall main agents.
When I went to collect the manifold from the engineers they pointed out that it was also cracked and offered to weld it up. I did consider scrapping it and getting another from the breakers, But decided that as they had already done a good job on the bolt hole to continue with it.
Inspected the old tappets and stripped the ones down from No2. One wouldn't come apart, so tried one of the others. The stuck one eventually came apart and I compared it with the good one. Showed signs of wear scratches on the piston. I decided that this was the probable cause of failure, wearing the cam lobe when the adjuster failed. Went to the breakers again and picked up 2 tappets to replace the suspect ones.
Removed the valves from No2 and lapped them in after inspection. The valves were in near perfect condition and needed very little lapping using fine paste. After doing the valves for No3 decided that I was probably doing more harm than good, as they were in such good condition and previous compression tests proved the other cylinders were fine.
Re-assembled the head with new gasket on the block. Tried a dry test fit with the new bolts and found that the bolt heads were different. I needed a star socket tool to fit them. CMD didn't have one but determined that it was an E12 socket. Sourced one at Halfords.
Picked up the manifold from Towlers. They had skimmed the mating surface as well, nice job.
Assembled the tappets, slippers and rockers on top of the valves. Found that one of the rockers supplied with the camshaft was odd. Same dimensions but a lighter construction. The new rockers were the same as the old, but had a hole drilled in each over the tappet socket. I chose the best of the old ones and put it in the No1 position. Its now the only one without an oil hole.
Applied green jointing compound to the top of the cylinder head and lowered the cam unit on top after checking the timing marks. Torqued down the head in sequence and the four stages, as per the book. An initial torque setting and then three 60ยบ angle turns in sequence, done by eye.
Assembled the inlet manifold on the head and found that it is very awkward to get the nuts on the lower side. Various pipes and a cooling water gantry get in the way and its impossible to see. It would have been better, with hindsight, to attach it to the head before mounting.
Gave up for the evening.

Third day,
Final assembly and start

Struggled with the inlet manifold and finally got it fastened. Re-attached all the various vacuum pipes and electrics to the manifold and plugged in the injectors. Hard to find home for two of the vacuum pipes. One went to a circuit board on the bulkhead and the other under the air mass meter.
Went to attach the exhaust manifold, after inserting new studs, where they were missing and found that it was distorted and didn't line up with the studs. The welding must have closed up the crack and pulled the sides in. Again, I wished that I had scrapped it.
Decided to re-drill the holes to elongate them using the gasket as a template. Re-fitted the manifold with new copper nuts and anti-seize compound.
Re-attached the down pipe with the new bolts and springs, using anti-seize on the bolts and fire paste in the pipe socket.
Inserted the spark plugs and wiring.
Using a new torx30 socket I managed to release the final screw holding the cam belt rear cover and gain access to the water pump. I managed to release the final hex socket bolt on the pump, so they were all loose. I didn't really want to release the water pump to adjust the cam belt unless absolutely necessary, as it was sealing OK and looks recently replaced. Re-fitted the bolts with anti-seize paste. I tried a dry fit of the cam belt as see if it was at the right tension. The marker on the idler showed that it was. Put the crank pulley back on using my chain wrench. Then discovered that I couldn't get the back cover on, so had to take it back off again. Re-attached the rear cover and then re-assembled the cam belt, holding the tensioner back using a drill in the holes in the pointer. The crank pulley was tightened using an 18mm ring spanner as tight as I could. The cam sprocket was tightened to the required torque using the tool I previously made to hold the sprocket still. Turned the engine over twice using the spanner on the crank bolt and re-checked the tension. Perfect.
Re-assembled the final parts, alternator, air cleaner and crank speed sensor wiring. Re-filled the water and a final check everything OK. Replaced the fuel pump relay and put the covers back. Re-connected the battery negative terminal. Fingers crossed.
Turned the starter. After about 2 revolutions the engine fired and then stopped. Second turn it ran for a couple of seconds. Third attempt it ran and ticked over. A slight rattle from the top until the adjusters pumped up and then sweet.
Job done.
Road test.
All OK, but the engine management warning light comes on for a few minutes and then off.

Installing a new Hard drive

How to install a new hard drive.
Sony Vaio with Windows XP Media edition and Sony back up.


To upgrade the hard drive from a 100GB to 250GB without loosing any data, keeping the operating system and back up system intact.

Perceived problems.

Due to it being a laptop project, there is no way to connect two hard drives in parallel. The hard drive is also SATA, so no ribbon cables.
The cloning of the drive would have to be via USB.
Need cloning software (preferably free).

Things to buy before starting.

