5V Supply Failure

[Andyroo56]

I built the 5V supply for the Wave Bubble RC1a according to the errata pages with the following components:
IC2 = LM2733, R3 = 36.5k, R4 = 12k, C20 = 560pF. The regulator would start off ok (~4.8V) but then would drop within a minute to around 3.4V. I then decided to remove C20, and changed R3 to 39k; this worked well for about a minute giving a supply voltage of ~5.3V; however after that the voltage jumped up to ~12.5V and burned the tantalum capacitor (C10).

I've replaced C10, changed R3 back to 36.5k and left out C20 unplaced; however, I am scared to turn it on again as I do not have another tantalum to replace this one.

Does anybody have any suggestions? What values for R3, R4 and C20 did anybody else use? I've seen other designs where people removed C20 but I'm not sure that it's okay to. Any suggestions are appreciated, this is quite frustrating.

[lightningserpent]

I had a lot of trouble when I built this section on my wavebubble. One thing to check is inductor; has it bridged, shorted, or otherwise failed. I would also check IC2 and see if it has shorted out.

The L1 value for the LM2733 should be 15uH. I used an LM2731Y and the inductor value was 7.3uH.

I am sure that part of my issues were with the inductors, but I have no evidence to support that claim. At one point I removed everything from the 5v section and placed all new components. I did place C20 (but I ended up removing C60 later on).

My RC1a Build Journal

[soram]

Definitely check the inductor! The contacts appear to be in diagonally opposed corners but actually they span the whole side underneath. It is easy to place them 90 degrees off and solder the 2 corners. It may appear to work fine but if the board flexes it will short across the pads............