John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
No, that's no helpful info at all. Cu thickness is not determined by country
of origin but by technical necessities. We see here another example of
this audio people "thicker==better" fetish.

Already 70u is seldom used, if at all for real power stages or inner
power planes for multilayers. Yes, I have counterexamples also, like
microwave substrates with 2 mm Cu or Al on the bottom side, but that is
more for thermal reasons and to make up for the bad properties of the
teflon prepregs (flows under pressure, bends easily, traces stick about
as good as an egg in a teflon pan, mechanical TC) or the brittleness
of an Al2O3 substrate. Making quality vias here is artwork.

You cannot mount 0603 class parts on a 100u metal layer. There would
be a lot of underetching; a 5 mil trace would be about as high as wide
nominally. In practice, the acid would eat away most of it from the sides.
The deeper you must etch, the more.

For my own experiments I use presensitized FR4 material from Bungard,
available from SEGOR-electronics GmbH, Berlin.
It is easy to produce clean 5 mil traces or micro-SO8 footprints. The mask is
usually "offset film" , produced by a local printer shop from .pdf.
And I love that double sided 0.5mm FR4 with 17u Cu. Ideal to make
stamp sized miniboards that can be soldered to a larger ground plane
for development. I have a whole library of regulators, oscillators, OpAmp stages etc.

If you lose pads, thicker copper would not hold them. It's underetching,
glue, roughness of the epoxy, not exceeding the fr4 glas temperature,
trying to lift the part b4 everything is liquid, pressure from soldering iron.

The new leadless solders sometimes require high temperatures. I then
add a little bit of low temp. solder, that makes it a lot easier. And do not use
tin pumps, just solder wick to remove the excess.
Also, preheating the board to 100°C on a hot plate helps a lot against
loss of heat via GND layers. And get a hot air solder station.

regards, Gerhard

Gerhard, all other things being equal, thicker is better simply because thicker allows for less stray capacitance.

Of course, we are talking only abut the the foundation, poor layout will still create problems no matter what you use.

For the last 25 years, I have NEVER used any other but German made glass epoxy boards, and my PCB service man hasn't bought any other. I'm happy with them and have no incentive to change.

My soldering irons, both standalone, battery powered and soldering station are all by ERSA, always have been since 1971 to this day. And I do not save on tips, I have a fine selection for various purposes and I do change them when they eventually wear out.

China was mentioned only in ontext of reported but not verified appearence of 255 micron copper tracing. Try working with that if you are a maschist or have no choice, because as Richard says, they tend to come off real easy. But they probably do use it as they are extremely cost-conscious and are always looking for ways to drop the price.
 
AX tech editor
Joined 2002
Paid Member
Accidents happen

The point is not that accidental discoveries are not made. The point is that such an accidental or serendipity only happens to those that know the field, know the details and recognise the lucky shot for what it is. SY's example is a good one: the guy who made the 'accident' happen had no clue - it was the experts in the field that recognised it for what it was.

Jan
 
So what are the leap forward improvements done in audio in the past 25 yrs and if any, why do so many find current Hi-end sound, horrible and over hyped... :xfingers:

I don't know who those 'many' would be, I haven't met them. The major improvement in audio in the past 25 years is that there really aren't that many bad amps around any more.The same goes for the source material. So, the general quality of audio in public places has gone up considerably. Also, the average quality of electronic components for home use has gone up a lot.

This might be why hi-end is suffering. To most consumers, electronics at mid level is good enough and cannot be distinguished from high end. I would even dare to say that you have a better chance to buy a technically sound component from one of the large producers, than from one of the niche marketeers.
 
My little bit of serendipity occurred through focused interest; it wasn't accidental, but was totally unexpected - getting convincing replay was not on my agenda, but it was there waiting, in the wings ... :).

My "discovery" resulted from my natural inclination to get things better, a little bit better, just a bit more - and one day the steady progression was just enough to flip the quality over the subjective hurdle. Others of course have "discovered" it also at times, but nearly always disregarded the significance of it; once experienced, my interest was locked on totally to this behaviour from then on - and remains so still ... ;)
 
BTW, a.wayne, if I try and go to your Public Profile I end up at Bruce Wayne's. Now, I don't think you're that batty, but this is the sort of thing that wore me out in the software game - dumb programming everywhere, and it grinds you down in the end; I haven't got the patience to tolerate all the sillinesses any more ...
 
www.hifisonix.com
Joined 2003
Paid Member
You are very tiring. Do you think this thing was invented by someone who never hunted, never thought about how to kill a prey, never speculated 'if I only could do it from farther away'? Basic R&D my friend.

jan

No I am not tiring Jan. I just challenge statements I don't think make sense, don't stand up to scrutiny.
 
Must be some detail missing here, you can't layout, cut masks and fab from scratch in three weeks . Did they have a prototyping array available that only needed a metal mask?

I remember in the late 70s when that was offered for just a few grand for your first chips. A few hundred for the design kits that contained the basic chip with different varients so you could breadboard your design and test it before they did the final interconnects.

I haven't looked but is this still offered?

I believe the original dbx chips were done this way.

P.P.S. Looking a bit more I see it is quite common for a finished packaged run of 10 pieces a production budget looks like $10,000. However it seems that a descrete design is probably a better approach for anything besides very large quantities.


P.S. I look and saw Mosis is still around doing it, although mostly digital processes.
 
Last edited:
Status
Not open for further replies.