The new "My Ref" Rev C thread

[ClaveFremen]

Originally Posted by linuxguru

The R3 heatsink issue is messy - maybe the best option is to use a high-mounted heatsink without lower fins, which should allow it to comfortably clear the components around it.

The heatsink in my photo has already the smaller footprint you can probably find....

I can't see enough space for the FKP2s (C12, C30, C32) that are part of the Ultimate BOM...

[linuxguru]

Originally Posted by ClaveFremen

I can't see enough space for the FKP2s (C12, C30, C32) that are part of the Ultimate BOM...

I've changed the outlines, but the pitch and the space around them are the same as before. Some space can be created for them, and I can revert to the original outlines easily - no problems.

I got the idea that C7 can be shifted to the left of the LM318 (next to pins 4 & 5), which will shorten the trace length from +Vs through C7 to -Vs of the LM318. This can be done by nudging the LM318 to the right by 2.5 mm or so, and rerouting one signal trace. Another advantage is that C6 and C11 can be brought closer.

Any comments/concerns about this?

[ClaveFremen]

Originally Posted by linuxguru

I've changed the outlines, but the pitch and the space around them are the same as before. Some space can be created for them, and I can revert to the original outlines easily - no problems.

I suggest you to use C5/4.5 from the capacitor-wima library so you have the exact layout.

[quadtech]

Originally Posted by madisonears

Linuxguru, you certainly do know your way around this amp, and I apologize for my earler comments in the other thread.
Tom E

Tom, that is really nice. I did wonder about your original comment.
Linuxguru's posts, esp on the solidstate forum, have been very informative
and educational.
I built the MyRef from Uriah's first group buy, and it has been excellent.
Appreciate every effort to improve it, from everyone here.

[quadtech]

Did anyone else find the PGND connector hard to get to, or
is it just my fat fingers?
How about an optional PGND connector in the space between C2 and C3?

[troystg]

Originally Posted by quadtech

Tom, that is really nice. ......
Appreciate every effort to improve it, from everyone here.

I second that!!

[linuxguru]

Originally Posted by quadtech

Did anyone else find the PGND connector hard to get to, or
is it just my fat fingers?
How about an optional PGND connector in the space between C2 and C3?

It is hard to get to - I soldered the PGND wire, so it's not a problem for me. But people who use a tag connector may find it tiresome. I'll see if I can re-route PGND to the bottom edge.

[regiregi22]

Originally Posted by troystg

I second that!!

+1!

Regarding the PGND issue, yeah, it is a very difficult location for plugging the connector. I soldered a cable, so I have the metal tab only for measurements purposes at the moment. It will be used for earthing purposes.

[linuxguru]

Originally Posted by linuxguru

I got the idea that C7 can be shifted to the left of the LM318 (next to pins 4 & 5), which will shorten the trace length from +Vs through C7 to -Vs of the LM318. This can be done by nudging the LM318 to the right by 2.5 mm or so, and rerouting one signal trace. Another advantage is that C6 and C11 can be brought closer.

OK, I made this change and everything fell in place smoothly. Well, almost everything - C7 and C10 still need to be cleaned up a bit, which should not be a problem. The big gain was that C6, C11, the zeners and the small-signal star ground are now in a near-optimal placement and routing configuration. A side benefit is that R39, R43 and C34 could also be shifted a bit to the right, gaining additional clearance for R3 and its heatsink. C12, C30 and C32 now have the correct outline for the fat FKP2 package - a tight squeeze in some places, but managed.

Edit: Ignore the unconnected capacitor outline at the bottom - I was going to use it for C17 and C18, but left it for later.

[regiregi22]

Capacitors around the opamp are really well placed now, the shortest traces are achieved.

What are you planning to stuff in R7, a miniature nuclear reactor for supplying the power IC? Now seriously, why such a really big resistor there?