RF inductors in OrCad

[Joerg]

Hello Joel,

The Pulsonix software I use has interactive spiral design support:

I've often wondered how many people actually use (meaning "purchase!") that
extra option. I suppose that if you're designing a spiral inductor for, e.g.,
a switch-mode power supply ...

Spiral inductors for switchers? Wow. Did you snap the core in through
cutouts in the PCB? Must have been a multi-MHz switcher. Now I am
feeling old...

--
Regards, Joerg

http://www.analogconsultants.com

 

[Joel Kolstad]

Hi Joerg,

"Joerg" <notthisjoergsch@removethispacbell.net> wrote in message
news:hx6%g.19594$6S3.16815@newssvr25.news.prodigy.net...

It is amazing what kinds of design tools we have these days. In fact, I
didn't have to pull my old Bronstein/Semendjajew math dictionary off the
shelf for over a month now. However, when it comes to really intricate stuff
there is not much progress.

In general I think you're correct, although even with the intricate stuff it's
a lot easier just by virtue of having tools like Matlab (or Scilab) around.

When doing wave digital filters I still have to resort to DOS tools.

With DSP filters taking over most applications these days, wave digital and
fancier passive and active filters definitely are something of a dying breed.

It would be smart for a company like TI or one of the other microcontroller
vendors to fund, e.g., some university to modernize those wave digital filter
tools you're using. I can't imagine it'd take more than one or two students
over the span of a year, so they'd be looking at less than the price of one
regular engineer.

BTW, did you find other good candidates for the RF engineer opening?

No, and I've been gone for the past couple of weeks, so I haven't been looking
(again).

---Joel

 

[Joel Kolstad]

Hi Joerg,

"Joerg" <notthisjoergsch@removethispacbell.net> wrote in message
news:Tz6%g.19600$6S3.12515@newssvr25.news.prodigy.net...

Spiral inductors for switchers? Wow. Did you snap the core in through
cutouts in the PCB? Must have been a multi-MHz switcher. Now I am feeling
old...

I haven't designed one like that myself (the last one I did used toroids with
Kool-Mu material -- traditional with a bit of high-tech thrown in), but I have
seen some built that way... the TI switcher "modules" (complete siwtching
power supply in a package, some even with, e.g., 7805 compatible pinouts and
roughly the size of a 7805+heat sink clipped on) that they acquired from
PowerTrends do it that way.

 

[Joerg]

Hello Joel,

"Joerg" <notthisjoergsch@removethispacbell.net> wrote in message
news:hx6%g.19594$6S3.16815@newssvr25.news.prodigy.net...

It is amazing what kinds of design tools we have these days. In fact, I
didn't have to pull my old Bronstein/Semendjajew math dictionary off the
shelf for over a month now. However, when it comes to really intricate stuff
there is not much progress.


In general I think you're correct, although even with the intricate stuff it's
a lot easier just by virtue of having tools like Matlab (or Scilab) around.

I wasn't enthused about those tools. Mathworks changes versions a lot
and then things become incompatible. Scilab and Octave are rather
incompatible with Matlab.

When doing wave digital filters I still have to resort to DOS tools.


With DSP filters taking over most applications these days, wave digital and
fancier passive and active filters definitely are something of a dying breed.

Sure but I have yet to see a micropower DSP. You can't use them on
anything that has to run off a couple of AA cells for weeks. Also, they
are above $4, usually.

It would be smart for a company like TI or one of the other microcontroller
vendors to fund, e.g., some university to modernize those wave digital filter
tools you're using. I can't imagine it'd take more than one or two students
over the span of a year, so they'd be looking at less than the price of one
regular engineer.

I suggested it to them more than once. Didn't happen, except that they
came out with a new app note where the older tool set has been compiled
into a single executable. But it's not interactive at all. Also, you
really have to know what you are doing. Mistakes won't cause any error
message, they will result in erroneous data.

It would be the perfect win-win situation for TI and a few interns. Even
in my days it was sometimes tough to find a good intern job. With good I
mean being able to create something that mankind really needs, not just
some scientific fluff.

--
Regards, Joerg

http://www.analogconsultants.com

 

[Joerg]

Hello Joel,

Spiral inductors for switchers? Wow. Did you snap the core in through
cutouts in the PCB? Must have been a multi-MHz switcher. Now I am feeling
old...


I haven't designed one like that myself (the last one I did used toroids with
Kool-Mu material -- traditional with a bit of high-tech thrown in), but I have
seen some built that way... the TI switcher "modules" (complete siwtching
power supply in a package, some even with, e.g., 7805 compatible pinouts and
roughly the size of a 7805+heat sink clipped on) that they acquired from
PowerTrends do it that way.

Thanks, have to take a closer look at those. I haven't used any switcher
modules lately because they couldn't compete in cost versus the usual
"roll your own".

--
Regards, Joerg

http://www.analogconsultants.com

 

[Joel Kolstad]

"Joerg" <notthisjoergsch@removethispacbell.net> wrote in message
news:wo8%g.19639$6S3.5300@newssvr25.news.prodigy.net...

Thanks, have to take a closer look at those. I haven't used any switcher
modules lately because they couldn't compete in cost versus the usual "roll
your own".

The board I used one of those modules on had a $600 FPGA on it, so no one much
cared about the price of the regulator.