Author Topic: DEX eval by free_electron  (Read 347964 times)

0 Members and 3 Guests are viewing this topic.

Offline free_electronTopic starter

  • Super Contributor
  • ***
  • Posts: 8550
  • Country: us
    • SiliconValleyGarage
Re: DEX eval by free_electron
« Reply #450 on: February 14, 2015, 08:14:01 pm »
At least on that i can agree with Iliya : Do go download it and see for yourself.

And please, if you find a bug or annoyance : do send em in. His responses make for a good laugh.

there are absolutely no other restrictions, no board size or pin limits.

but ...

- no paste layers
- no mask layers
- no group operations
- no from-to operations
- no working library
- no 'real' sample files apart from a 'typical horn circuit': (see the picture i posted earlier.)

'Ford Horn Ciruit'  (typo -1-)
for a ..dodge viper (car image  -2-)
with three horns (yeah sure, typical cars have three horns. -3-)
a shorted relay coil (that relay never does anything -4-)
missing connection dots (at least 5 of em -5,6,7,8 and 9 -)
a switch that does nothing but connect the horns, through a resistor and on a mystery box labeled 'y') to ground. -10-
on a sheet labeled 'Horn Ciruit' (typo -11-)
Designed by someone appropriately named S.M Art : Sado Masochistic Art ... At least Art is happy  with Dex: it fits his tendencies to a dot. So there is at least one known customer that uses it...

all at revision 1:8

still with all those mistakes in it ? you'll need a couple more revisions matey...

then again , what do i know ? i can't even type correctly
« Last Edit: February 14, 2015, 08:34:30 pm by free_electron »
Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).
 

Offline nali

  • Frequent Contributor
  • **
  • Posts: 713
  • Country: gb
Re: DEX eval by free_electron
« Reply #451 on: February 14, 2015, 08:20:49 pm »
your downloads may go up ... but the important question is : do your sales go up ?


A few of them are probably like me :-

1. Download & Install
2. Take a look at the example
3. Move a resistor up & down, crazy rubberbanding
4. Read this thread
5. Uninstall




 

Offline ElektroQuark

  • Supporter
  • ****
  • Posts: 1245
  • Country: es
    • ElektroQuark
Re: DEX eval by free_electron
« Reply #452 on: February 14, 2015, 08:45:53 pm »
Thank you Iliya.
You have simplified my list of possible PCB software candidates.
Bye DEX. Uninstalling...

Offline elgonzo

  • Supporter
  • ****
  • Posts: 688
  • Country: 00
Re: DEX eval by free_electron
« Reply #453 on: February 14, 2015, 09:04:35 pm »
My downloads go up. :-+ Thanks.

No reason for thanks. Glad i could contribute to your download counter going up.
What else i did with DEX after downloading:

1. Install in VM.
2. Fiddling around a bit.
3. Shutting down VM.
4. Followed this thread and the Best PCB Design Practices and Books/Articles thread.
5. Deleted VM (i can even rightfully claim that i did install DEX and so far have not uninstalled it again  :P )

P.S.: Bye!
P.P.S.: The incredible "Ford Horn Circuit" makes me really want to download Dex again, just to see it with my own eyes...
« Last Edit: February 14, 2015, 09:15:43 pm by elgonzo »
 

Offline zapta

  • Super Contributor
  • ***
  • Posts: 6289
  • Country: 00
Re: DEX eval by free_electron
« Reply #454 on: February 14, 2015, 09:08:01 pm »
The non-commercial license's only restriction is that it must only be used for non-commercial designs; there are absolutely no other restrictions, no board size or pin limits.

How does this, works, if I design an open source project with dex, do I need to restrict other people from using that design commercially?
« Last Edit: February 14, 2015, 09:10:21 pm by zapta »
 

Offline iampoor

  • Frequent Contributor
  • **
  • Posts: 500
  • Country: us
Re: DEX eval by free_electron
« Reply #455 on: February 14, 2015, 10:28:44 pm »

It's astonishing how such a self-proclaimed expert can't master a simple program.


Im just amazed that you think your program is usable. Seriously, it makes Eagle's UI look good.  |O
 

Offline IconicPCB

  • Super Contributor
  • ***
  • Posts: 1546
  • Country: au
Re: DEX eval by free_electron
« Reply #456 on: February 14, 2015, 11:34:18 pm »
"... If he is immodest enough to proclaim himself an expert it wouldn't make him any less expert. ..."

