Powered by Invision Power Board


Forum Rules Forum Rules (Please read before posting)
  Reply to this topicStart new topicStart Poll

> Best Choices For Schematic Capture And Pcb, Your opinon please
phin
Posted: December 20, 2016 12:14 am
Reply to this postQuote Post


Newbie
**

Group: Members+
Posts: 11
Member No.: 40,315
Joined: December 19, 2016




Hi All,

I am new to your forum, but an experienced EE now retired. I looked around and did not see any current info on my topic in your forums. Please forgive me if I missed something and direct me to the requested information.

I would lie to do some hobby type electronic design work using schematic capture and PCB layout software. I have read a little on my choices for free and low cost software, but cannot determine which path is best for me. I would like your input please in making a selection for capture and pcb layout. This could be one or two packages, free or low cost.

My designs will be fairly low tech. no boards over 6 inches square. I am not planning to do any RF work, but who knows what may come? I will probably use some simple CPU and/or FPGAs. I may need to use surface mount, but only if needed. Hand assembly only. I would like to be able to do simulation/spice type testing, but that would be new to me. I run win 10.

Thanks in advance

This post has been edited by phin on December 20, 2016 12:16 am
PMEmail Poster
Top
Sch3mat1c
Posted: December 20, 2016 09:25 am
Reply to this postQuote Post


Forum Addict ++
Group Icon

Group: Moderators
Posts: 20,538
Member No.: 73
Joined: July 24, 2002




Probably:
1. Eagle: free, size limited, quirky to use.
2. KiCAD: free-as-in-GPL open source, unlimited, under active development. Maybe quirky, at at times buggy? (I've not used it, personally. I've seen excellent work done in it, and I've seen it being used without too much weirdness. Seems pretty good as free goes.)
3. Circuit Studio, Altium's entry to the free market (normally, Altium Designer is in the $8k range -- CS is stripped down). No size limits, but tied to the cloud, unless you pay for a professional license (quite a lot cheaper than AD and other enterprise-level software, mind).

And a couple of other things, that range from toys (Fritzing?) to weirdness (gEDA?) to more or less practical things.

Tim


--------------------
Answering questions is a tricky subject to practice. Not due to the difficulty of formulating or locating answers, but due to the human inability of asking the right questions; a skill that, were one to possess, would put them in the "answering" category.
PMEmail PosterUsers Website
Top
phin
Posted: December 20, 2016 11:52 pm
Reply to this postQuote Post


Newbie
**

Group: Members+
Posts: 11
Member No.: 40,315
Joined: December 19, 2016




Thanks Tim

From what I have read KICAD does seem to be my best choice. The weaknesses of EAGLE keep popping up in postings that I have seen.

Does anyone know how useful KICAD is for simulation? Nothing real fancy, but it just makes sense to simulate rather than breadboard and I don't know how well it does in that area. I'm talking about basic digital and basic analog. Nothing real complex.

Thanks
PMEmail Poster
Top
Sch3mat1c
Posted: December 22, 2016 02:00 am
Reply to this postQuote Post


Forum Addict ++
Group Icon

Group: Moderators
Posts: 20,538
Member No.: 73
Joined: July 24, 2002




Actually, SPICE integration is new in the latest version, I think?

If you like to stay in the FOSS family, something like NgSpice would work. I guess they recommend it?
http://mithatkonar.com/wiki/doku.php/kicad...ice_quick_guide
There are a few others as well; including the original Berkeley 3f5 and (Georgia Tech) XSPICE. Though I think those are command-line-only tools, so you'd probably want to integrate them somehow, rather than driving them by hand. tongue.gif

Not sure if you'll have the same experience with KiCAD and supporting tools, as, say, LTSpice IV (which has schematic entry, if a rather odd one that takes some getting used to), or any of the pro ones. Meaning, there may be more steps from schematic to sim results, maybe including command line options as well.

I don't know if LTSpice and KiCAD schematics are interoperable. I would be kind of surprised if no one wrote a conversion script or something; that way you can at least have one resource driving the next, and so on (e.g., LTSpice model and schematic is the original source; import to SCH for documentation; transfer to PCB for fabrication).

Tim


--------------------
Answering questions is a tricky subject to practice. Not due to the difficulty of formulating or locating answers, but due to the human inability of asking the right questions; a skill that, were one to possess, would put them in the "answering" category.
PMEmail PosterUsers Website
Top
mrk
Posted: December 22, 2016 03:58 am
Reply to this postQuote Post


Jr. Member
***

Group: Trusted Members
Posts: 59
Member No.: 38,803
Joined: January 25, 2014




I like kicad, I'm mostly used to it now but all the cad programs have a bit of a curve to them... PITA the first while.
PMEmail Poster
Top
phin
Posted: December 22, 2016 04:37 am
Reply to this postQuote Post


Newbie
**

Group: Members+
Posts: 11
Member No.: 40,315
Joined: December 19, 2016




QUOTE (Sch3mat1c @ December 21, 2016 08:00 pm)
Actually, SPICE integration is new in the latest version, I think?

