Broken power button

[dchang0]

If anyone has examined the switch wiring with the cover off, can you confirm true or false for that?

It might be a multilayer PCB (more than two layers). If so, then just looking might not be enough.

When I desolder the old, broken switch, I can use a multimeter in continuity testing mode and try touching nearby solder pads to see if pin 1 does indeed go to something else. There's a capacitor right next to the power switch.

I'll get the batch of new switches May 6 and report back with my findings. The $4.99 shipping cost more than the five switches.

 

[Deleted member 22229]

It is my understanding that in at least some models there IS an important difference between the switch and just pulling the plug. Some models apparently use the second set of switch contacts to short the power supply to ground when turned off, thus instantly discharging the capacitors.

Just soldering the contacts closed but leaving the switch physically in the off position would create a situation in which the power supply was trying to energize but was looking at a short circuit to ground--not good for power supply life.

If anyone has examined the switch wiring with the cover off, can you confirm true or false for that?

Thank you.

That is just not true. There is NO drain cap built in to the switch.

 

[Grisu]

It is my understanding that in at least some models there IS an important difference between the switch and just pulling the plug. Some models apparently use the second set of switch contacts to short the power supply to ground when turned off, thus instantly discharging the capacitors.

Just soldering the contacts closed but leaving the switch physically in the off position would create a situation in which the power supply was trying to energize but was looking at a short circuit to ground--not good for power supply life.

Yes confirmed, desolder switch in that case, did not mention this and you made a valid objection!
I included this part in my thoughts but didnt write down, sorry.
Dont know if it will be relevant for the circuit in this particualar case, anyway better to remove the switch to be on the safe side without any further analyzing for no more than wasting time.

Never thought this will be read that carefully if at all

 

[dchang0]

I've got good news.

The E-Switch LC-1258-EE-NP switch is a perfect fit and functions correctly as a replacement power button in my RT-N66U (outside label is RT-N66R, but PCB says RT-N66U Rev. 3.20).

A few notes:

1) The ground plane on the PCB makes desoldering the switch a real pain. I could not do it with a 25W pencil soldering iron nor with just a Hakko 808 desoldering gun. I had to use a heat gun with a small nozzle that was cranked up to 1000 deg. F to pull out the switch. I clamped some forceps on the metal body of the switch and simply heated up the pins until the solder visibly melted, then it pulled straight out. It came out quite easily.

Clearing the holes of solder was a pain. I had to use the heat gun PLUS the Hakko 808 to heat up the area and suck the spare solder out of the holes.

Resoldering was easy except for the two metal housing pins. I didn't want to heat the switch up too much since the innards are plastic, so I didn't solder those as well as I could have.

2) After removing the switch completely, I used a multimeter in continuity mode to try to trace the pins. Pin 1 (nearest the plunger) is definitely connected to the ground plane. I repeat, pin 1 definitely goes to ground. So, when the actuator is out (switch is off), it shorts pin 2 to pin 1's ground.

Pin 3 appears to go to one leg of the nearby large electrolytic capacitor.

I could not figure out where pin 2 goes. The PCB does appear to have more than 2 layers (top and bottom). If you look closely at the top and bottom, you cannot see any copper traces for pin 2, so that means one of the inner layers routes pin 2 somewhere. I tried randomly touching just about every exposed land or pin to find where pin 2 goes and never did find it.

3) I suspect that the original switch is a fake E-Switch. Even though it is a perfect match in dimensions and functionality, it appears more cheaply made. My bet is that some third party Chinese company cloned the E-Switch and ripped off ASUS.

Looking at the way the old switch and the new switch work, there is a tiny copper latch that moves back and forth and catches on the blue plastic plunger. You can see the latch move by looking through the V-shaped hole in the top of the metal housing.

The broken switch's copper latch doesn't move back and forth any more.

So, it is possible that with this genuine E-Switch LC-1258-EE-NP, it may last forever.

I'll post up photos if I can. Not sure if this forum will allow me to.

 

[dchang0]

Here are the photos.

1) New and old switches. Note that the bent metal pins that clamp the metal housing to the PCB are better formed on the new E-Switch. The old one could be a fake or could be a prior revision, but it looks like a fake to me.

NOTE: the old switch has had its spring and spring retainer removed.

 

[dchang0]

2) Side view of new and old switches. The metal housing is clearly better on the genuine E-Switch. Even the blue plastic plunger is better formed--it has faceted corners while the old switch has sharp corners.

 

[dchang0]

3) Bottom of PCB after switch removal. Solder is still stuck in the holes. I took this mainly to show that there are no visible copper traces going to or from pin 2.

