Diptrace/PCB design first attempt - sanity check

[ftumch]

Hello forum, this is my first go at Diptrace/PCB design. I'd appreciate a sanity check

My objective is to use a pile of low gain (30-50 hfe) Russian PNP germanium transistors I picked up

The EQD Dream Crusher circuit is designed around low gain NPNs, so I started there and added a voltage inverting charge pump, JFET input buffer and AMZ pickup sim. There may be some mistakes in my schematic

For the PCB itself I've tried to keep the board somewhat symmetrical at the expense of some trace lengths

I made an effort to group the different parts of the circuit together as well as the component types. The power filtering and charge pump are at the top, the biasing resistors are down the right hand side, the input stages are at the bottom. There may be a better way to lay the components out

Russian transistor pinout is back-to-front so I made a new component for them in the component editor. This is something I may have screwed up

For the JFET I've allowed for either SMD pads and through holes. I feel like the routing here is messy and I am 90% confident with my pinout

I was also confused about the best way to route the three way junctions at R5/7 and C5

I made ground planes on of both sides of the PCB. There aren't any specific ground traces and the planes are not separated by circuit function. Power traces are 0.6mm and signal is 0.33

I've enjoyed doing this so far. It reminds of the game Freeways

 

[ftumch]

I've learnt that by unpouring the ground plane I can see both layers at once so I've rejigged some of the traces to keep them apart

Fully aware that I may be rearranging the deck chairs on the Titanic at this point

 

[finebyfine]

Didn't have time to look over the schematic while I'm at work so below is just from visually going off the PCB for now. It looks better than my first PCB and I'm glad I'm not the only one who thinks of laying out a PCB like a video game lol.

For the JFET, you can rotate the SMD footprint 90 degrees counter-clockwise and then move it up so that the drain and source pads are on top of the TO-92 through-holes - that way only the gate pin will need an actual trace to connect them.

I would avoid routing non-power traces under the charge pump if possible. I've never had issues picking up switching noise but I only route under them if I can't find any other way to route the board.

I would nudge the bottom most 9v trace up a bit so it's surrounded by your ground pour. For 3 way meets on the same layer I will just have them meet to form a "T".

 

[szukalski]

Your input and output cross over each other. If you have a high gain signal then that could introduce oscillation.

 

[ftumch]

@finebyfine @szukalski thanks both, much appreciated. I've jigged it around and applied your suggestions. I haven't done much over on the left hand side so I might give that another pass tomorrow

It needs straightening up now, and I can probably tuck it in to reduce the overall dimensions