Monday, July 26, 2010

Acer 5516 - Dead, No Power, No Lights

Finally a breather before the storm came. (Load of work)

This jack was damaged by drop and or push. The lead broke/shattered as if it was hit while running. Copper can shatter too, you know.

This should be clear enough LarryS.

Saturday, July 24, 2010

Acer 5515 Works off Battery, Charges, but will not run off AC

This unit arrived working from a charged battery, but will not power on from AC. 
My first thought was bad leg on the DC Jack, Easy job! Finally

Upon opening the case and removing the keyboard I saw the beginning of the failure. Burnt MOSFET under the keyboard directly behind the battery connector.

So took the rest of the panels off to see how bad it was. Out of a scale of 1-10 with 10 being the most difficult I rate this as a 7.5 because it was more annoying than difficult. The copper under the MOSFET was actually welded to the legs of the FET, heating solder would not break the weld.

Click to see larger image
The visual shows almost the worse case scenario for a shorted MOSFET. Burnt and totally unreadable numbers.  Seeing the circuit I surmised that it was an identical pair of chips but I had to verify against another motherboard to be sure.  Also there was a scorch mark under the keyboard, but it didn't affect the keyboard or any keys.

Removing part and making a copper bridge for the legs of the replacement MOSFET was a little challenge getting it right and isolating it
from the center plane of the motherboard if copper was exposed. In this case a small piece of mylar tape will be placed under the copper retrofit.

Sizing and fitting the copper 'jig'. This was made from thin copper foil and cut with exacto knife and placement test made to see if the Gate lines were shorting or any other pads before making it permanent and not being able to remove it. 

Replacing the MOSFETs and soldering broken traces. In this case the  GATE line was blown away on both FETs so I got 2 small segments of wire and made the necessary bridges to the GATEs.

Soldering the Gate line and cleaning up.  Now if you wanted to be fancy you can cover the solder and copper with some green fingernail polish, or clear polish and use a green marker over the polish after it dries. Then the repair will appear less noticeable.  

Notice the line at marking PR152 which was not so clear as being broken because it goes under the chip.

After doing all this I tested with the battery in place and with AC before closing up the laptop. No worse feeling than to have to redo after assembling things thinking you are done. ARRRGH It happened to me and so I started testing before closing.

As always you can click an image to see a larger version of it, and use the BACK button/Arrow to return to the page. Conversely- Shift and Click works best and automatically loads image in new browser.

Thursday, July 15, 2010

Compaq 6735b - Dead No Lights

This was a strange cookie. 
This unit showed no sign of life and was a weird design for power in button. 

I have to admit this was a 'lucky' repair. While diagnosing where was the DC power, I discovered that a power MOSFET was open, No Power anywhere, no lights, no charge and stone cold dead.

On the bottom of the board it was a mosfet FDS6676 that was open.  When I replaced it with a substitute I was able to follow the DC circuit, but it dropped off at the power button. So testing that area I noticed that when I put the meter positive on point on the board the light behind the pwr button would flash briefly.

Recalling a similar experience where I had to modify another CPQ 6735 when they first came out/released. I thought I would try my luck again. 

1: With the meter set to VOLTAGE and holding the meter's positive lead on this point, the unit would come on. My first guess was 47k Ohm. This worked but the laptop would not turn off, or would come on when it wanted to.
2: I would need another meter to see the resistance the meter itself presented. This turned out to be near 100k Ohm. Getting another 100k Ohm chip resistor from a dead unit, then making the leads, and adding clear heatshrink to insulate it from shorting. 
3: Always test thoroughly when out of case. This way you won't have to redo, or find out something else don't work.

First picture show resistor with background for clarity. The leads were added because there was no 'safe place' to tack to board. The white connector in picture is for the power button ribbon connector.

Click pictures for larger view

This shows placement where the part will lay. Heat shrink will be slid over resistor and hot glued down.

