1. Home
  2. Electronics Design
  3. PCB layout for ADC
  4. Page 2

PCB layout for ADC

Can you examplify this?

Guess the best setup for a A/D is to have two seperate circuits for analog and digital with only the A/D chip connecting them through a shielding wall in between. Galvanic isolation with transformers, but then there's the issue of ripple ;)

Reality is a harsh and consistent teacher. So I guess that there will be investment made in this arena ;).. let's see how about a analogue-boom? :)

I wouldn't even call them sw guys ;) They mainly tend to obfuscate information makeing it hard to access by introduceing various extensions not neccesary to communicate information.

Reply to
pbdelete
Loading thread data ...

In article , wrote: [....]

No, you power it with a photocell driven by a laser and use opto-isolators for the digital output.

But seriously, it is often how the op-amp driving the ADC is connected that really matters most:

O---------/\\/\\/--------+------/\\/\\/------- ! ! --!-\\ ! Xducer ! >------------+--- ADC --!+/ ! O--------/\\/\\/---------+------/\\/\\-------- ! ! ! Some people think this ground plane ! -+-----------------------------------------+----------- is all at the same voltage

--
--
kensmith@rahul.net   forging knowledge
Reply to
Ken Smith

Not really because I can't publish client's layouts and schematics.

I found that isolation rarely works. The reason is that there will always be paths where noise makes it across. ADCs are never really autonomous, at some point there will be a common power source between front and back (latest at the battery or the wall outlet), digital controls, probes that are being grabbed and thus capacitively coupled to by humans or machines etc. All this creates loops.

If you keep the 5V to the analog section clean and leave the 5V to the digital section as "whatever it is" chances for a clean conversion are slim. An ADC is small, it's all on one chip. Even if it was internally completely separated between analog and digital (typically they aren't) the impedance of the bond wires alone makes for a nice "noise voltage divider" smack dab into the heart of the chip.

--
Regards, Joerg

http://www.analogconsultants.com
Reply to
Joerg

Hello Jeff,

http://66.102.9.104/search?q=cache:ZV9hu2nZK5IJ:

formatting link
*-*-*-smartest-*-in-the-room+artificial-shortages

But why was that possible? Because there were no long term contracts to speak of. That was the cardinal mistake. If you don't have long term contracts and hedges in place anybody can have you over the barrel. Not every company will take advantage of that but some (not only Enron) did.

One problem is that too many government decision makers are lacking business experience. Even in the rather small division that I was leading we made 100% certain that we had long term delivery contracts for crucial parts of our business (in our case chips, PZT transducers etc.). It blew my mind when it was reported in the press that our electricity grid was running almost without that.

--
Regards, Joerg

http://www.analogconsultants.com
Reply to
Joerg

I was thinking about that recently. It would be very handy if I could power a box over optical fiber, as well as communicate with it; separate fibers would be fine. But it gets awkward to couple more than

5 or maybe 10 mW into a regular 62/125 micron fiber, and the photocell will be maybe 10% efficient, so there's not a lot of power available at the end. Is there anything better around?

John

Reply to
John Larkin

In article , John Larkin wrote: [....]

You can use a fiber bundle or a "light pipe".

Endoscopes get the light into the "victim" somehow.

I had to solve the problem of bringing the light from an LED about 6 inches to the outside world. A 1/4" plastic rod does it quite nicely. If you don't need things to flex, you could try that.

--
--
kensmith@rahul.net   forging knowledge
Reply to
Ken Smith

Hello John,

The only thing that comes to mind (and is in production since over a decade now) is magnetic coupling. You can bridge large gaps, especially when adhering to the series resonant concept.

--
Regards, Joerg

http://www.analogconsultants.com
Reply to
Joerg

JeffM wrote: > >>>Well, at least our mains supply remained nearly constant > >>>after the last governor was kicked out :-) > >>> Joerg > >>

http://66.102.9.104/search?q=cache:ZV9hu2nZK5IJ:

formatting link
*-*-*-smartest-*-in-the-room+artificial-shortages

Jeff, with long term contracts, Enron would be required to provide the agreed to power at the contract price. If they shut down their generators, they would have to buy power on the spot market to make up the shortfall. They would LOSE money by dicking around so there would be every incentive to keep the system up and running. I believe that's what Joerg meant.

GG

Reply to
Glenn Gundlach

Hello Glenn,

Exactly. Except I'd have said "They would have lost money ..." because those long term contracts were not in place. Every properly run business has these, including penalties for failure to deliver and whatnot. Yet the politician in those days took it upon themselves to not consider that proper. With disastrous results. IIRC Andy Grove from Intel went straight to the governor and told him in no uncertain terms that if there'd be one more outage there would be zero job creation by the company and that there would be lots of other places to conduct business.

--
Regards, Joerg

http://www.analogconsultants.com
Reply to
Joerg

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.