So from this I can assume that simulation is not integral to either Eagle or Kicad? I guess if I am serious about using simulation, I should dig deeper into the ability of both packages to interact with the possible choices of simulators and how "closed-loop" the relation is (similar to reverse annotation)

I am asking about this for a basic reason: it is my "assumption" that we are beyond the point of doing truth tables on the back of a napkin. Debugging after the fact seems to be a bad approach with a current day piece of electronics.

Am I thinking this out correctly? It just seems to me that design without simulation would be just a bad way to do a project.

I did this stuff back in the 70's with lots of paper and pencils. Timing diagrams and fan-outs till I was blue. I am guessing this is now all done in simulation. So selecting a capture package really requires the selection of a simulator too?

Am I heading the right way? If not please set me straight.



PMEmail Poster
Top
Sch3mat1c
Posted: December 22, 2016 11:12 am
Reply to this postQuote Post


Forum Addict ++
Group Icon

Group: Moderators
Posts: 20,538
Member No.: 73
Joined: July 24, 2002




SPICE level simulation is pretty useless these days.

No one designs with transistors anymore. There's an IC for everything. (And the few ICs that do have models, are almost exclusively locked into proprietary-free, or enterprise grade, simulators, not generic SPICE compatible.)

On the few occasions I've reasonably been able to use an analog circuit, I've typically been met with hopeful confusion. Hopeful because I'm that good and they have confidence in my ability, but confusion because, not knowing better, everyone treats it as a black blob that doesn't have a datasheet.

No one designs with discrete gates anymore. There's an MCU or FPGA for everything.

If you want to resolve digital logic, you're welcome to do so, and you're better off doing it, rigorously, on paper, understanding all the truth tables and propagation delays and race conditions, if the circuit is small enough that you can still understand it that way (i.e., the design doesn't span a pile of sheets).

Otherwise, learn a Hardware Description Language, like VHDL or Verilog. You can write down combinatorial logic much faster and much more accurately than by hand, with truth tables, and the synthesizer handles optimization and implementation for you. And when it comes to propagation delays and race conditions, you can make just as many mistakes, just as quickly, if you're ignorant of the process either way. wink.gif But you can edit, recompile and test the design within seconds, not minutes, no physical circuit changes needed.

The number of gate ICs that's worth rolling into a programmable array has always been decreasing over time:
Back in the early days, you might've only used a PAL where it was convenient, and never where time or power consumption was critical: the propagation delays stink and the EPROM cell design sucked many mA of supply current.
When they were introduced, FPGAs were expensive, but worthwhile for moderate speed, large scale, flexible logic designs: something you might need for a video graphics controller, or a signal processor.
The highest performance applications had always required high speed discrete gates (74F or 74LVC, ECL), or custom ASICs, or full-custom bare silicon designs.

All of these are still true today, but the thresholds are lower than ever.
The cost, and PCB area, of about four gate ICs, is about where I would seriously consider using a programmable device over discrete logic.

Lattice FPGAs I think are the cheapest and smallest, in the buck-each range. You're doing yourself a massive disservice if you remain ignorant of these things!


Or, use an MCU. Logic functions can be implemented in software, at relatively large expense in propagation delay. (CPUs are all clocked, so state changes only follow clock edges; and at that, calculating a state change might require dozens of instructions.) Where MCUs really shine is complexity of the operation: with arbitrary access to a relatively large amount of memory (just a dozen bytes worth of memory, is already enough that, the CPU could never possibly increment through all accessible states, over the age of the universe!), and arbitrary looping and recursion possible, you can solve literally any problem -- given enough time and memory.

Tim


--------------------
Answering questions is a tricky subject to practice. Not due to the difficulty of formulating or locating answers, but due to the human inability of asking the right questions; a skill that, were one to possess, would put them in the "answering" category.
PMEmail PosterUsers Website
Top
phin
Posted: December 22, 2016 10:14 pm
Reply to this postQuote Post


Newbie
**

Group: Members+
Posts: 11
Member No.: 40,315
Joined: December 19, 2016




Thank you VERY much Tim.

I will chew on that excellent advice. I had been confused on the FPGA as I had forgotten that the logic you develop for them is done in their SW and not gate by gate. This had been my concern.

From what you have said,on the analog side, I am better to breadboard then to simulate. The is certainly a shorter learning curve and my analog portions should be fairly simple.

So, unless you correct me, I will skip simulation, buy some breadboarding supplies and maybe some wire wrap.

I am thinking of using Raspberry Pi as my core and building any add-ons that are not off the shelf. FPGA is a definite possibility.

Any other suggestions?