Indeed..but ... he also claims the right to tell others what they should do based on his self styled expertise rather than accepting a tool and and applying it with the same patience and care he exercised when he touched the software which formed his expertise.

He should understand his expertise is just ...his opinion and accordingly tone down his attitude.  Stone tablets are a part of another story.

All the other member who have taken on the partisan position ought to consider what happens when group-think takes over and where it leads. Keep an open mind, exercise the tool ( DEX) and form your own critical opinions of the tool don’t just wank off the confrontation developing before You.

To the friendly community contributor... next time You find Yourself heading into a confrontation of this magnitude. Don't. Just remember more flies with honey than vinegar.. although shit attracts them too.

A friendly comment from a REAL SELF STYLED EXPERT even if I say so. 
 

Offline elgonzo

  • Supporter
  • ****
  • Posts: 688
  • Country: 00
Re: DEX eval by free_electron
« Reply #457 on: February 14, 2015, 11:57:28 pm »
All the other member who have taken on the partisan position ought to consider what happens when group-think takes over and where it leads. Keep an open mind, exercise the tool ( DEX) and form your own critical opinions of the tool don’t just wank off the confrontation developing before You.

Iliya noticed an increase in downloads of DEX lately. Mere coincidence?
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 28059
  • Country: nl
    • NCT Developments
Re: DEX eval by free_electron
« Reply #458 on: February 15, 2015, 12:29:34 am »
"... If he is immodest enough to proclaim himself an expert it wouldn't make him any less expert. ..."

Indeed..but ... he also claims the right to tell others what they should do based on his self styled expertise rather than accepting a tool and and applying it with the same patience and care he exercised when he touched the software which formed his expertise.
I agree. FE cleary doesn't allow software to be different to what he is used to. And that is a big part of the problem. I have not tried DEX myself but I have encountered software which doesn't work the way I was used to. It turns out that a different way of doing things can actually be a step forward. As a software developer it is hard to deal with somenone who wants your software to follow a different phylosophy. That just doesn't work.

To add to the pissing contest: I have been making PCBs for 28 years  O0
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11713
  • Country: my
  • reassessing directives...
Re: DEX eval by free_electron
« Reply #459 on: February 15, 2015, 01:18:55 am »
i know this is going to happen. if i tell this earlier, people will call me partisan, prejudice, self proclaim or whatever. if i tell this now, people will say of course you know it, it already happened! ::D

All the other member who have taken on the partisan position ought to consider what happens when group-think takes over and where it leads.
somebody asked how to do something and then somebody reply... "use mspaint", whose head in the ass here?

have you tried dex? instead of making corrective preaching, why dont you help iliya to help fe with some problems/questions he stumbled upon? continue, see below...

I agree. FE cleary doesn't allow software to be different to what he is used to. And that is a big part of the problem. I have not tried DEX myself..
ah that is the problem, the bolded one... for those of you who havent tried dex, dont preach on the "partisan" side of the dex, probably you dont get what fe (and the rest of us who tried dex) tried to tell/convey/ask you people. agreed dex is different, but "different" in different way, different in a way that way way less efficient than what fe/we used to. lets just say as analogy, one guy today invent car with square wheel, and then people back him up saying we are wrong because we dont let car to be different than round wheel... but in the end, i agree... mourning and lengthy review/comment or whatnot is useless (with this kind of inventor attitude).. i have more to say/review/complaint if i want to but i've given up, just not worth my time. if you think you havent tried square wheel car and want to buy it, then buy it its your money and time... save the good thing now is dex let us try it before buying (just like the rest of packages) and trial text reminder embedded in gerber output only too! 100++ thumbs up for that "novelty"! i hope it remains that way from now on (until dex is becoming useful to what we used to). havent try it? go try it! its free, we recommended it (just for the try), iliya recommended it, no partisan here. you will contribute to more dex download count :P iliya hopes to get more users, we hope you get what we meant all this while, the more and the longer you enjoy the program and better engineer you are ;) edit: or maybe you'll never get what we meant, because you never drive a round wheel car and hence forever happy with the "different and novel invention" of the square wheel car. ymmv.
« Last Edit: February 15, 2015, 01:35:54 am by Mechatrommer »
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 

Offline bigmik

  • Newbie
  • Posts: 6
  • Country: 00
Re: DEX eval by free_electron
« Reply #460 on: February 15, 2015, 01:34:59 am »
Hi All,

Whilst I am sad that there could be no reconciliation between FE and Iliya, I can see both points of view, I think FE went about the `feedback' in a manner that could be deemed to be `rubbing your face in it' and that is not easy to take sometimes, I also Feel that Iliya did not respond professionally and allowed himself to be involved in a `slanging match' which is not conducive to either side.

