That looks very fine, Winfried. You basically have designed a 2.5-layer board with those supply jumper wires ;-)
Thank you Klaus, your positive feeback makes me feel confident that this circuit will work and do what I need from it with decent quality. I'll probably be back here when the boards are assembled and ask for advice for some measurements to verify performance.
Thanks and Regards,
Winfried
Thanks and Regards,
Winfried
^
To avoid plane slots, and it's a one-off PCB anyway.
But true, shorts slots should not be a problem here as the GND plane still is quite solid now and any GND signal references are RF only.
Usually one would bridge the slots as much as possible on the top side with GND fills and vias and it looks like there would be just enough empty space between the L and R channels to do that (after moving that one mounting hole)...
To avoid plane slots, and it's a one-off PCB anyway.
But true, shorts slots should not be a problem here as the GND plane still is quite solid now and any GND signal references are RF only.
Usually one would bridge the slots as much as possible on the top side with GND fills and vias and it looks like there would be just enough empty space between the L and R channels to do that (after moving that one mounting hole)...
Folks,
currently I don't see/understand a technical reason requiring me to avoid Supply wires. And "yes" the PCB is a one-off (or better 2-off 😉 ) project for me personally, because I modify two DEQX PDC units and that's it... By the way: Mounting holes are not movable as the PDC main board "dictates" their placement.
If substantial technical downside exists against the supply wires, please state clearly, so I understand the reasoning and I may reconsider, but at this time I feel confident to order PCBs (already having received all parts to populate the PCBs 🙂 ).
Thanks for all the help, the thoughts and considerations!
Winfried
currently I don't see/understand a technical reason requiring me to avoid Supply wires. And "yes" the PCB is a one-off (or better 2-off 😉 ) project for me personally, because I modify two DEQX PDC units and that's it... By the way: Mounting holes are not movable as the PDC main board "dictates" their placement.
If substantial technical downside exists against the supply wires, please state clearly, so I understand the reasoning and I may reconsider, but at this time I feel confident to order PCBs (already having received all parts to populate the PCBs 🙂 ).
Thanks for all the help, the thoughts and considerations!
Winfried
No technical downside... assuming you are keeping them short and snug fit to the PCB (no coils ;-)
Inductance should be as low as possible. The ESL of the 330uF plus any wiring inductance will form a LC-tank circuit with the C of the 100nF, increasing supply impedance to several or even tens of ohms at MHz frequencies. Best way to mitigate this is a large(!) ESR of the 330uF, > 100mR or so. That's why tantalum caps are often used, and/or standard grade small aluminum electrolytics.
Paralleling caps always invites impedance issues the moment there is little damping and enough inductance. You'd be surpised how many circuits and layouts are affected by this as most audio designes don't check supply impedance at the chip with a network analyzer.
https://www.electronicdesign.com/te...-tantalum-and-ceramic-capacitors-can-be-risky
Inductance should be as low as possible. The ESL of the 330uF plus any wiring inductance will form a LC-tank circuit with the C of the 100nF, increasing supply impedance to several or even tens of ohms at MHz frequencies. Best way to mitigate this is a large(!) ESR of the 330uF, > 100mR or so. That's why tantalum caps are often used, and/or standard grade small aluminum electrolytics.
Paralleling caps always invites impedance issues the moment there is little damping and enough inductance. You'd be surpised how many circuits and layouts are affected by this as most audio designes don't check supply impedance at the chip with a network analyzer.
https://www.electronicdesign.com/te...-tantalum-and-ceramic-capacitors-can-be-risky
OK. These are interesting details! So, as the supply input RC is through-hole, change of these parts would be simple. Measuring the actual implementation's HF behavior is much harder, though, as I don't have the necessary equipment, so may/will have to go with BKMs to be on the safe side 😉
Greetings,
Winfried
PS: PCBs ordered.
Greetings,
Winfried
PS: PCBs ordered.
Brief update: PCBs arrived. Had ordered three to have one for potential experiments and one for each DEQX PDC unit. I'll be back.
Cheers,
Winfried
Cheers,
Winfried
Progress is a snail...
The 1% C0G 100p and 220p Caps only arrived today, will proceed with first build now and be back.
Please keep fingers crossed that the board will work! I'll be back.
Greetings,
Winfried
The 1% C0G 100p and 220p Caps only arrived today, will proceed with first build now and be back.
Please keep fingers crossed that the board will work! I'll be back.
Greetings,
Winfried
I'm back as promised 😉🙂 ...but unfortunately with sort of bad news:
Failure of the DEQX PDC (it seems age took its tall...) has inhibited tests and due to unavailability of schematics self repair seems next to impossible and payed repair would be done in Australia for lots of money...
So let me thank you all for the very engaged and competent advice received here for my project. I'm now thinking about an alternative which I can afford, but that is obviously off-topic here and I may start another thread about this topic.
Thanks and best regards,
Winfried
Failure of the DEQX PDC (it seems age took its tall...) has inhibited tests and due to unavailability of schematics self repair seems next to impossible and payed repair would be done in Australia for lots of money...
So let me thank you all for the very engaged and competent advice received here for my project. I'm now thinking about an alternative which I can afford, but that is obviously off-topic here and I may start another thread about this topic.
Thanks and best regards,
Winfried