Thanks again
PMEmail Poster
Top
MacFromOK
Posted: December 22, 2016 10:38 pm
Reply to this postQuote Post


Forum Addict ++
*******

Group: Spamminator Taskforce
Posts: 15,180
Member No.: 5,314
Joined: June 04, 2006




QUOTE (phin @ December 22, 2016 03:14 pm)
From what you have said,on the analog side, I am better to breadboard then to simulate.

As a geezer and (somewhat) electronics newbie, that's always the best course in my experience, especially below RF levels. Simulations don't always show the correct "error of your ways" and at times can just drive ya nuts... biggrin.gif

Welcome to the forum btw. beer.gif


--------------------
Mac *

"Basic research is what I'm doing when I don't know what I'm doing." [Wernher Von Braun]

* is not responsible for errors, consequential damage, or... anything.
PMEmail Poster
Top
dmg
Posted: December 23, 2016 01:58 am
Reply to this postQuote Post


Forum Addict ++
*******

Group: Trusted Members
Posts: 1,265
Member No.: 36,860
Joined: June 06, 2012




well i would wish Sch3mat1c would be teaching nearby at the electronics school..
i -rarely- popp in to give a class or 2 for free, if the teachers are missing.
it seems they are still teaching tubes and transistors witch is not bad, but don't even get close to ICs witch is.. in my opinion a verry bad idea.
you see, if you need nothing more than to amplify something like 1 volt to 2 volts, these kids seem to think that a transistor is a "good" solution.
nowdays i would question if an opamp would not be more suited.
at this moment a tl071 is cheaper here then the cheapest transistor i could find.
not kidding here. and that opamp is actually not a badone.

i think that specially youngsters in eelctronics should learn about tubes and transistors and FETs and so on, but as soon as possible start working with ICs.
i think it won't take too long till it would be pointless to even search for a discrete transistor at all.
maybe high voltage and/or high curent jfets are going to stay for a while, but even that is questionable to me as to how long.

i think it is simular to writing stuff.
surely carving into stone was one methood but nowdays its quite obsolete for general public. and later on came the pen and the paper, but.. even so i have to be honest in the last decade i would swear i have not wrote anything by hand expect my signature.
hand drawn schematics are not bad for sketching up things, but even that became questionable with all these tablet things going around.

interesting world we have today, the hit single:
video killed the radio star comes to my mind quite often.
and nowday internet kills the tv star biggrin.gif
PMEmail Poster
Top
phin
Posted: December 23, 2016 02:18 am
Reply to this postQuote Post


Newbie
**

Group: Members+
Posts: 11
Member No.: 40,315
Joined: December 19, 2016




QUOTE (MacFromOK @ December 22, 2016 04:38 pm)

As a geezer and (somewhat) electronics newbie, that's always the best course in my experience, especially below RF levels. Simulations don't always show the correct "error of your ways" and at times can just drive ya nuts... biggrin.gif

You know, I am kind of surprised by the sad state of simulation as stated in this thread. I was really kind of expecting it to be super smart! Kind of like: name your CPU, copy in your code and see the outputs move.

What a shame.
PMEmail Poster
Top
Sch3mat1c
Posted: December 23, 2016 02:39 am
Reply to this postQuote Post


Forum Addict ++
Group Icon

Group: Moderators
Posts: 20,538
Member No.: 73
Joined: July 24, 2002




QUOTE (phin @ December 22, 2016 08:18 pm)
You know, I am kind of surprised by the sad state of simulation as stated in this thread. I was really kind of expecting it to be super smart! Kind of like: name your CPU, copy in your code and see the outputs move.

What a shame.

IIRC, Multisim can do that with 8051 and PIC. Not sure how "copy and paste" it is.

MCU emulation tends to be very slow (but very accurate). I think most dev kits come with emulation, or if not, you can just as easily push it to a real device and watch it with a debugger.

It's really quite sad to me that simulators have stagnated; there's literally nothing new in the free sector since the early 90s. And no EDA companies have contributed to that codebase. Some are in the same place they were, one or two decades ago (e.g., Altium); some made a few proprietary additions (Altium's XSPICE backend is extended with SimCode; Multisim's with a proprietary code-model language; PSPICE uses hidden or built-in library components), few made any real contributions to stability or speed. (LTSpice is the only one I know of that uses multithreading.)

Advanced techniques (like partial-matrix evaluation: i.e., the high frequency oscillator section chugs away on its own, independent from the slow part of your circuit) haven't filtered down.

The market has fragmented, using simulators as a means of controlling customer base: LTSpice is chock full of LT models (that are usually encrypted, and even if not, use "undocumented" logic macros that aren't compatible with anything else), ADI made their own thing, TI has TINA, Multisim's still around but it's I guess more web- and Mouser-oriented now (I haven't checked it out), and a few others.

You can pay a lot for enterprise scale simulators (HSPICE, Ansoft anything, etc.), but you're unlikely to get any more models with them. They're mainly used for RF and IC design. For example, AFAIK, HSPICE has support for IC foundry authentic models -- you merely have to sign an NDA and buy a license to use the device model! doh.gif

Tim


--------------------
Answering questions is a tricky subject to practice. Not due to the difficulty of formulating or locating answers, but due to the human inability of asking the right questions; a skill that, were one to possess, would put them in the "answering" category.
PMEmail PosterUsers Website
Top
MacFromOK
Posted: December 23, 2016 02:40 am
Reply to this postQuote Post


Forum Addict ++
*******

Group: Spamminator Taskforce
Posts: 15,180
Member No.: 5,314
Joined: June 04, 2006




QUOTE (dmg @ December 22, 2016 06:58 pm)
well i would wish Sch3mat1c would be teaching nearby at the electronics school..

Hehe, I'm sure several of us do. I've said this before, but the movie "Good Will Hunting" often comes to mind regarding Tim. He's scary smart on all aspects of electronics, and has been since a teenager.

However... we do remind him from time to time that he's dealing with mere mortals, and also try to prevent his head from getting too big. laugh.gif


--------------------
Mac *

"Basic research is what I'm doing when I don't know what I'm doing." [Wernher Von Braun]

* is not responsible for errors, consequential damage, or... anything.
PMEmail Poster
Top
kellys_eye
Posted: December 23, 2016 03:06 pm
Reply to this postQuote Post


Forum Addict ++
*******

Group: Spamminator Taskforce
Posts: 11,418
Member No.: 2,735
Joined: June 21, 2005




I'm just grateful Tim (and others of course) are willing to SHARE his knowledge here - when I was industry it was seen to be career suicide to reveal your knowledge to other staff members doh.gif

Thanks Tim thumbsup.gif


--------------------
May contain nuts
PMEmail PosterYahoo
Top
Sch3mat1c
Posted: December 23, 2016 11:48 pm
Reply to this postQuote Post


Forum Addict ++
Group Icon

Group: Moderators
Posts: 20,538
Member No.: 73
Joined: July 24, 2002




My experience has been:
The people who think their knowledge is hard-won and should be guarded, are usually not the smart ones.

You can try to teach them new knowledge, but it usually falls on deaf ears.

So, the ones who would benefit most from it, reject it anyway (for any number of reasons).

The ones who do learn from it, are the ones you want as your allies. Because when they learn something, they'll probably teach it to you, too.

Tim


--------------------
Answering questions is a tricky subject to practice. Not due to the difficulty of formulating or locating answers, but due to the human inability of asking the right questions; a skill that, were one to possess, would put them in the "answering" category.
PMEmail PosterUsers Website
Top
phin
Posted: December 24, 2016 03:43 am
Reply to this postQuote Post


Newbie
**

Group: Members+
Posts: 11
Member No.: 40,315
Joined: December 19, 2016




QUOTE (Sch3mat1c @ December 22, 2016 08:39 pm)
[/QUOTE]
IIRC, Multisim can do that with 8051 and PIC. Not sure how "copy and paste" it is. .....

It's really quite sad to me that simulators have stagnated; there's literally nothing new in the free sector since the early 90s.

Tim,

This makes me think that I might want to experiment with some different simulators and see what they can, or cannot, do. Maybe different ones for different purposes.

However; this would be a waste of time if then cannot accept the output of my capture program. I am not willing to re-enter my schematic to get a simulator to work as the chance of error ( and the result of such an error ) are too large to make it sensible to try.

Are the output capabilities of either Eagle or Kicad any more universally usable by the suggested simulators? Or is this just a crap shoot? Lacking any suggestions as to how best to tie a capture program to a potential simulator(s), I would probably pick a capture program and just see what the simulators can do.

This territory is not new to me as I saw similar issue with CAD outputs in my time in industry. All output files are NOT created equally!

Any suggestions for a starting point? I am somewhat of a willing experiment, just don't hold your breath waiting for the results as my plan is fairly long term. I would just prefer to NOT pick the wrong starting point (ie capture program).

Thanks again
PMEmail Poster
Top
Sch3mat1c
Posted: December 24, 2016 11:50 am
Reply to this postQuote Post


Forum Addict ++
Group Icon

Group: Moderators
Posts: 20,538
Member No.: 73
Joined: July 24, 2002




Dunno. I don't work in them so see what others have done about it. smile.gif

Tim


--------------------
Answering questions is a tricky subject to practice. Not due to the difficulty of formulating or locating answers, but due to the human inability of asking the right questions; a skill that, were one to possess, would put them in the "answering" category.
PMEmail PosterUsers Website
Top
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

Topic Options Reply to this topicStart new topicStart Poll

 


:: support us ::




ElectronicsSkin by DutchDork & The-Force