R-L-C structures on PCB

Just wondering why? SMD R & C are so cheap and light weight. PCB area is more expensive than the parts.

Ansoft Designer might do what you want:

Leon

Hi,

recently I was assigned with the job to design a Printed Circuit Board (PCB) which does a kind of analog signal formatting through on-board designed capacitances, inductances and resistances. These components must be designed not as discrete components (SMD, etc.) but deliberately as stray capacitances in the form of specially layouted transmission line segments and 3D arrangements. The material of the PCB is predefined to be FR4. Surely I can design the circuit with tools like PSpice etc. but when it comes to design the calculated R, L, C values on the PCB I'm out of luck with my tools. So my question is: Is there a tool (Mentor?) which contains both circuit simulation AND PCB layout where I can design my defined capacitances, inductances and resistances in the form of specially routed transmission lines? What would be the best way to perform that task, anyway? In my opinion it must be an iterative process. Has anyone experience in this particular application field? Any help, suggestions and examples (maybe) would be appreciated.

Many thanks in advance and best regards,

Melanie

Hello Melanie,

up to 2 or 3 GHz it's possible to compute by hand. It works good if i use a thickness of PCB of 1 mm.

Computation-formulas i've found in ISBN 3-723-6545-0, but only in german language.

regards

baumann engineering

Hi Melanie,

Many years ago I designed a few CAT V components with such printed capacitors. I tested two types:

1. Two copper planes (SMD-pads from stack;-) on each side of the board (we used only doublesided FR4 1.6 mm)
2. traces with vias in a matrix

---O O O O O X X X X ...

---O O O O O X X X X ... O O O O O

Pretty much all the "big boys" of RF design software do this: They contain a frequency domain simulator that has some reasonably good models of microstrips, spiral inductors, etc. and easily link up with a field solver for when you need a more detailed answer. I.e., Agilent's ADS does this, as does Eagleware Genesys, Sonnet, Ansoft's RF Designer, etc.

Since you're just starting out, I'd suggest you go over to Sonnet and download a copy of Sonnet Lite for free

I believe they're the only one of the companies listed above that has ANY free offerings. Agilent and Ansoft, however, will hook you up for next to nothing if you're part of a Real University and either are or can get a professor to contact the appropriate people. Eagleware isn't quite as generous in my (limited) experience; I don't know about the others.

If you're buying any of these packages for "professional" usage they all come in a dizzying number of different configurations and start with 4 digit price tags that can rapidly creep up to 5.

One final note: Although all these RF simulators will let you do "PCB layout," they're really not optimized for the generic layout task. What most people do is use the RF simulators to design just the, uh, critical RF portions and then export the copper layers into a more conventional PCB package such as PADS, Protel, etc. Most RF things these days hit a bunch of digital circuitry sooner or latter, and you really don't want to use ADS to layout some 676 ball FPGA with hundreds of digital I/O lines coming off of it. :-)

---Joel Kolstad

Hello Melanie,

Just FWIW: R will be a problem here. For controlled resistor values you would have to deposit carbon and laser trim. I have never done that on FR4 but a lot on alumina (thick film hybrid).

I still do that by hand. Sometimes I even use the old slide rule calculator.

Regards, Joerg

Puff is an old, free (not sure if it's still free) program that lets you simulate planar transmission-line-based filters

And there's also Sonnet Lite, the free version of the Sonnet e-m simulator.

A bit of googling should turn up the basics of planar transmission-line-based filters.

John

You might find something of interest on this site:

HP (Agilent) has a suite for RF design. I don't know if this will cover your needs.

--
Mark

No, I can't afford Mentor products. They did buy out PADS, which we use, and I'm wondering what they'll do with that.

John

You mean the Autocad/GenericCad maneuver?

John

Hi Marte,

thanks for your response. Some questions still:

How did you do this? I mean, if I calculated the needed capacitance, how would I get the corresponding PCB layout for that?

Again, what was your approach to model the layout in maths? Sorry, if I bore you but I never did it before and thus I have no experience on that topic. For a seminar work at university I have to suggest a compensation network on PCB for a certain case. I don't want any of you to do my work, I just want to comprehend how to get there. :-)

Bye Mel

They'll give it the same treatment that Cadence gave PSpice ;-)

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
|  Analog/Mixed-Signal ASIC\'s and Discrete Systems  |    manus    |
|  Phoenix, Arizona            Voice:(480)460-2350  |             |
|  E-mail Address at Website     Fax:(480)460-2142  |  Brass Rat  |
|       http://www.analog-innovations.com           |    1962     |
             
"Winners never quit, quitters never win", Jack Bradley Budnik ~1956

I thought about Mentor Hyperlynx since the university has some Montor licenses but I am not quite sure if this tool is exactly what I need? Any suggestions / experiences?

Bye Mel

"John Larkin" schrieb im Newsbeitrag news: snipped-for-privacy@4ax.com...

"Melanie Nasic" schrieb im Newsbeitrag news:do94fl$3kv$ snipped-for-privacy@mamenchi.zrz.TU-Berlin.DE...

(PCB)

designed

anyway?

this

Hello Melanie,

I am sure that nobody wants to make R with the PCB-Traces. It may be Z(transmission line impedance) but never R. L and C can be stubs for example or other odd shapes along the "main" trace.

Please give a clear example(with numbers) what you have to design. All the postings so far are pure speculations about your real task. It's your turn to make the picture more clear.

Best regard, Helmut

In article , Helmut Sennewald wrote: [....]

I'm sure that isn't true.

--
--
kensmith@rahul.net   forging knowledge

Reg Edwards keeps some useful utilities for this kind of stuff on his website. Check out this page and see if it's covered...

--

"What is now proved was once only imagin\'d" - William Blake

It sounds totally impractical to me, at any rate, to try to make R from copper cladding. Anyway, I don't think the OP has thus far actually even told us the frequency of interest!

--

"What is now proved was once only imagin\'d" - William Blake

"Ken Smith" schrieb im Newsbeitrag news:dp463q$3e2$ snipped-for-privacy@blue.rahul.net...

Hello Ken,

I am interested if there are other applications. On what application do you think except current sensing?

Best regards, Helmut