My only perhaps useful comment is to make the pads for the small stuff as BIG as possible, especially if there is the potential for having to take parts in and out.
Also make the traces on the small stuff NOT as thin as possible, actually make them as thick as reasonably possible.
The trend is to make these boards with thinny thin thin thinny traces because you see it in commercial PCBs where they try to slide traces inbetween things like IC pins. Bad idea for audio, imo.
Imo, FAT traces and BIG pads are good. traces as wide as the pad are fine.
( I prefer a jumper or two over thinny traces )
On the pads, alternately make them oblong rather than just a single pad diameter...
hard to illustrate in ascii, but, single pad = o oblong pad = ooo hope that illustrates the idea...
YMMV.
_-_-bear
Also make the traces on the small stuff NOT as thin as possible, actually make them as thick as reasonably possible.
The trend is to make these boards with thinny thin thin thinny traces because you see it in commercial PCBs where they try to slide traces inbetween things like IC pins. Bad idea for audio, imo.
Imo, FAT traces and BIG pads are good. traces as wide as the pad are fine.
( I prefer a jumper or two over thinny traces )
On the pads, alternately make them oblong rather than just a single pad diameter...
hard to illustrate in ascii, but, single pad = o oblong pad = ooo hope that illustrates the idea...
YMMV.
_-_-bear
What is it with some of you guys and this "gold plated is sooo much better" mentality?
Gold is inferior in conductivity to both copper and silver. Gold rules in corrosion resistance, in harsh environments, and is very well suited for "finger type" and clamp-down type connectors where the gold surface is one of the two contact points. Otherwise gold sucks, and is not a very solderable surface.
Soldered gold pads are usually avoided in high-tech instrumentation.
Gold is inferior in conductivity to both copper and silver. Gold rules in corrosion resistance, in harsh environments, and is very well suited for "finger type" and clamp-down type connectors where the gold surface is one of the two contact points. Otherwise gold sucks, and is not a very solderable surface.
Soldered gold pads are usually avoided in high-tech instrumentation.
Hi mate.
Absolutley agree with you
As evvery time I try to post same definitive numbers I get flied and give a chance to cumpolsory poster to fell superior I was asking questions insted.
There are to many factors involved for straight answer
One need to consider
Normall max listening volume (Search for Pano tread to get Idea)
Impedence of speakers (normaly much worst than rated especialy at low frequency where signall normaly present maxymum energy demand)
Those will be values that aply only to each individual case
In my case I have Polk LS15i nominal 4 homs in my smalish room 87dB is fine after Pano test I will rarely ask the F5 to give me more than 12 rms out
The music I listen to is mostly not Classical so dinamic range is not big issue.
I am quite happy with 35 volts rails and Bias of 2.5 A on each rail
Yess at time the amplifier will go in to class B but I do not think I will be able to distinguish the difference at such level
I still have the option of going balanced for 3dB more head room (Ops there it cames)
This will be different from one having 40 V pluss rails when diodes became needed maybe
Or someone with home made super duper speakers with 90 + dB efficency and impedance curve that gos down to fraction of homs.
The difficulties with gold are ussually the type of plating that is done. There is electrolisis gold and some other palting method(Brain Fart). Personally, I agree with some posters that gold plated PCBs are extra raz-ma-taz that does not improve anything except the vendors pocket! Solderability sucks! (especially electrolsis gold) and it grows wiskers(just pretend your a space satalite) etc.etc.
The most important factor in PCB construction for our audio circuits, other than layout related crosstalk issues, is low DCR and inductance along with the ability to cary away unwanted heat from components! That boils down to wide, 2 3 or 4oz. copper! And, good vias!
just my 2cents
The most important factor in PCB construction for our audio circuits, other than layout related crosstalk issues, is low DCR and inductance along with the ability to cary away unwanted heat from components! That boils down to wide, 2 3 or 4oz. copper! And, good vias!
just my 2cents
I don't see a point using 4 layers in this PCB. The fabrication fees need to concern too.
Actually, and for the record, generally one needs MORE headroom for classical recordings than most other. I'd suggest about 20dB above average listening level, whatever that power level is.
So, if that is 1 watt, then 64 watts.
But that's only if ur average is actually 1 watt!
What if it is 2 watts? Then 128 watts.
For 4 watts? Only 256 watts.
Hmmm...
Well the good news is that most peaks that are 20dB up from average are actually transients, and end up in the tweeters, and with an amp like the F5 tend to have nicely rounded clipped waveforms (as do true triode tube amps) so the effect is more of compression (if heard at all) than hard clipping.
Imo, this is the reason that people say that tube or triode amps "sound" more powerful than they really are...
In the system that I would use an F5 (and I will be, and it will happen) the compression driver on the widerange horn is 109dB/1w/1m rated. So, my average listening level is under 1 watt!! So, my "first watt" is really critical and the headroom is a non-issue. Of course merely having a compression driver does not in and of itself assure that you will like anything about how it sounds, it's more complex than just using a compression driver...
_-_-bear
Exactley the same ting I did say
I do not listen to much classical music so I do not need the head room.
The 20dB is prety much right as evveribody know
I am not realy shure about the tweeter bit especialy when sometink like 1812 is on the spinner but have to agree with the compression part.
Or if you prefer this is another of the many variable involved in the compromises one can choose to make or not to when building the amplifier.
From what I have experienced the big party popper for any amp is clipping pheraps current clipping more than voltagge clipping goes hand in hand with listening fatigue.
Since I had the F5 I find myself listening for longher periods of time and some how the volume keeps going up to the point that I get suprised that I have to shout Just to keep a conversation going.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.