Hot glue holds part and wires to board. Makes it easy to remove additional work is needed later. This area is covered with a L-shaped piece of black plastic insulation sheet 
Final and ready to test again and then close.

Personal Note:
I always wondered why my high school Math/Algebra/Calculus teachers said 'Show Your Work'.  I was one of those students that seen solutions in my head, and couldn't always say, tell why or how I got I got the answers. They were just visible in the front of my mind.  Besides how can you explain something when you don't know the terms/rules and conditions of an equation, before you are taught 'the normal way'?

I quit that calculus class because I couldn't show the work, I didn't flunk I just quit it and picked another advanced science.  Their loss not mine.

Sunday, July 11, 2010

DV6000/DV9000 Lines in LCD and External monitor -unreadable

A fairly common problem with this series is lines in the LCD and external monitor. Sometimes it will exhibit scrambled characters on the screen. Similar issue.  This is because the video memory has fail or lost connection and the addresses cannot be read properly by the system. The rest of the system may work but video is not proper and screen unreadable.

Lines and streaks on POST screen
Should show BIOS info and CPU info

This shows the step by step on how to correct this problem. There are other videos on YouTube dealing with the 'no-video' issue. Some are dangerous, others are goofy. I will be getting back to that production.  Not wanting to knock anyone for trying, but be careful, some of these guys are not techs or engineers and have no clue of consequences. 

Thursday, July 8, 2010

Laptop Rebooting- DV6000, F700

While working on a DV6000 today I decided I should also address other problems with these models while working on the DVD production. 

As it seems this problem is also quite common, Rebooting, looping, restarting.  Never getting to the Windows startup screen.

OK there are several causes of this scenario with the ribbon cable being shorted or defective.  The ends of the ribbon cable can lift away from the surface when inserted and shorting the lead from the adjacent wire.  Some times it will short and cause the wire to burn. Other times it will cause the REBOOT/LOOP issue.  

Also verify that the Coin Sized CMOS battery has 2.5v- 3v, if not or voltage is low, all kinds of problems may present themselves.  Correct operating voltage is needed to read the CMOS data, or the data may be corrupt or deleted.

How to test/bypass these?
1: Disconnect the PWR connector cable, and turn the system on with the remote control (if you have one).  Make sure the battery is good in the remote.* 
2: Substitute with another cable.

Other issues presenting same problem
CMOS is bad** - If the CMOS is bad it cannot be read or written to- and the system continues to try to write/read data from it.

The Dual Channel MOSFET - one output is shorted or open. (or both)
When the MOSFET has about 3.25v on pins 1 & 3 it should be considered working. There should also be the same voltage on the Drain side (pins 5-8). This will take the voltage up to the legs on the PWR button and when the button is pressed this voltage drops to ZERO and the unit turns on.  The original part was a AO4812 and sometime has been replaced with a ME4920.

*How can I test the IR remote?
Since most people have a digital camera nowadays the cheapest and easiest way is to press the power button on the remote and look at the camera's LCD display.  You should see the remote blinking.

**How can I test the CMOS to see if it is working correctly
Disconnect AC adapter! 
With the ribbon cable connected to the power button, hold the button for 15 seconds to clear the CMOS. Now hold the PWR button down, and at the same time insert the adapter into the DC JACK.  The lights should come on for about 5 sec. then go off. Release the PWR Button. Now attempt to power on normally. Sometimes this works when the CMOS settings are bad.

Watch the orientation of PIN 1.
On DV6000 the leads are up 
On the DV9000 the leads are down
On the F700 the leads are down.
If your memory/recall is bad- take a picture before starting.

Note:  The black KOTEL ribbon cables are flimsy and the strain-relief/support is too close to the leads to help matters.  Remove and insert less often as possible.  If they break you'll wind up looking all over for a substitute. Though similar pitch and length cable can be found on GATEWAY MA3/7 touchpad. (Techs at shops should have a few of them around)  

Also the Gateway MA3 has two of the AO4812 chips on the back side near the 28 pin power output IC. Incase you were needing one.