Polyimide non woven aramid would be a perfect match 6-9, most others are between 10-17 depending on the base epoxy.... Glad I don't have to worry about really cold stuff.... Under the hood automotive is one of the biggest challenges for non esoteric design these days, not a nice environment.
A lot of the bottom terminated packages and BGAs are done on a interposter, quite often with a BT resin so are a good match for the base laminate, smaller ceramic parts the distances involved still make for reliable joints, as long as the correct volume of solder is used...
But then most of us don't have to worry about the silly temps you play with...
That's because you are 'ignorant' Marce, of some research made by Tektronix years ago. The basic problem is that the DA of the board material cannot be easily compensated for by simple capacitance compensation, as is normally used in probes, attenuators, etc, and you get problems with square waves, for example.
The problem is real, and often measurable, but ultimately something that should be avoided with the best quality designs.
The problem is real, and often measurable, but ultimately something that should be avoided with the best quality designs.
www.tek.com/dl/LOSS-0601_0.pdf
They needed 36" of FR4 to get a serious effect with 50ps rise-times. An RF problem.
They needed 36" of FR4 to get a serious effect with 50ps rise-times. An RF problem.
As digital electronics further penetrates the consumer market ---
https://www.speedingedge.com/PDF-Files/tutorial.pdf
-RNM
https://www.speedingedge.com/PDF-Files/tutorial.pdf
-RNM
Last edited:
I've never even seen two sided PC boards in the economy class. Just peek inside any Yamaha integrated amp from their line which is below the mddle and what you'll see is a forest of jumpers inside it makes you wonder how is that more cost effective even using robots for soldering, but it seems it is.
With two sided PCBs, I don't even remember when I had to use any jmper at all, anywhere.
From 1999, things changed big time with the move to lead free, laminates have evolved, we are now doing boards with DDR5 running at silly speeds all on FR4....
Here is a good source of material...from the master...
Home
A broader response not aimed at any one person...
I cant understand what the problem is with areas of consumer audio (high end seems to be the culprit) and all this, there is a wealth of information out there, there is a world wide organisation that has all the info you need for your PCB assemblies...
of course us ignoramuses don't have a clue what we are talking about or what we are doing because we aren't doing esoteric audio design, just mundane run of the mill stuff.....
I would think 10G backplanes for Cisco's routers might be a little more demanding than audio.
Dick, what board is in the generator you tweaked to < -130dB?
hook and quality control
The article on hook that JC references I remember well, and I wish I could find my copy. But it was a long time ago, and it's claimed that proper modern processing has eliminated it as a concern.
However, some problem fixes can arise again in the global economy if one is not vigilant. Although there is nothing intrinsically wrong about sourcing materials from the Far East, we all know what can happen if you lose control and fail to monitor your suppliers. Perhaps raw laminates and so forth are pretty well-controlled. One would hope that low cost passives are similarly consistent. But only about twelve years ago a batch of bad dielectric used by an unscrupulous manufacturer of 100nF caps led to a huge problem as they turned into little heating elements and melted the sidewalls of powered speakers, Investigation indicated the supplier had turned the tester off because it enhanced yields!
The article on hook that JC references I remember well, and I wish I could find my copy. But it was a long time ago, and it's claimed that proper modern processing has eliminated it as a concern.
However, some problem fixes can arise again in the global economy if one is not vigilant. Although there is nothing intrinsically wrong about sourcing materials from the Far East, we all know what can happen if you lose control and fail to monitor your suppliers. Perhaps raw laminates and so forth are pretty well-controlled. One would hope that low cost passives are similarly consistent. But only about twelve years ago a batch of bad dielectric used by an unscrupulous manufacturer of 100nF caps led to a huge problem as they turned into little heating elements and melted the sidewalls of powered speakers, Investigation indicated the supplier had turned the tester off because it enhanced yields!
I've occasionally pointed out that driven guard layout can nearly eliminate PCB dielectric concerns at audio for the few nodes high enough Z to matter
in fact someone got a patent on driven guard PCB traces "for audio amplifier"
in fact someone got a patent on driven guard PCB traces "for audio amplifier"
What the TEK article showed was that a WAVEFORM waveshape could be noticeably changed by circuit board material. This is beyond a single tone, but more like what music is actually like. The effect goes down to low frequencies in some instances.
Care to provide an example of a measured audio waveform that is changed going from, say, FR4 to PTFE?
Take a minute and go over to Marsh Headphone Amp to see what Walt Jung is doing and visit his new site.
-RNM
-RNM
I have no idea, really..... does not look like Teflon nor microwave substrate material.
But if your material and or layout causes the materials' DA to show itself as waveform change, then that is not a good thing for audio.
High Z circuits are more vulnerable to DA affects than low Z circuits. So it all 'depends'. The effects at 50 Ohms are minimum but not so at high Z... scope inputs are high Z. Cap mic preamp are high Z, RIAA feedback values might be high Z, filters, etc.
THx-RNMarsh
Last edited:
No, because my oscilloscope waveform is less accurate that what the circuit board material might contribute. But if one wants to see a similar effect, move the comp cap on your oscilloscope probe with a 1KHz square wave (it can be bandwidth limited) and note the change. The addition of the dielectric absorption puts in a SECOND ORDER EFFECT in parallel to the comp cap, and in oscilloscope attenuators, for example, and that is why TEK wrote the article, warning other engineers of the problem. It is NOT necessarily a high frequency effect, it happens at all frequencies.
Of course, those of you who use cheap wire, marginal caps, etc. might dismiss what a board might change. '-)
Of course, those of you who use cheap wire, marginal caps, etc. might dismiss what a board might change. '-)
OK, so the answer is "No, I have no data to show this is relevant to audio."
- Status
- Not open for further replies.