I think FE has ideas on what he expects DEX to do and how it should be done. I rather, try to work with what there is rather than try to force the author to my way of thinking. Bugs are definitely there (I have never seen the artifacts FE showed when selecting pads.. that seems like a video driver conflict to me) but it is still a work in progress. I ask you, take WORD or EXCEL, is every function the way you think it should be? If I had hair I would pull it out when I cannot find how to do something I think should be easy to do.. From my experience no printed circuit board program worked the way that others do, admittedly I started with DOS based ones and then when my company bought Protel 98 then 99se I was locked into that for 15 years. But Protel was NOT intuitive either, zooming and panning etc were a right royal pain.. It didn't do what I expected it to do but I would put up with its idiosyncrasies and work with it. I feel the same with DEX, it is NOT easy to learn from scratch but the more you play the more you start to see the way that it works.. I have had 11 months use of DEX and am only now (in the last month) starting to feel quite comfortable with using it.. Admittedly I might use it today but might not use it again for several weeks, I usually do small tweaks here and there but unless I have a new board I don't put in all my efforts 24/7, I use it as a hobby and it fulfills all that I can ever see me requiring EVER and more. I will also sum up with if DEX doesn't do the high level `NASA' type boards you require because a solder mask is 1mil out or something then don't waste your $49 and go and pay for Altium, I cannot even get a guide to its price as you have to contact your local dealer but as they sold Protel for around $4k in 1999 I can only surmise that it is around the $15-$20k now.

All that aside this is Free_electron's Thread and maybe if we `fork' away from his review a new Autotrax DEX thread should be started where what DEX can do rather than what it cannot, unless as suggestions, can be discussed. I have found Iliya to be open and approachable with problems I have found, more often than not my fault (not knowing how to do something) rather than DEX bugs, which I think goes a long way to making me feel comfortable that I have paid for a product that I can get support for when I need it.


Regards,

Mick

 

Offline free_electronTopic starter

  • Super Contributor
  • ***
  • Posts: 8550
  • Country: us
    • SiliconValleyGarage
Re: DEX eval by free_electron
« Reply #461 on: February 15, 2015, 01:38:42 am »
Software is allowed to be different. I can perfectly fine deal with different pieces of software. Some of the things i have shown during this review were clearly bugs, and fixed. good.  :-+
Some other things i commented on are things that one can reasonably expect a cad program to be able to do. Like moving a selection of elements to a new position by entering a numerical offset or aligning them to a grid. Other things are , for lack of a better word, defective; like the autowiring That thing just puts stuff down where-ever, just play with it.

And yes, i have opinions about these things. Opinions that are formed by using various CAD programs over the years. They all share some common base functionality.  Like being able to move a selection of elements without disturbing the spatial relations inside that group.

Also, along this topic i have given some insight in best practices which are used in the PCB industry , and pointers where to get more information as well only to be met with more hostility.

DEX, as a one man project is amazing. The UI is fresh and the program works smoothly , but it has quirks and bugs. Bugs that are fixed quickly . Good. I have to grant that to Iliya : he is quick fixing bugs. Quirks : some are livable and have workarounds or a 'less-standard' why of doing it. Some are just extremely annoying. Whenever i encountered such a blocker i have shown where i encountered it and given reasons why it is a blocker.

I've posted links to good resources and reading material as well and made a suggestion to visit a PCB fab ,an assembly house. Maybe even a tradeshow like APEX and take a look at what is out there , technology wise.

It is not normal for a PCB tool to not have the capability to create paste masks. All you can do is create mask on or off. You can't size elements , you can't even add or remove elements. So there is clearly room for improvement. The same goes for moving groups of elements without disturbing what is already in the group. That just is not workable.