(Pin 1 has the square pad. Pin 2 is the center pin.)

 

[dchang0]

4) Top of the PCB after the switch was removed. Also can't see any copper traces to or from pin 2. Pin 3 appears to go to one leg of the green electrolytic capacitor shown next to the power jack in this photo. The copper trace that does this is on the bottom of the board and is easily visible (see photo #3 above).

 

[dchang0]

5) Pin and housing holes with solder sucked out. You can see clear through them.

 

[dchang0]

6) New switch installed and working!

I guess my advice is this: if you have the right tools (heat gun with small nozzle attachment and solder sucker), go for it. If not, then this repair will be too hard due to the large ground plane acting as an efficient heat sink.

DO NOT use a soldering gun (not the same as a soldering iron) to work on delicate circuitry like this. Use only an ESD-safe pencil soldering iron.

You probably don't need a fancy solder sucker like the Hakko 808 desoldering gun. Just a $5 manual squeeze bulb or syringe type solder sucker with a metal tip will probably work fine. The heat gun is the one keeping the solder liquified. You just need a solder sucker that can handle the prolonged heating.

 

[NoelS]

Amazing job and documentation. Thank you for doing this!

 

[dchang0]

I took apart the old, failed switch.

There is a thin copper wire bent in a sort of upside-down W shape. The shorter rear end of this swings back and forth in a little molded indentation in the tail end of the blue plastic actuator. The longer front end of this W goes up and over the top of the copper spring, down through the copper spring (see the tiny hole) and then down into the long center slot in the top of the blue plastic actuator. This longer end of the W essentially stays still because of the tiny hole in the spring holding it in place.

It looks like the blue plastic either melted or wore out in the indentation at the tail end so that the copper wire's end no longer toggles back and forth. What this means is that people can't open up their original, failed switch and repair it, then put it back in the router.

This view is of the top of the blue actuator.

When the switch is assembled, you can see the tiny hole in the copper spring and the longer end of the W wire at the vertex of the V-shaped hole in the top of the housing but not much more.


The genuine E-Switch is rated for 85 deg. C, but I measured the temperature of the switch and found that it sits around 48 deg. C. (Ambient temp is around 20 deg. C.)

But I know that I didn't press the button very often, so it is more likely that the failure was due to melting rather than wear. The genuine E-Switch is rated for 10,000 cycles. The failed switch might never have been cycled more than 200 times in its entire life, but it has sat there for 3 years straight, 24-7, through some really hot summers.

Perhaps the old switch was made with a far inferior plastic that failed at much less than 85 deg. C, but I find that hard to believe. Most of the industrial plastics would never melt anywhere as low as 50 deg. C. And if the interior had gotten very, very hot, the router's plastic housing should have at least warped.

This would suggest that it was mechanical wear, but again, I didn't press this button nearly enough to wear it out.

I did not take apart a genuine E-Switch to see if the failed one had changed the mechanical design. That's a possibility, that they cheated and cut some corners on the actual latching mechanism.

Strange problem inside the switch that will probably never be diagnosed.

 

[RMerlin]

You know what the chart says... Duct Tape and WD-40 can fix every thing.

- Does it move? If yes, Is it supposed to move? If yes, then it's OK.
- Does it move? If yes, Is it supposed to move? If not, use duct tape.
- Does it move? If not, is it supposed to move? If not, then it's OK.
- Does it move? If not, is it supposed to move? If yes, then use WD-40.

 

[NoelS]

That's "DUCK" TAPE! (quack)

 

[ColinTaylor]

That's "DUCK" TAPE! (quack)

Except when it isn't. http://archive.boston.com/bostonglobe/ideas/articles/2010/03/14/tale_of_the_tape/

In the UK at least, "duct tape" is the generic name of the sticky fabric tape we're familiar with today. "Duck tape" is a trademarked brand name. "Duct tape" appears in the dictionary but "Duck tape" doesn't.

 

[NoelS]

Gotta love the English Language (King's or otherwise)!
(quack-quack)!

 

[dlandiss]

That's "DUCK" TAPE! (quack)

I have used it to repair a duct, but never to repair a duck.

 

[RMerlin]

Quack, damn you.

(bonus points to anyone who can track down that reference, which also ties to other duct tape references )

 

[dlandiss]

Quack, damn you.

(bonus points to anyone who can track down that reference, which also ties to other duct tape references )

 

[AndyN]

I've got good news.

The E-Switch LC-1258-EE-NP switch is a perfect fit and functions correctly as a replacement power button in my RT-N66U (outside label is RT-N66R, but PCB says RT-N66U Rev. 3.20).
(...)

Does anyone know switch symbol fin asus rt-ac68u?