Toshiba 250Gig SATA drive from
SATA docking station with USB from

Scan the Web

To find free software for cloning.
As part of the search, found free back up software in the form of DriveImage XML from ‘Runtime’ and boot disc software BartPE, from Bart Lagerweij.
Found the ideal program in ‘PC Disk Clone’ from ‘PC Disk Tools’
in Australia. They do several versions for sale and a free one. The free one is fully functional but SLOW.
Downloaded the free version and ran it on the Sony.

The program creates a boot disk in order to copy the hard drive. A nice logo of a sheep. No doubt reference to ‘Dolly’ the cloned sheep.
Follow the on screen instructions and put a blank CD in the drive. After a few seconds after clicking on the ‘Burn a CD’ and the disk logo button, a disk was created.


Set up the computer where it can’t be disturbed for a few hours.
Set up the caddy with the brand new, unformatted SATA drive in it, plugged in, switched on and connected the USB.
Put the disk just created, into the drive and re-start the computer. No need to set the BIOS to pick up the CD ROM drive as it is already set to look there first, if there is a bootable disk in the drive.
When it sets up it looks for the external drive. It said that, if it can’t find the USB just press any key.
This I did. I did try unplugging and reconnecting the USB but no visible acknowledgement on the screen, so I pressed any key.
Not to worry, the next screen showed the existence of the new drive on the USB.
You select the ‘source’ drive from the table, which was at the top of the list.
It showed the partitions on that drive including the hidden one where Sony had stored the recovery data.
Go to ‘next’, then select the Destination drive.
The drive was shown as ‘removable’. Then click next.
The cloning then started and showed 0% for several minutes.
It was SLOW. It took 10 hours!, to copy 100Gb to the new drive. But who cares, it was free and I don’t expect to do this on a regular basis.

Replace the drive.

Once finished and the computer was shut down, it was time to swap the drives.
On the VAIO the hard drive is accessed from under the computer. Remove 2 screws and then the drive slides out from the front. It has to be separated from the mounting plate by 4 screws.
Re-attach the new drive to the mounting plate plug it back in and replace the 2 screws diagonally opposite on the rectangular location.
Switch on. Hey Presto! The computer fired up as if nothing had happened to it. All the desk top and programs exactly in place.
So perfect was the copy that we had exactly the same capacity as before. What has happened to the extra size on the drive?!…..


Using Windows, I went to the Computer management. Found in ‘Control Panel’ then ‘Administrative Tools’ then ‘Computer Management’. Select the ‘Disk Management’ under the ‘Storage’ tree. And there it is!.
The Partitions are shown in a table, all as it should be. The hidden drive, the ‘C’ drive, the ‘D’ drive and there at the end, in black, ‘UNALLOCATED’ all 137Gb of it!
How to use it?

A bit of research on the net, all refers to using proprietary software such as ‘Partition Magic’, all at a price. Ok, if you are doing this as a living, but as a one off I want it for free!
I found a program called ‘Cute Partition Manager’ at
I downloaded it and ran it. It makes a bootable disk so that it can handle messing with the hard drive without falling over itself.
It failed to make the boot disk properly so I selected it to make an ‘*.ISO" file, which it put in my root drive under CPM. I then burnt this to a CD using ‘Nero’.
The disk booted OK and gave a table of the partitions. I couldn’t work out how to expand the ‘C’ drive to make use of the unallocated area, at least not without loosing all the data it had taken me 10hours to put on there!
More research on their site on the web revealed that it couldn’t do it. It couldn’t do much more than the windows system would do anyway and to Re-partition would loose the data.

Back to the drawing board.

Research by Googling ‘Free partition software’ found a company called ‘Easeus’ at or
Navigate to the FREE Partition Manager, Home Edition.
It downloaded and installed itself on the computer.
When you run it, it shows the partitions and you can simply drag with the mouse, using grab handles on the graphics. Brilliant!
I had to do it in several stages as the unallocated partition was at the end, after the ‘D’ drive.
I wanted a big ‘C’ drive and not so big ‘D’ drive.
So I first made the ‘D’ Huge, dragging the end of the partition right to the end of the line, using up all the unallocated space.
Then I made it small again by dragging the beginning away from the end of the ‘C’ drive.
I then expanded the ‘C’ drive to take up the slack. Almost as easy as saying it!
Then click on the ‘apply’ button. Then computer then shuts down and reboots.
Very clever that it doesn’t need an external drive like a floppy or CD. (It must retain the program in memory and run from there).
Anyway it re-booted and shows a display on how it’s getting on. It did it in three stages and took little more than 10 minutes.
It then re-booted and everything was OK. After a second re-boot to acknowledge what it saw as a new hard drive, a look in ‘My Computer’ shows a nice big ‘C’ drive and ‘D’ drive.
All is well.