The documentation of the program is a bit lacking which makes it hard to learn and discover what is there. There are some video's but those touch on very simple stuff , or using the wizards. As soon as you want to do something 'manually' dex falls short of expectations.

Dex has some really cool idea's. one thing i like is the linking between schematic and pcb: No need to do things like pull a netlist to go from schematic to board. Place a part in the schematic and it appears in the pcb. No questions asked. That is nice. On the other hand similar 'automatic' processes can be a hindrance. For example : make a new component. Place three pads , place the pins in the schematic symbol , add some artwork. Go back to the footprint editor and place a pad , go back to schematic symbol editing and see what happened. Dex has placed another rectangle and shifted the pins that were already placed. That is a hindrance. Such hindrance is easily solved by adding an attribute in the program to the objects. : 'user manipulated'. Before the automated routine modifies the part it should check for each object if that object has already been touched by the user. If it has: hands off. only add the new stuff. If it has not been touched by the user the routine can move it. That is a fairly logical assumption to make. 

Now, you may say , DEX doesn't work that way, you are supposed to draw pads first , symbol second. And that is fine. If that is a limitation of the program, or simply the way the program works than so be it. But that should not take away my right to complain about that, or tell you why, in my opinion this is 'stupid'. We live in a non-linear world There may be a point in time where i need to modify a footprint a bit. Add two reinforcement pads for mechanical strength for example.

Software developers listen to their users and they look at their competition for ideas as well. I tried to be a user. i failed. Too different , lots of bugs ( now fixed, at lest the ones i reported ) and , for my taste, shortcomings that make it unusable for me.  This is an open forum. I documented every step i took and you can try the same for yourself. Everyone is entitle to an opinion.

Iliya also still is convinced i am 'after him'. I am not. Like i said many times : DEX intrigues me as it is very impressive as a one man show. It looks promising, but the images in the documentation and the sample designs do not reflect well. Sub-par routing is one aspect of it. So i thought maybe i can create a projec tin DEX that shows off better. iliya himself has asked for people to post their designs. Has anyone stepped up to the plate ? (one guy did, but they aren't quite state of the art designs) At least i tried. I've put a lot of hours into this and am now about at the point where i can make a footprint, but i'm stuck as i can not make it according to even manufacturers drawings.
i have not commented on anything beyond that point because i am focused on footprints first : they are the foundation of a PCB layout. Limiting factors in dex make it impossible to even make the footprint according to what the part manufacturer specifies , for example the paste mask settings: The part manufacturer specifies the overprint or underprint and the lattice structure for the thermal pad. Dex has no ability to edit the paste mask. That is clearly a huge shortcoming. State of the art designs require these things.

The foundation is there , there are a bunch of rough edges and a bunch of shortcomings. I helped polish some of the rough edges and iliya was quick to solve these. the shortcomings ... those are met with a lot of resistance. I do understand they may not be easy to code up. But that is the life of software developing.

All i have done is point out  what some of those shortcomings are. If the software creator does not like that ... so be it. Freedom of speech. This is a review and an open forum. Other users of DEX could have pitched in showing me how to do it. sadly none came forward.
Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11713
  • Country: my
  • reassessing directives...
Re: DEX eval by free_electron
« Reply #462 on: February 15, 2015, 01:48:45 am »
in economic theory, $20K price is probably about right, but in practice who want to pay 20K for that sheet? i have that trial with crack version but i wont touch it with "20ft pole tongue at the right angle". i suggest you get acquainted with diptrace to save more precious time in the future that you unknowingly waste, if you are comfortable within 300 pins 2 layers board boundary or think will have/gain small amount of money in the future to upgrade, diptrace is way more efficient. diptrace has slightly dated GUI (protel is wayyy dated) but who give a sheet as long as its more efficient. and who give a sheet on fancy modern GUI if its not that efficient? ps: i love word1997 better than word2015 because the GUI is more efficient.
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 28059
  • Country: nl
    • NCT Developments
Re: DEX eval by free_electron
« Reply #463 on: February 15, 2015, 02:15:35 am »
i know this is going to happen. if i tell this earlier, people will call me partisan, prejudice, self proclaim or whatever. if i tell this now, people will say of course you know it, it already happened! ::D

All the other member who have taken on the partisan position ought to consider what happens when group-think takes over and where it leads.
somebody asked how to do something and then somebody reply... "use mspaint", whose head in the ass here?

have you tried dex? instead of making corrective preaching, why dont you help iliya to help fe with some problems/questions he stumbled upon? continue, see below...

