chief mushroom cloud wrote: ↑Sun Aug 26, 2018 3:34 pm
Nice work! I like those Phoenix/Weidmuller type screw terminals, too. But be sure to dab a dot of flexible acrylic nail polish into the setscrew wells....they can vibrate loose. I know this from experience.
Thanks and good tips! I've added some comments below:
But I'm curious....did you allow for creepage clearance around power supply diodes? When you fire it up, watch for potential arcing....just turn off the lights to really see it. If you don't see it, don't worry about it.
Ah, this is a great point - I'll be sure to check on first power-up. It's this sort of input that would have been valuable early on! There's always a v2.
What copper thickness did you specify in your fabrication guidelines doc?
The high current traces (so heaters and B+) are 2mm wide on 1oz copper. The rest of the traces are 1mm.
On the PI board....what PI variations can it accommodate?
Restricting to a traditional LTPI, it has spaces for negative feedback / presence control, as well as pads for a fizz cap and space for a 'scale' control as mentioned in Merlin's book. I've also added the heater elevation components to this board as the grounding for it returns to the PI filter cap. Schematic attached.
Also, I would recommend adding a wirepad/Fast-on or ??? right at the neg side of 1st rectifier filter cap....and run PT HT CT there or neg side of bridge rectifier, whichever is pertinent. Reference a second gnd wire to the chassis from there. I always put two wirepads (or fast-ons) at this point on every power supply. The rest of gnd plane can attach here as well.
If you have gnd planes on both sides...it's good practice to stitch them with additional vias, regardless of signal rise times (or frequency, in this case).
My grounding is a bit overkill. I have separate audio and chassis grounds throughout. On the power supply board I have multiple connecting points so each filter cap's ground for B+1 and B+2 isn't actually connected through the PCB - they can be configured as required - with the OT connection made to point
x or
y, e.g., whatever you prefer. Scheme attached.
The copper pours (chassis ground) can be attached to the chassis through normal standoffs, or isolated using a plastic one at one corner. Each board has its audio ground returning to the stage's filter cap negative. Also, there are connecting points on each board for the 'correct' returns to attach - ie, the gain control ground is returned to the following stage's filter cap -ve. The sole connection between audio ground and the chassis can be controlled as one wishes - I'll put it at the furthest point along the chain from the power supply. The grounds can all feed back to a central point (star) or daisy-chained. I tried to keep it generic!
Good to know we have a KiCad expert in da haus....
Ha - I wish I was. Many hours of frustration and googling.
Once I know it all works as expected I'll tidy up the files and share - hopefully someone can improve them and we can all benefit with improved versions.
You do not have the required permissions to view the files attached to this post.
. Are your KiCad files shareable? I'm trying to learn it, your right about the steep learning curve.
