Video Teardown and Repair of an Agilent E3634A Power Supply

[Hugoneus]

In this episode Shahriar investigates the cause of failure of an E3634A DC Power Supply. The video features a step by step troubleshooting and teardown of the damaged power supply. Various schematics and block diagram of the power supply are presented. Will he be able to fix it?

http://thesignalpath.com/blogs/2012/06/04/teardown-and-repair-of-an-agilent-e3634a-power-supply/

The Signal Path
http://www.TheSignalPath.com

[metalphreak]

Aaaaah! As soon as I saw you going for that SMD diode with the cutters I was yelling "NOOOO THE PAAAAAADS". Lucky it was one of the unnecessary pads (probably part of why it lifted as well, having no tracks holding it under the soldermask).

Perhaps time to invest in a $100 hot air tool like the Atten Dave reviewed?



It's amazing how many products can be rendered inoperable by a single part failure. I'm sure lots of consumer hardware is thrown out simply because a diode, capacitor, or resistor has failed, and people simply have no interest in repairing them. Something still valuable like test gear is always worth saving!

[Hugoneus]

Yes, given the amount of space I had to work with, I took the risk and decided to break it off. I also knew that one pad was unused and since it had no trace at all, it just came off.

[Rufus]

A lucky or temporary fix IMO.

The zener looked like it had let some smoke out. Hard to imagine a 2.4v zener failing high voltage so it could over dissipate and cook. Mostly likely the 15v rail was overloaded.

I expected the replacement zener to immediately smoke but I guess the overload must and been temporary or intermittent. If it were my supply I would be keeping my fingers crossed.

[codeboy2k]

 .... "29 minutes and a couple of bottles of beer" ...

Thanks for the video... I always enjoy watching.

I expected the replacement zener to immediately smoke but I guess the overload must and been temporary or intermittent. If it were my supply I would be keeping my fingers crossed.

It seems to work. If it goes out again, he'll at least know to look for something overloading the 15V rail next.

[MikeK]

So which do you say shows correct voltage...The readout on the power supply, or the readout on the multimeter?

[Hugoneus]

So which do you say shows correct voltage...The readout on the power supply, or the readout on the multimeter?

Both show the correct voltage.

[Hugoneus]

A lucky or temporary fix IMO.

The zener looked like it had let some smoke out. Hard to imagine a 2.4v zener failing high voltage so it could over dissipate and cook. Mostly likely the 15v rail was overloaded.

I expected the replacement zener to immediately smoke but I guess the overload must and been temporary or intermittent. If it were my supply I would be keeping my fingers crossed.

I appreciate your comment. I don't do things half-way. In the past 48-hours I have put the power supply in an automated loop through its GPIB interface where every 15 minutes a random voltage is selected to deliver 150W to an active load. So far the supply has not failed. After 48 hours of such stressed operation I am more confident to use the supply for my regular experiments. If it does fail again, at least I know where to start looking!

[Rufus]

After 48 hours of such stressed operation I am more confident to use the supply for my regular experiments. If it does fail again, at least I know where to start looking!

So how do you think it failed?

Briefly looking that the circuits there seem to be some reasonably direct paths from the output and sense terminals to op-amps powered from that +15v. Maybe a zap on the terminals caused one of them to latch up and killed the zener before anything else failed.

[Hugoneus]

After 48 hours of such stressed operation I am more confident to use the supply for my regular experiments. If it does fail again, at least I know where to start looking!

So how do you think it failed?

Briefly looking that the circuits there seem to be some reasonably direct paths from the output and sense terminals to op-amps powered from that +15v. Maybe a zap on the terminals caused one of them to latch up and killed the zener before anything else failed.

Perhaps it suffered a power surge, or simply failed due to aging. Of course I can't completely rule out that there is some bigger underlying problem. I will keep you guys posted if it fails again.