I agree. FE cleary doesn't allow software to be different to what he is used to. And that is a big part of the problem. I have not tried DEX myself..
ah that is the problem, the bolded one... for those of you who havent tried dex, dont preach on the "partisan" side of the dex, probably you dont get what fe (and the rest of us who tried dex) tried to tell/convey/ask you people. agreed dex is different, but "different" in different way, different in a way that way way less efficient than what fe/we used to. lets just say as
I've seen FE write very negative about other PCB software (a package called Layo1) which has a large and happy user base in the NL. Actually there are many parallels to the way he tried to use Layo1 and DEX. Many of the problems he spotted in Layo1 where just user error.
« Last Edit: February 15, 2015, 02:17:45 am by nctnico »
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline DerekG

  • Frequent Contributor
  • **
  • Posts: 883
  • Country: nf
Re: DEX eval by free_electron
« Reply #464 on: February 15, 2015, 02:55:29 am »
There is at least one things that works correctly in DEX : the uninstaller   :-DD

Unfortunately FE, this is also not correct.

Do you have Revo Uninstaller loaded? If you do, you will find there are legacy bits left all over the place & that includes LOTS in your Windows Registry.

Now I wish you had spent your time doing an evaluation on either DipTrace or Proteus.
I also sat between Elvis & Bigfoot on the UFO.
 

Offline IconicPCB

  • Super Contributor
  • ***
  • Posts: 1546
  • Country: au
Re: DEX eval by free_electron
« Reply #465 on: February 15, 2015, 03:44:22 am »
 I have installed DEX.

I have had a look at the user interface and found it daunting coming from zero experience with DEX.

I have used rubylith..Bishop graphics blue, red and black library parts on a bifolding mylar user interface.

I have used autotrax under DOS, protel under various windows flavours, i have licenced copies of protel 98, 98SE and 2004XP as well as the almost current flaghip AD14.

DEX is nothing like any of the above and that is OK too. At first blush too many things happening , i feel like I am fighting my way through clutter. I am sure that sensation will vanish as I become familiar with various functionalities.

You know what they say ... what is on the outside reflects what is inside. In this instance I do not think chaos in my mind reflects DEX.
 
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11713
  • Country: my
  • reassessing directives...
Re: DEX eval by free_electron
« Reply #466 on: February 15, 2015, 03:52:49 am »
I've seen FE write very negative about other PCB software (a package called Layo1) which has a large and happy user base in the NL. Actually there are many parallels to the way he tried to use Layo1 and DEX. Many of the problems he spotted in Layo1 where just user error.
i didnt read all his comment but i am just as annoyed in some dex aspects. not wishing to detail it because it will be too long to type hence waste time. iliya is correct about it... stfu, download, install and feel it for yourself. if it works for you then it works. not trying to be too negative, let me just say again, one man achievement of software at the scale like dex is very rare no doubt there. but then again, what good it is a ferrari looking beauty if the wheel (basic things) is square? i wish dex success in the future.
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11713
  • Country: my
  • reassessing directives...
Re: DEX eval by free_electron
« Reply #467 on: February 15, 2015, 04:50:49 am »
now my antivirus reports suspected Win32.Evo Gen virus for both version 5.01 and 5.02 i just downloaded :palm: latest version need password to install :palm: so, BYE dex for now.
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 

Online Monkeh

  • Super Contributor
  • ***
  • Posts: 8061
  • Country: gb
Re: DEX eval by free_electron
« Reply #468 on: February 15, 2015, 05:13:16 am »
I wonder how much he makes per email..
 

Offline free_electronTopic starter

  • Super Contributor
  • ***
  • Posts: 8550
  • Country: us
    • SiliconValleyGarage
Re: DEX eval by free_electron
« Reply #469 on: February 15, 2015, 05:41:26 am »
if anyone wants to try : here is the footprint i was trying to make in DEX. I can't figure it out. my limitation..

http://www.st.com/web/en/resource/technical/document/datasheet/CD00000956.pdf

It's not like it is that complicated a part but it has a few quirks. The thermal slug under the part is not a rectangle.

