A couple thoughts; the pads for the variable could be larger as could the ones for the stakes on the heat-sink. You can see I just made a nasty big solder ball on the stake for stability. Also, you can see where it's not a good idea to bend the 5W diode when mounted flush. A slightly larger hole for the stake hole too. It's just able to fit the diode, and makes removal precarious with regards to pads. This is an older design from Mark, but if you are planing on having a board made, you might consider these few changes.
I got around to testing the Tuba today, something isn't right, I have 35V on the outputs and a drop of 1v across the test points, and VR does nothing. I checked all the diodes and resistors, all appears to be right there. All the caps are correct. The relay works as it should. I have LM1084 in place of the LD1084, as far as I can tell that should be okay?
It's a pretty simple circuit, but I'm missing something.
I followed Marks power up and test procedure too.
It's a pretty simple circuit, but I'm missing something.
I followed Marks power up and test procedure too.
Last edited:
Update & Warning!: The LM1084 is not a direct swap, for those that can't find the LD1084. I'm sure it would work with some tweaks, but not as designed. I found some LD1084s I didn't know I had and swapped them in... adjustment perfect, no issues at all.
The holes are pretty close to the component size on these and I pulled a trace and had to pull a cap and solder that pin from the top. Hint: if you have very close ID/OD I found it best to heat, or even add solder then heat all and pull the part. After the part is removed, you can go back and clear the holes.
The holes are pretty close to the component size on these and I pulled a trace and had to pull a cap and solder that pin from the top. Hint: if you have very close ID/OD I found it best to heat, or even add solder then heat all and pull the part. After the part is removed, you can go back and clear the holes.
Last edited: