From: Spehro Pefhany on
On Wed, 11 Nov 2009 11:07:09 -0500, Phil Hobbs
<pcdhSpamMeSenseless(a)electrooptical.net> wrote:

>George Herold wrote:
>> On Nov 11, 10:19 am, MooseFET <kensm...(a)rahul.net> wrote:
>>> On Nov 11, 5:40 am, Phil Hobbs<pcdhSpamMeSensel...(a)electrooptical.net> wrote:
>>>
>>> [....]
>>>
>>>> LTSpice is free, and so is Gnu Octave. Why doesn't the budding designer
>>>> download a copy of each?
>>> Yes to both of those.
>>>
>>> Octave is a great language for quickly coding up the math of
>>> something.
>>>
>>> Also learn a little about gnuplot. It is great for making graphs of
>>> things you can't do with a spread sheet.
>>>
>>> Also you want to have a copy of the latest OpenOffice spreadsheet. It
>>> can do really big ones and will import ASCII.
>>
>> "> Octave is a great language for quickly coding up the math of
>>> something."
>>
>> Cool, I never heard of Octave. Steep learning curve? I've never used
>> Matlab or similar software. When I need math functions more
>> complicated than my calculator I fire up an old version of QuickBasic
>> (4.5?) under DOS. I did a little bit of C coding back in the 80's,
>> but never needed the speed, (or the slows.. I found it easier to debug
>> the basic code.)
>>
>> George H.
>
>Octave is an open-source Matlab clone, originally developed as
>courseware. It has a pretty good user base and reasonable developer
>support. It can run most Matlab M-files unaltered. Earlier versions
>had trouble with plotting, but the latest ones are quite good, though
>not as good as Matlab. (On the other hand, they don't cost $2k.)
>
>Octave/Matlab are matrix-oriented, so they work more efficiently with
>vector and matrix operations (not necessarily linear ones). They do
>have loops and so on, but they run at QuickBasic type speeds instead of
>Matlab type speeds. Octave/Matlab are nearly as fast as compiled code
>for vectorish stuff.
>
>Matlab is great if you need the more advanced extensions, but paying $2k
>for the basic program is completely unnecessary otherwise.
>
>Cheers
>
>Phil Hobbs

There's also Scilab. http://www.scilab.org/ which might be easier for
Windows-based users.

From: D from BC on
On Wed, 11 Nov 2009 12:13:49 -0000, "Phil O. Sopher"
<invalid(a)invalid.invalid> wrote:

>Over a 40 year period of interest, I've gathered some
>test equipment, AF Genny, RF Genny, Freq Counter, High
>Impedance Voltmeter, Wheatstone Bridge, Oscilloscope.
>
>None of these is particularly small and all are at least a 6" cube.
>
>It seems to me that the same functionality could be achieved
>these days with perhaps just a few inches of bench space, but,
>would it be of any use?
>
>The eqpt I gathered together dates very much from the days of
>designing circuitry with individual componenents (R, L, C, BJT)
>and offers test and validation at that level, but nowadays we don't
>work at that level (even op amps have been around for that 40 year
>period).
>
>So, what do you have on your test bench these days, how big is it,
>did you design it yourself, and what would you recommend to the
>budding circuit designer of today who isn't in the industry and therefore
>does not have access to Spice or Matlab to validate their designs?
>
>

If I tell should I install extra locks on the door afterward? :P

D from BC
Amateur smps designer
British Columbia, Canada
Posted to sci.electronics.design
From: a7yvm109gf5d1 on
On Nov 11, 10:19 am, MooseFET <kensm...(a)rahul.net> wrote:
>
> Also you want to have a copy of the latest OpenOffice spreadsheet. It
> can do really big ones and will import ASCII.

While OO is great, it also has bloat and annoyances (UI gets confused
with multiple monitors).
You can get Gnumeric as a great stand alone spreadsheet.
From: MooseFET on
On Nov 11, 7:58 am, George Herold <ggher...(a)gmail.com> wrote:
> On Nov 11, 10:19 am, MooseFET <kensm...(a)rahul.net> wrote:
>
>
>
> > On Nov 11, 5:40 am, Phil Hobbs<pcdhSpamMeSensel...(a)electrooptical.net> wrote:
>
> > [....]
>
> > > LTSpice is free, and so is Gnu Octave.  Why doesn't the budding designer
> > > download a copy of each?
>
> > Yes to both of those.
>
> > Octave is a great language for quickly coding up the math of
> > something.
>
> > Also learn a little about gnuplot.  It is great for making graphs of
> > things you can't do with a spread sheet.
>
> > Also you want to have a copy of the latest OpenOffice spreadsheet.  It
> > can do really big ones and will import ASCII.
>
> "> Octave is a great language for quickly coding up the math of
>
> > something."
>
> Cool, I never heard of Octave.  Steep learning curve?

The learning curve isn't too bad. If you can do basic, you should
be able to learn to use Octave.

Octave defaults to printing the results unless you put a semicolon on
the end.
Here's a plot of sin()


octave:1> X=1:0.1:6 # Make an array of 1 to 6 in 0.1 size steps
octave:2> Y=sin(X) # Find the sin() of each value
octave:3> plot(X,Y) # Make a pretty plot

Functions are made like this:

octave:7> function Z=DoThing(A,B)
> Z=A+B
> endfunction

A B and Z are all matrix values so DoThing would add two matrixes.





 I've never used
> Matlab or similar software.  When I need math functions more
> complicated than my calculator I fire up an old version of QuickBasic
> (4.5?) under DOS.  I did a little bit of C coding back in the 80's,
> but never needed the speed, (or the slows.. I found it easier to debug
> the basic code.)
>
> George H.

From: MooseFET on
On Nov 11, 7:46 am, Fred Bartoli <" "> wrote:
> MooseFET a écrit :
>
>
>
> > On Nov 11, 4:13 am, "Phil O. Sopher" <inva...(a)invalid.invalid> wrote:
> >> Over a 40 year period of interest, I've gathered some
> >> test equipment, AF Genny, RF Genny, Freq Counter, High
> >> Impedance Voltmeter, Wheatstone Bridge, Oscilloscope.
>
> >> None of these is particularly small and all are at least a 6" cube.
>
> >> It seems to me that the same functionality could be achieved
> >> these days with perhaps just a few inches of bench space, but,
> >> would it be of any use?
>
> >> The eqpt I gathered together dates very much from the days of
> >> designing circuitry with individual componenents (R, L, C, BJT)
> >> and offers test and validation at that level, but nowadays we don't
> >> work at that level (even op amps have been around for that 40 year
> >> period).
>
> >> So, what do you have on your test bench these days, how big is it,
> >> did you design it yourself, and what would you recommend to the
> >> budding circuit designer of today who isn't in the industry and therefore
> >> does not have access to Spice or Matlab to validate their designs?
>
> > Do we also count the floor near the test bench?
>
> > On the bench I have:
>
> > A Tek digital phosphor scope and an old-old Philips scope that is only
> > good for 25 MHz.  Both are connected to the system being developed.  I
> > need to watch signals in two unrelated time domains.
>
> > There is a Fluke 45(IIRC) DVM with RS-232 output that I can record on
> > the PC.  I need to monitor a DC voltage over the time frame of hours
> > and I don't want to do all that writing.
>
> > The PC has 4 RS-232 ports.  Two of them run to the system. (Not
> > counting the Fluke)
>
> > Beside the PC is a laptop that is currently not in use but its serial
> > port is the reason it is sitting there.
>
> > The power supply is a B&K.
>
> > On the floor is a shield can the size of a modest water heater.
>
> > Filed away in a cabinet is a bunch of stuff but one that comes to mind
> > is a calibrated hair drier.  The temperature of the air it puts out is
> > just about exactly 60C.
>
> What? It's not calibrated in Fahrenheits?

No strangely enough the 45C and 60C temperatures needed to be reported
in C degrees.

We do have some tape measures in decimal feet just in case you need to
measure off 63.5 feet of #40 copper wire.