I will try to describe as best as i can and give you the numbers ( i don't want you to spend time figuring them out )

Draw the two rows of pads bottom row: 1 to 10 , top row 20 to 11. Pads are all 0.6mm wide and 2.2 mm tall.
Pin one is a rectangle, all others are 'oblongs' (rounded rectangle where the outer edge makes a 180 degree arc, so there is no 'flat' portion on the short side of the pads)

Pin pitch is 50 mils
Row pitch is 14.2mm

The center thermal pad is number 21 and measures 6.3mm by 9.5mm
For simplicity : All pads have a solder mask opening of 4mils over the edge of the pad.

the package is a silkscreen rectangle of 10mm by 16mm. no need for anything fancy here. Silkscreen rectangle center coincides with center of thermal pad.

Now, I need a piece of copper 1cm 'over' both sides of the package , we'll call em wings. so left side of package + 1 cm and right side of package + 1cm. The height is from the toe of the top pad to the toe of the bottom pad. Don't ask me what that number is. the datasheets don't give it. use the cad tool to adjust the edge to coincide  (actually that height is the pad pitch + 1 height of the pad but i want you to try to adjust it without punching numbers. basically using snap operations like snap to edge , snap to near object, snap to center and the likes. basic CAD operations i have trouble in finding in DEX.

the 'spoke' ( the piece of copper that connects the thermal pad to the two outer 'wings' ) has a height of 170 mils

The backside of the board should have a pad that is the size of the entire structure on top. Basically take the top left corner coordinate of the left wing and the bottom right corner coordinate of the right wing. that size. i don't know what the numbers are. simply use the cad tool to make coinciding points. no need for calculations, leave those to the computer.
this back structure has a solder mask opening of 4 mils over its side.

Note that neither the wings nor the back structure are 'pads'. they are simply pieces of copper and have no pin number assigned to the, ,only the center thermal pad has a number

For manufacturability we need two soldermask 'dams' just outside the package so solder cannot wick onto the top 'wings'. The dam should be 15 mils wide. (roughly, i don't mind if it's 18 or 14 so you can position it as you see fit to avoid complex calculations) . This dam sits at the edge of the thermal pad. it will not overlap the thermal pad nor leave a piece of copper of the wings exposed to the thermal pad ( so one edge of the soldermask sliver must coincide with the edge of the thermal pad. again , i have no numbers to give you. Use basic operations like snap to point, snap to edge to make coinciding structures.

We need an array of 3 x 6 thermal via's in the main portion of the heatslug. Via hole : 10 mils. Spaced 50 mils apart horizontally and 75 mils vertically . The array is placed around the center of the thermal pad. (The center of the array coincides with center of the thermal pad )

A paste lattice is to be applied with a minimum distance of 5 mils from the thermal via holes and a lattice spoke width of no less than 20 mils ( to avoid the mask 'curling')
I know DEX does not have Paste layers so i'm fine if you can draw this structure on the silkscreen layer or on a mechanical layer or documentation layer. At least i can export it to Gerber and use a gerber editor to throw it on a different layer (or-ing the two layers) .

Paste shall be retracted 4 mils from the pad edge.

The courtyard shall be 8 mils over the edge of the entire structure including the soldermask gaps.. so the left side of the courtyard is at the left edge of the left wing + 4 mils of soldermask opening + 8 mils of space and so on.

for visual reference i have added a couple of pictures what that looks like.

First two are with step file and without ( no need to do that for this exercise , they are so you can understand what this package is. many stepper motor controllers or H-bridge drivers come in this package. Even the L298 is available in this package )

Picture two shows the thermal via lattice in the thermal pad and the solder solder mask dams at the edge of the thermal pad. We don't want paste bleeding out.

Picture 3 shows the solder mask (purple) so you can clearly see where the dams are.
Picture 4 shows the paste mask. i know dex does not have this, but draw it as rectangles on the silkscreen layer if you must. Just to experience the level of difficulty to create such a thing.

My workflow :
I heavily rely on things like snap-to-point, snap-to-edge , from-to copying. once i have the centerpoint established i place the origin there and then start playing with grid spacing . Having snap-to grid on makes positioning and scaling of objects really easy. Making a from-to copy of a collection and pasting it mirrored around x or y axis. Scaling of groups and other typical stuff found in CAD programs ( not just PCB cad. ) Entering numerical data in fields where one field reads mm and the other mils.  That is my way of working. I rely on the CAD tool to crunch the numbers and come up with the actual coordinates.

