If you elect to use a bulky PP cap, you could easily re-locate it in-line, rather than on-board. Often, a grounded wrap of copper or tinplate foil around a large film cap will just about eliminate any hum induced by a noisy location or wandering power leads etc. in the vicinity. Otherwise, with shunt capacitors, you can usually identify the outermost foil layer on a clear plastic type and make the corresponding lead the grounded end.
A question though; how big a cap do you actually need? At the input impedance of a P3A, you don't really need a massive coupling cap to get down below audio frequencies. Anything larger than specified. is likely for bragging rights rather than for any sonic benefit.
A question though; how big a cap do you actually need? At the input impedance of a P3A, you don't really need a massive coupling cap to get down below audio frequencies. Anything larger than specified. is likely for bragging rights rather than for any sonic benefit.
An interesting suggestion, thanks for that.
Ideally I would use the 4.7uF suggested in the BOM.
The only thing I could find that would fit was a 1uF.
Iirc that puts the corner frequency a little above 20Hz hence the desire for a larger value.
I think Rod's build, as seen in fig 3 of the P3A ESP article, actually shows only a 1uF cap, though it's more a square footprint and tall, perhaps to illustrate what is required to fit the board layout. Perhaps he even started out with a polyester film version. Vishay and other competing PP film cap brands aren't cheap in any size or value but several similar types are available from Mouser in the 10 mm pitch and it should be possible to to find something that fits from one of the many PP film type ranges. Just don't expect local resellers to stock anything more than typical replacement types in the cheapest brand ranges.
FWIW, I've used various caps in several P3A builds, from 1-10 uF cheap polyester to mica and film/foil, with little difference in perceived bass strength, unless the speakers were large enough to be only 3dB down at 20Hz anyway. Few loudspeakers anywhere are really that good and I'll assume your headphones aren't heavy bass reproducers either. Try a 100V 1uF cap that fits then double up to see if you perceive any difference in your system.
FWIW, I've used various caps in several P3A builds, from 1-10 uF cheap polyester to mica and film/foil, with little difference in perceived bass strength, unless the speakers were large enough to be only 3dB down at 20Hz anyway. Few loudspeakers anywhere are really that good and I'll assume your headphones aren't heavy bass reproducers either. Try a 100V 1uF cap that fits then double up to see if you perceive any difference in your system.
Nikos,
Do you have a schematic to accompany this? This doesn't appear to be the standard P3A. From a quick review I can see what appears to be a ground lift and current mirror on the IPS.
Two comments on the layout:
Do you have a schematic to accompany this? This doesn't appear to be the standard P3A. From a quick review I can see what appears to be a ground lift and current mirror on the IPS.
Two comments on the layout:
- I'd suggest a different ground layout for the 100u caps. The positive and negative caps should tie then join to the ground point. Any other ground connections would connect between this and the PSU ground connections (T grounding). As you have it layout out, the 22K CCS ground resistor connects to a leg of the "T", not after the center point. It will be subjected to large ground currents between the main caps.
- The main rail caps should ideally be close to the output devices. As you have it laid out, the caps are at the opposite end of the boards. I realize this is tricky to change with the layout you are using.
Those are good suggestions. Also, with a CFP output stage, the Vbe multiplier transistor (Q9) should be thermally coupled to one of the drivers to track their temp. You can it isolated. If you look at Rod's PCB's you can see that he has Q9 & Q5 next to each other. This allows the builder to use a zip tie to bond them together.
If you want to improve the grounding further, here's a quick drawing showing how to use the "T" grounding approach with a star point. This is the same grounding architecture the Wolverine Amp uses.
This approach allows the heavy currents from the rail caps to converge and resolve before flowing to the return.
The Wolverine team also found that tying the ground lift to the speaker ground and allowing that to resolve before joining the star point produced better results.
This approach allows the heavy currents from the rail caps to converge and resolve before flowing to the return.
The Wolverine team also found that tying the ground lift to the speaker ground and allowing that to resolve before joining the star point produced better results.
Attachments
