Jneutron, I offered you a fax because it is available from me. I have a brand new HP scanner/fax/printer, but only the fax is operating at this time. Some people are more flexible. Scott's graph is adequate.
Relative to jfet noise, I doubt that you are more knowledgeable than the average engineer.
Relative to jfet noise, I doubt that you are more knowledgeable than the average engineer.
Noise quote related to the graph that I cited:
"As gold is used in the fabrication of commercial JFETS, it is likely that some gold is present in the silicon host lattice of such devices. Now gold in silicon is a well-known amphoteric center, with an acceptor level very close to the intrinsic Fermi level at the middle of the bandgap and a donor level about 0.4eV above the valence band edge. Haslett and Kendall suggest that the gold acceptor level is responsible for the peak in the noise at temperatures in the vicinity of 300K."
Well, that explains it everybody. No QM here! '-)
"As gold is used in the fabrication of commercial JFETS, it is likely that some gold is present in the silicon host lattice of such devices. Now gold in silicon is a well-known amphoteric center, with an acceptor level very close to the intrinsic Fermi level at the middle of the bandgap and a donor level about 0.4eV above the valence band edge. Haslett and Kendall suggest that the gold acceptor level is responsible for the peak in the noise at temperatures in the vicinity of 300K."
Well, that explains it everybody. No QM here! '-)
Thanks for the link, very interesting pictures. Looks like the boards were made on a prototype oriented cnc type router that was popular for lab use at the time. I cut my teeth on Orcad in the early 90's, then Protel...
Mike
I think 6 to 8 layer PCB's handling signals at upwards of a few hundred MHZ with single digit ns rise times are impressive and require real engineering skills. And especially if you are going to make millions and they all work, every time. At audio frequencies, and the associated circuit simplicities, the required skill levels don't even come close.
So? How many audio boards go through the trouble of making 100% double groundplane's for example? Of course, Carl uses other approaches today for surface mount, and high speed digital designs, but they are no concern here.
What we tried for is not obvious to the casual observer, but I doubt we have ever been matched in getting the maximum audio quality from these layouts.
What we tried for is not obvious to the casual observer, but I doubt we have ever been matched in getting the maximum audio quality from these layouts.
In my day job, my boards are implemented in 20 layers of Panasonic Megtron 6, and the traces must carry upwards of 20 GHz.
Rogers dielectric is the old standby, but it does not hold up well to the higher temperatures of 6/6 ROHS reflow. The relatively new Megtron 6 is used in most 40 Gb/s and 100 Gb/s system designs and is just as good as Rogers.
BTW, calculate the skin depth at 20 GHz and you get an amazingly small number.
Cheers,
Bob
20 layers!!!! 20 layers!!!!
Believe it or not there are still some fabricators who only do one!
20 layers!!!! 20 layers!!!!
I newer needed more than 2 for audio devices (analog, of course!)
Well, everyone, I am sorry to bore you with the layout of a successful product that was designed about 28 years ago. Of course, our layouts are somewhat more complex now, including added surface mount and tighter fit, but nothing has worked better than these original boards in my experience, for audio reproduction.
Why people take cheap shots at my former business partner's work is beyond me, since it wins awards and listening contests. Actually, more novice people could learn a trick or two in improving their next layout by studying these boards a little bit.
A worst case example of amateur layout is the original JC-80. What a mess! And trouble too.
Why people take cheap shots at my former business partner's work is beyond me, since it wins awards and listening contests. Actually, more novice people could learn a trick or two in improving their next layout by studying these boards a little bit.
A worst case example of amateur layout is the original JC-80. What a mess! And trouble too.
Actually, if novice people don't start experimenting understanding that traces have resistances, inductions, capacitances between them, they will learn nothing except how to participate in great marketing happening offering something different made using unobtanium computers.
Are you again attempting to insult me, Wavebourn? It is not working.
For everyone else, I don't know much about circuit board layout. I have only done a few layouts, myself over the decades. I don't know the in's and out's of perfect soldering, and my soldering stations are usually Weller, and not Metcal or Pace. I brought up the Vendetta layout, for the first time, as an early example (28 years old) of how we do things. I just talked to Carl Thompson about boards. He was doing 25 layer boards for video broadcast 24 years ago. He could care less what you people think here, "waste of time' he says.
The unfortunate thing is we are now not easily able to converse about the good and bad points of groundplanes, different circuit board material, or using gold, tin, or solder-mask. Oh well, you guys know better.
For everyone else, I don't know much about circuit board layout. I have only done a few layouts, myself over the decades. I don't know the in's and out's of perfect soldering, and my soldering stations are usually Weller, and not Metcal or Pace. I brought up the Vendetta layout, for the first time, as an early example (28 years old) of how we do things. I just talked to Carl Thompson about boards. He was doing 25 layer boards for video broadcast 24 years ago. He could care less what you people think here, "waste of time' he says.
The unfortunate thing is we are now not easily able to converse about the good and bad points of groundplanes, different circuit board material, or using gold, tin, or solder-mask. Oh well, you guys know better.
I never do that John, you know that. But I agree with Pavel, layout looks quite amateurish. Despite of ground planes and 1 million dollar computer that was used, according to your clarification.
Good layout always looks nice in terms of current paths, capacitive couplings and heat dissipation,, mechanical resonances. Components are placed accordingly, traces are made accordingly.
When the device has a single PCB it's size is enough to satisfy requirements of particular functions. If the PCB is one of many standard size PCBs in the device, it is a different matter. If the PCB has components that need to be placed according to rear, or front panel controls and indicators, it may change optimal positioning of them on the PCB. But I don't see any logic on that picture, except to do something that does not look like the rest of PCBs in competing devices.
- Status
- Not open for further replies.