HP/Agilent 3634A Repair

[sahko123]

This thread is basically for anyone who has the issues listed below and what I did to repair them.
1. Non-sensical/maxed out reading beyond range.
2. Collapsing voltage at any current

1. One of my 3634As when powered read beyond maximum voltages at around 54V and 7A. Part of the manual tells you to test the bias voltages. In mine the +15 rail read 0.8V which was traced to a broken CR9 2.4V zener diode. I repaired it with a through hole zener diode with the same voltage but higher power dissipation. This was because I couldn't be arsed waiting for the actual part to arrive and it made no difference. But anyways fixing this diode made the power supply work fine with no strange readings. I think the measurements are based of these rails but thankfully the 3634A has a calibration mode that can be accessed a power cycle while holding error for 5 seconds at power up.

1. (solution) Replace CR9 2.4V zener

2. The second 3634A would output the displayed voltage perfectly and even voltage calibration worked perfectly the problem was that when a current of more then 1 or 2mA was applied the voltage would collapse completely. This problem was slightly stranger to troubleshoot everything on the board measured fine at first I thought that the power regulator was at fault. I decided to test the components directly on the output by feeding 10V into the input and tracking it back. There was a resistor directly across the output which is 4.7K i measured the voltage across the resistor which should have been ten while inputting 10V from another supply. This resistor should have been directly across the output. Then i noticed that the sense wires are connected to the output through a resistor internally, I had them shorted externally on the binding posts. So I continued tracking the output voltage and noticed that there was no 10V anywhere but the actual binding posts. As it turns out the binding posts are not one piece from output to the board they are soldered just on the inside. So measuring the continuity from the binding posts to the board meant that the post itself was at fault. The solder was reflowed and more was added to make sure this fixed the supply. The reason i measured the correct voltage was because the sense and output was shorted to each other with the power output going through the resistor connecting them internally was output to the sense binding posts which was in turn connect to the output through the short. So the collapse was caused by the dicky binding posts

2. (solution) Reflow or add solder to the binding posts or specific dysfunctional binding post.