Re: Dell 2407WFPb Monitor - Bad Cap
I found an identical working monitor and made some progress debugging this situation (and get my monitor working again). But I did not have permission to desolder the zener on the working unit, only to measure normal voltages.
The drain of the hexfet should be at 5VDC. (Remember this is a p-channel hexfet, with drain going toward load.) This goes to an inductor which then goes to a smoothing capacitor (two 220uF in parallel) in parallel with the zener to ground. In other words, the zener is after the drain, and across 5VDC and ground. So it is very likely to be larger than 5V, and intended as a shunted load overvoltage protection device should the hexfet fail (as it did). So the hexfet was on for too long, the load saw more than zener voltage, the zener shorted out, which then made the hexfet melt. I think the designer intended it exactly this way.
I have replaced the unknown zener with a 6.8V 5W zener. I also removed the inductor and burnt hexfet, and wired up an external 5VDC supply to the top of the zener. The monitor works! (My recapped SMPS supplies 19V and 7V fine. In the SMPS board, on the 19V output line, I am planning to add a 27V 5W zener in parallel with a 30V MOV after the 5A fuse to ground.)
My only remaining question is, should I continue like this with an external 5V power supply, or should I try to replace the hexfet and inductor back in the board. It would be nice to get back to one power supply. The danger is, if the gate control circuit was damaged, the zener and hexfet will do fireworks again, and this time I may not be lucky and downstream components may fry.
Given the reports here and elsewhere, Dell should have designed the hexfet and zener as snap-in components ;-)
Originally posted by chappers
View Post
The drain of the hexfet should be at 5VDC. (Remember this is a p-channel hexfet, with drain going toward load.) This goes to an inductor which then goes to a smoothing capacitor (two 220uF in parallel) in parallel with the zener to ground. In other words, the zener is after the drain, and across 5VDC and ground. So it is very likely to be larger than 5V, and intended as a shunted load overvoltage protection device should the hexfet fail (as it did). So the hexfet was on for too long, the load saw more than zener voltage, the zener shorted out, which then made the hexfet melt. I think the designer intended it exactly this way.
I have replaced the unknown zener with a 6.8V 5W zener. I also removed the inductor and burnt hexfet, and wired up an external 5VDC supply to the top of the zener. The monitor works! (My recapped SMPS supplies 19V and 7V fine. In the SMPS board, on the 19V output line, I am planning to add a 27V 5W zener in parallel with a 30V MOV after the 5A fuse to ground.)
My only remaining question is, should I continue like this with an external 5V power supply, or should I try to replace the hexfet and inductor back in the board. It would be nice to get back to one power supply. The danger is, if the gate control circuit was damaged, the zener and hexfet will do fireworks again, and this time I may not be lucky and downstream components may fry.
Given the reports here and elsewhere, Dell should have designed the hexfet and zener as snap-in components ;-)
Comment