router troubleshooting

[clearchris]

hi.

I have a router that died some years ago. Recently, I replaced all the caps, some were obviously bulged. I also upgraded the ram chips.

When I plug it in, I get nothing. No power light, no leds. A quick once over on the board shows that the rt9202 is hot to the touch. Its a pwm buck voltage controller.

So, how do I start here? Is a hot voltage controller indicative of a short somewhere else?

I'm also considering pulling the ram chips off and check if I see the same symptoms.

Thanks!

[Agent24]

Re: router troubleshooting

So, how do I start here? Is a hot voltage controller indicative of a short somewhere else?

Very likely. Check your work on the capacitors, there may be solder bridges, make sure the capacitors are not backwards. The short may even be inside the IC itself.

I'm also considering pulling the ram chips off and check if I see the same symptoms.

Probably easier to look around the DC-DC converter first.

Photos would be good

[clearchris]

Re: router troubleshooting

Hot chip is closest to the power jack.

How would you chase down a short? Remove components until the chip stops heating? Would that work for memory chips? Capacitors?

[Mad_Professor]

Re: router troubleshooting

Hot chip is closest to the power jack.

How would you chase down a short? Remove components until the chip stops heating? Would that work for memory chips? Capacitors?

Start with basic components, diodes and transistors then check for possible shorts from caps and resistors and so on, then work up to logic and gate components.

[Agent24]

Re: router troubleshooting

That RT9202 drives the dual 9936 MOSFET right next to it. I would check it is not shorted first

Also check the RT9202 is not shorted from supply to ground etc.

[clearchris]

Re: router troubleshooting

On the rt9202 I get a connection to ground on Ground, ugate and phase.
On the 9936, i get a connection to ground on s1, g1, s2, and both d2 pins. That assumes the triangle on the chip marks pin1.

Would that indicate a problem with the mosfet? I wasn't expecting to see a connection to ground on the drain. Or g1 for that matter.

[Agent24]

Re: router troubleshooting

It could be a problem, best to remove the MOSFET and check it out of circuit for sure.

[clearchris]

Re: router troubleshooting

http://www.decibelcar.com/menuelectro/93-fettest.html
Tested mosfet using these instructions. Mosfet passes all tests.

My suspicion turns back to the rt9202. The spec sheet looks pretty complex, Im not sure there is an easy way to test this out of circuit. If I plug the router in without the mosfet that could eliminate the rest of the board as the problem, right? If the rt9202 gets hot, its fried or one of the resistors or diodes (those black boxes with the Z are diodes?) around it are the problem. If the rt9202 stays cold, the problem is with the mosfet or something downstream of the mosfet.

Comments?

[Longbow]

Re: router troubleshooting

Why would you plug it in with the 9936 removed? That is not a test of anything. Pull it out and check the two devices. Most of the time a faulty MOSFET will have a short between the gate and S or D. Post your readings.

[clearchris]

Re: router troubleshooting

on the rt9202 in circuit I have 5v at boot and ocset pins. 300mv at ugate and 600mv at lgate. If the output is pulsef, would this be an expected reading?

On the mosfet out of circuit, I got a reading of .1 from signal to drain with the mosfet discharged. After charging the gate, the reading went to .6. No continuity detected between gate and drain or signal. Is that what you are looking for?

[Longbow]

Re: router troubleshooting

.1 ohms from source to drain in both directions? They are shorted. These fet's are enhancement mode. That means they will be o/c until a forward bias is placed on the gate, in this case +3 to +4 volts. Note the protection diode across d-s. You should see near infinite resistance in one direction and a diode junction in the other direction.

[clearchris]

Re: router troubleshooting

Deleted post, was going from memory, and I don't know this well enough to do so. Desoldering chip again, will post full results.

[momaka]

Re: router troubleshooting

On the rt9202 I get a connection to ground on Ground, ugate and phase.

Check again with the 9936 MOSFET removed from the board. If you still get short circuit to ground on either UGATE or LGATE of the RT9202, then the RT9202 is toast.

When a MOSFET develops a short between S and G or D and G, the results are often a fried buck controller as well. For this reason, I suggest replacing both the RT9202 and the 9936 MOSFET.

Also, Phase should be connected to that large toroidal coil next to the MOSFET, so depending on the static resistance of the router CPU (or whatever is powered from that buck rail), the resistance may appear as a short circuit without there necessarily being a short circuit. Usually, this only applies to computer CPUs, GPUs, and other moderate-to-high power chips. But it's something to keep in mind when testing, anyways.

[clearchris]

Re: router troubleshooting

Ok. desoldered the chip again, and through a minor miracle, found it when it popped across the room.

+Drain to -source: no circuit
+drain to -gate: no circuit
+source to -gate: no circuit
-gate to +source: no circuit
+gate to -drain: no circuit

after discharging chip (placing it on a piece of metal, pins down)
+source to -gate: .5v forward voltage drop

after charging chip (positive probe on gate, negative on drain in diode mode)
+source to -gate: .1v forward voltage drop

[clearchris]

Re: router troubleshooting

Last part of last post should have read:

after discharging chip (placing it on a piece of metal, pins down)
+source to -drain: .5v forward voltage drop

after charging chip (positive probe on gate, negative on drain in diode mode)
+source to -drain: .1v forward voltage drop

[clearchris]

Re: router troubleshooting

So, I have a possible lead here. I have been going back to basics as I have never had any formal electronics training, and I have also been centering my troubleshooting back on the rt9202 chip.

Vcc, by the spec sheet, expects 5v. On the circuit board, to the right of RA1 (the side with the printing) I see 5v. To the left of RA1, I see 1.5v. I don't see a voltage divider here. I'm at a loss as to what would cause a voltage drop like that.

If resistors burn, would this be expected? Would a shorted ceramic capacitor pull the voltage low like that?

[Longbow]

Re: router troubleshooting

This method of testing is very confusing. You seem to be measuring between the Source to the Gate, or between the Drain and the Gate. Regardless of any external voltage or static charge, these readings should be infinite, since there should be no connection at all between the channel and the gate. Go back to the Source and Drain with the chip discharged. One direction S-D should give you a diode junction, the other direction (meter probes reversed) should give you an infinite reading.

If all the above tests look good, the device is probably ok. Then you can play with biasing the gate, although that method is hit and miss with the chip floating in air.

As suggested, if you are getting any reading from the channel to the Gate, that means ugly voltages have traveled through the Gate back to the driver IC and smoked it.

[clearchris]

Re: router troubleshooting

Yeah, I got the verbiage wrong, but it was past an hour, so I couldn't edit the post. I corrected it in the following post.

Testing was done from source to drain. I think the mosfet is ok.

Going to pull the rt9202 and possibly the resistor and some smt caps off the board and see what I can learn. Hopefully I can chase down why I'm not getting 5v to vcc.

[Agent24]

Re: router troubleshooting

Going to pull the rt9202 and possibly the resistor and some smt caps off the board and see what I can learn. Hopefully I can chase down why I'm not getting 5v to vcc.

Either RA1 is has gone high value or a decoupling capacitor or the RT9202 itself is shorted.

Since the RT9202 is what gets hot, it is most likely the problem. I suppose it's *possible* that the decoupling capacitor(s) could go open circuit and cause the controller to become unstable and overheat, but more likely it's shorted.

[clearchris]

Re: router troubleshooting

I pulled the rt9202 off and tested the board without the mosfet or the rt9202. Now I get 5v on both sides of RA1. The rt9202 is toast.

Seems I can actually source some of these, going to end up coming from china though.

Thanks guys!