I can't find the CAD operations is seek in DEX. Maybe they are there and i am too stupid to find em, maybe they are not there. Maybe that way of working is not compatible with DEX. I don't know. I give up. It is too time consuming to figure it out and we have known limitations like group moves (both graphically and numerically)  that don't work like a traditional CAD program, so they hamper this workflow and make designing this footprint harder than it should be.

One important thing : do NOT use the wizard, not even to create a base structure of the two columns. This is not an exercise in running the wizards. I know that you could take advantage of running the wizard and then build on that, and in real life that is what i do too. But that is not the point here.

The goal of this exercise is to get a feel of what base object manipulation toolset is available in DEX ,and find out where the necessary controls and commands are located to efficiently play with grids , scale and position objects and set their properties.
« Last Edit: February 15, 2015, 05:59:21 am by free_electron »
Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).
 

Offline firewalker

  • Super Contributor
  • ***
  • Posts: 2452
  • Country: gr
Re: DEX eval by free_electron
« Reply #470 on: February 15, 2015, 08:12:02 am »
Ok. We should now force free_electron to "evaluate" kicad or GEDA.

 >:D >:D >:D >:D

Alexander.
Become a realist, stay a dreamer.

 

Offline Iliya

  • Frequent Contributor
  • **
  • !
  • Posts: 396
  • Country: 00
Re: DEX eval by free_electron
« Reply #471 on: February 15, 2015, 09:24:46 pm »
I wonder how much he makes per email..
Who?
 

Offline Iliya

  • Frequent Contributor
  • **
  • !
  • Posts: 396
  • Country: 00
Re: DEX eval by free_electron
« Reply #472 on: February 15, 2015, 09:28:29 pm »
if anyone wants to try : here is the footprint i was trying to make in DEX. I can't figure it out. my limitation..

http://www.st.com/web/en/resource/technical/document/datasheet/CD00000956.pdf

It's not like it is that complicated a part but it has a few quirks. The thermal slug under the part is not a rectangle.

I will try to describe as best as i can and give you the numbers ( i don't want you to spend time figuring them out )

Draw the two rows of pads bottom row: 1 to 10 , top row 20 to 11. Pads are all 0.6mm wide and 2.2 mm tall.
Pin one is a rectangle, all others are 'oblongs' (rounded rectangle where the outer edge makes a 180 degree arc, so there is no 'flat' portion on the short side of the pads)

Pin pitch is 50 mils
Row pitch is 14.2mm

The center thermal pad is number 21 and measures 6.3mm by 9.5mm
For simplicity : All pads have a solder mask opening of 4mils over the edge of the pad.

the package is a silkscreen rectangle of 10mm by 16mm. no need for anything fancy here. Silkscreen rectangle center coincides with center of thermal pad.

Now, I need a piece of copper 1cm 'over' both sides of the package , we'll call em wings. so left side of package + 1 cm and right side of package + 1cm. The height is from the toe of the top pad to the toe of the bottom pad. Don't ask me what that number is. the datasheets don't give it. use the cad tool to adjust the edge to coincide  (actually that height is the pad pitch + 1 height of the pad but i want you to try to adjust it without punching numbers. basically using snap operations like snap to edge , snap to near object, snap to center and the likes. basic CAD operations i have trouble in finding in DEX.

the 'spoke' ( the piece of copper that connects the thermal pad to the two outer 'wings' ) has a height of 170 mils

The backside of the board should have a pad that is the size of the entire structure on top. Basically take the top left corner coordinate of the left wing and the bottom right corner coordinate of the right wing. that size. i don't know what the numbers are. simply use the cad tool to make coinciding points. no need for calculations, leave those to the computer.
this back structure has a solder mask opening of 4 mils over its side.

Note that neither the wings nor the back structure are 'pads'. they are simply pieces of copper and have no pin number assigned to the, ,only the center thermal pad has a number

For manufacturability we need two soldermask 'dams' just outside the package so solder cannot wick onto the top 'wings'. The dam should be 15 mils wide. (roughly, i don't mind if it's 18 or 14 so you can position it as you see fit to avoid complex calculations) . This dam sits at the edge of the thermal pad. it will not overlap the thermal pad nor leave a piece of copper of the wings exposed to the thermal pad ( so one edge of the soldermask sliver must coincide with the edge of the thermal pad. again , i have no numbers to give you. Use basic operations like snap to point, snap to edge to make coinciding structures.

We need an array of 3 x 6 thermal via's in the main portion of the heatslug. Via hole : 10 mils. Spaced 50 mils apart horizontally and 75 mils vertically . The array is placed around the center of the thermal pad. (The center of the array coincides with center of the thermal pad )

A paste lattice is to be applied with a minimum distance of 5 mils from the thermal via holes and a lattice spoke width of no less than 20 mils ( to avoid the mask 'curling')
I know DEX does not have Paste layers so i'm fine if you can draw this structure on the silkscreen layer or on a mechanical layer or documentation layer. At least i can export it to Gerber and use a gerber editor to throw it on a different layer (or-ing the two layers) .

Paste shall be retracted 4 mils from the pad edge.

The courtyard shall be 8 mils over the edge of the entire structure including the soldermask gaps.. so the left side of the courtyard is at the left edge of the left wing + 4 mils of soldermask opening + 8 mils of space and so on.

for visual reference i have added a couple of pictures what that looks like.

First two are with step file and without ( no need to do that for this exercise , they are so you can understand what this package is. many stepper motor controllers or H-bridge drivers come in this package. Even the L298 is available in this package )

Picture two shows the thermal via lattice in the thermal pad and the solder solder mask dams at the edge of the thermal pad. We don't want paste bleeding out.

Picture 3 shows the solder mask (purple) so you can clearly see where the dams are.
Picture 4 shows the paste mask. i know dex does not have this, but draw it as rectangles on the silkscreen layer if you must. Just to experience the level of difficulty to create such a thing.

My workflow :
I heavily rely on things like snap-to-point, snap-to-edge , from-to copying. once i have the centerpoint established i place the origin there and then start playing with grid spacing . Having snap-to grid on makes positioning and scaling of objects really easy. Making a from-to copy of a collection and pasting it mirrored around x or y axis. Scaling of groups and other typical stuff found in CAD programs ( not just PCB cad. ) Entering numerical data in fields where one field reads mm and the other mils.  That is my way of working. I rely on the CAD tool to crunch the numbers and come up with the actual coordinates.

I can't find the CAD operations is seek in DEX. Maybe they are there and i am too stupid to find em, maybe they are not there. Maybe that way of working is not compatible with DEX. I don't know. I give up. It is too time consuming to figure it out and we have known limitations like group moves (both graphically and numerically)  that don't work like a traditional CAD program, so they hamper this workflow and make designing this footprint harder than it should be.

One important thing : do NOT use the wizard, not even to create a base structure of the two columns. This is not an exercise in running the wizards. I know that you could take advantage of running the wizard and then build on that, and in real life that is what i do too. But that is not the point here.

The goal of this exercise is to get a feel of what base object manipulation toolset is available in DEX ,and find out where the necessary controls and commands are located to efficiently play with grids , scale and position objects and set their properties.

You can't make this in DEX? I assume you are no fool so you are just being awkward. Why are you no nasty?

P.S. I'm glad to see your written English has improved.
 

Offline Iliya

  • Frequent Contributor
  • **
  • !
  • Posts: 396
  • Country: 00
Re: DEX eval by free_electron
« Reply #473 on: February 15, 2015, 09:30:00 pm »
The non-commercial license's only restriction is that it must only be used for non-commercial designs; there are absolutely no other restrictions, no board size or pin limits.

How does this, works, if I design an open source project with dex, do I need to restrict other people from using that design commercially?

Yes or pay the $49. You need to add an appropriate Open Source license.
 

Online Monkeh

  • Super Contributor
  • ***
  • Posts: 8061
  • Country: gb
Re: DEX eval by free_electron
« Reply #474 on: February 15, 2015, 09:30:38 pm »
I wonder how much he makes per email..
Who?

Only one person here is collecting emails for 'no' purpose.

Why are you no nasty?

P.S. I'm glad to see your written English has improved.

Seriously, Iliya, stop digging.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf