Memory limits?
Not sure that the original post made it through, apologies if this is a
duplicate:
I’m new to this list, but I’ve been using openSCAD for a number of
years. I think I’ve run into a few limits that I’d like confirmed. I
need to give you a background to put the questions in context.
I’ve been programming in C++ for quite a while.
I design PC boards using Eagle (7.7, the last one before Autocad
took it over, and have a hobbiest license (100 mm by 200 mm).
I also use OpenSCAD for model visualization, case design, and the like.
To do that, I have a program that reads Eagle PCB files, and
produces a description in SCAD of the PC board.
I’ve got the following working:
1.
rounded corners on the board
2.
parts placement where needed with proper rotation per part
3.
parts are modeled using either nopSCADlib or in some cases, my
own models
1.
there are assumptions built into the nopSCADlib that don’t
work for me
2.
nopSCADlib is wonderful, but somewhat limited in the parts
choices
3.
no, I don’t completely understand how it all works, but
that’s not the issue
4.
board reading and parsing, parts lookup, part generation for
openSCAD, parts placement
Now, what doesn’t work:
1.
I think there’s a limit in OpenSCAD about number of parts, which
seems to work into a “only this complex” limit.
What I do:
1.
run a board through the program.
2.
Look at the output
3.
parts are missing. (its a nopSCADlib file for an SO16 IC)
4.
Another such chip is on the board and properly placed.
5.
Commenting out lines eventually gets to the point where the
“lost” part pops back in.
6.
This suggests a memory limit somewhere.
I’d love to give you the source files, but it’s an entire project
with subdirectories, referring to other libraries of parts I’ve
designed. Since I think it’s a memory limit, a partial file would
not help all that much.
So I have a board model that is not complete. Anyone have any ideas?
I have some ideas for nopSCADlib, but those should be another post.
Harvey
On Sat, Jul 13, 2024 at 10:24:22AM -0400, Harvey white via Discuss wrote:
I’d love to give you the source files, but it’s an entire project
with subdirectories, referring to other libraries of parts I’ve
designed. Since I think it’s a memory limit, a partial file would
not help all that much.
How many parts are there? Eagle has a board-stats ULP. Then we know
how many parts we need to "simulate" to try and reproduce this.
How about you put say a resistor, led, and an sot23 transistor in a
schematic, switch-to-board, place them compactly, in schematic view
load another design, but DO NOT load the board. Now back in board
save-as with a new name, and then copy-paste your three components.
select everything, repeat doubling the number of components until
you've filled your allowed design area. Now run it through your
rendering pipeline. Now there are only three models that need to be
included.
OK. F*** it. I did that for you. Attached is a board with 192 instances
of a resistor-cap-sot23. Does this show the problem?
Roger.
--
** R.E.Wolff@BitWizard.nl ** https://www.BitWizard.nl/ ** +31-15-2049110 **
** Delftechpark 11 2628 XJ Delft, The Netherlands. KVK: 27239233 **
f equals m times a. When your f is steady, and your m is going down
your a is going up. -- Chris Hadfield about flying up the space shuttle.
I have never hit any memory limits with NopSCADlib and many parts. One of
its tests has a single instance of all the parts in the library. My biggest
design has about 2500 parts, maily fasteners, rather than electronic
components but I don't see why they would be different.
Regarding the components available I only model what I need for my own
designs. Pull requests are welcome.
On Sun, 14 Jul 2024 at 08:02, Rogier Wolff via Discuss <
discuss@lists.openscad.org> wrote:
On Sat, Jul 13, 2024 at 10:24:22AM -0400, Harvey white via Discuss wrote:
I’d love to give you the source files, but it’s an entire project
with subdirectories, referring to other libraries of parts I’ve
designed. Since I think it’s a memory limit, a partial file would
not help all that much.
How many parts are there? Eagle has a board-stats ULP. Then we know
how many parts we need to "simulate" to try and reproduce this.
How about you put say a resistor, led, and an sot23 transistor in a
schematic, switch-to-board, place them compactly, in schematic view
load another design, but DO NOT load the board. Now back in board
save-as with a new name, and then copy-paste your three components.
select everything, repeat doubling the number of components until
you've filled your allowed design area. Now run it through your
rendering pipeline. Now there are only three models that need to be
included.
OK. F*** it. I did that for you. Attached is a board with 192 instances
of a resistor-cap-sot23. Does this show the problem?
Roger.
--
** R.E.Wolff@BitWizard.nl ** https://www.BitWizard.nl/ ** +31-15-2049110
**
** Delftechpark 11 2628 XJ Delft, The Netherlands. KVK: 27239233 **
f equals m times a. When your f is steady, and your m is going down
your a is going up. -- Chris Hadfield about flying up the space shuttle.
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
OK, I think that establishes that there isn't a memory limit, it's about
250 or so parts, but they are a bit complex. So let's not think that
it's a memory limit since the demo of the openSCADlib uses much the same
parts, and more of them
I think I have something, not related to parts limits.
the following two lines:
//translate([6.191,68.262,1.600]) rotate([0,0,270])
rotate([0,0,-90]) green_terminal(gt_5p08, 3) //X2 0
translate([40.640,70.802,1.600]) rotate([0,0,90])
smd_resistor(RES0603,"22"); //R2 22
when put in a module will demonstrate the problem. The code shows one
part, the terminal block. Comment out the terminal block and the
resistor pops in.
I've done the same thing with the original code that I made the board
with, and the next part past the green terminal doesn't get on the
board. No error messages, thank you.
The way the code is formatted (as an explanation) is that the position
from the board is the first translate, the rotation is from the board.
Another translation/rotation pair may happen to adjust each part
depending on how it was modeled.
I suspect that I'm not using the terminal block correctly.
I do have some requests for nopSCADlib, but perhaps those may be better
taken off list.
I too do the parts that I need, but I'm doing more microprocessor stuff
on custom boards. I'm also doing a PNP machine which got modeled in
OpenSCAD first, I have an algorithm that is particularly useful for
moving stepper stage.
Harvey
On 7/14/2024 4:16 AM, nop head via Discuss wrote:
I have never hit any memory limits with NopSCADlib and many parts. One
of its tests has a single instance of all the parts in the library. My
biggest design has about 2500 parts, maily fasteners, rather than
electronic components but I don't see why they would be different.
Regarding the components available I only model what I need for my own
designs. Pull requests are welcome.
On Sun, 14 Jul 2024 at 08:02, Rogier Wolff via Discuss
discuss@lists.openscad.org wrote:
On Sat, Jul 13, 2024 at 10:24:22AM -0400, Harvey white via Discuss
wrote:
I’d love to give you the source files, but it’s an entire project
with subdirectories, referring to other libraries of parts I’ve
designed. Since I think it’s a memory limit, a partial file would
not help all that much.
How many parts are there? Eagle has a board-stats ULP. Then we know
how many parts we need to "simulate" to try and reproduce this.
How about you put say a resistor, led, and an sot23 transistor in a
schematic, switch-to-board, place them compactly, in schematic view
load another design, but DO NOT load the board. Now back in board
save-as with a new name, and then copy-paste your three components.
select everything, repeat doubling the number of components until
you've filled your allowed design area. Now run it through your
rendering pipeline. Now there are only three models that need to be
included.
OK. F*** it. I did that for you. Attached is a board with 192
instances
of a resistor-cap-sot23. Does this show the problem?
Roger.
--
** R.E.Wolff@BitWizard.nl ** https://www.BitWizard.nl/
<https://www.BitWizard.nl/> ** +31-15-2049110 **
** Delftechpark 11 2628 XJ Delft, The Netherlands. KVK:
27239233 **
f equals m times a. When your f is steady, and your m is going down
your a is going up. -- Chris Hadfield about flying up the space
shuttle.
_______________________________________________
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
You are missing a semicolon after the green_terminal module.
On Sun, 14 Jul 2024, 16:29 Harvey white via Discuss, <
discuss@lists.openscad.org> wrote:
OK, I think that establishes that there isn't a memory limit, it's about
250 or so parts, but they are a bit complex. So let's not think that
it's a memory limit since the demo of the openSCADlib uses much the same
parts, and more of them
I think I have something, not related to parts limits.
the following two lines:
//translate([6.191,68.262,1.600]) rotate([0,0,270])
rotate([0,0,-90]) green_terminal(gt_5p08, 3) //X2 0
translate([40.640,70.802,1.600]) rotate([0,0,90])
smd_resistor(RES0603,"22"); //R2 22
when put in a module will demonstrate the problem. The code shows one
part, the terminal block. Comment out the terminal block and the
resistor pops in.
I've done the same thing with the original code that I made the board
with, and the next part past the green terminal doesn't get on the
board. No error messages, thank you.
The way the code is formatted (as an explanation) is that the position
from the board is the first translate, the rotation is from the board.
Another translation/rotation pair may happen to adjust each part
depending on how it was modeled.
I suspect that I'm not using the terminal block correctly.
I do have some requests for nopSCADlib, but perhaps those may be better
taken off list.
I too do the parts that I need, but I'm doing more microprocessor stuff
on custom boards. I'm also doing a PNP machine which got modeled in
OpenSCAD first, I have an algorithm that is particularly useful for
moving stepper stage.
Harvey
On 7/14/2024 4:16 AM, nop head via Discuss wrote:
I have never hit any memory limits with NopSCADlib and many parts. One
of its tests has a single instance of all the parts in the library. My
biggest design has about 2500 parts, maily fasteners, rather than
electronic components but I don't see why they would be different.
Regarding the components available I only model what I need for my own
designs. Pull requests are welcome.
On Sun, 14 Jul 2024 at 08:02, Rogier Wolff via Discuss
discuss@lists.openscad.org wrote:
On Sat, Jul 13, 2024 at 10:24:22AM -0400, Harvey white via Discuss
wrote:
I’d love to give you the source files, but it’s an entire
project
with subdirectories, referring to other libraries of parts I’ve
designed. Since I think it’s a memory limit, a partial file
would
not help all that much.
How many parts are there? Eagle has a board-stats ULP. Then we know
how many parts we need to "simulate" to try and reproduce this.
How about you put say a resistor, led, and an sot23 transistor in a
schematic, switch-to-board, place them compactly, in schematic view
load another design, but DO NOT load the board. Now back in board
save-as with a new name, and then copy-paste your three components.
select everything, repeat doubling the number of components until
you've filled your allowed design area. Now run it through your
rendering pipeline. Now there are only three models that need to be
included.
OK. F*** it. I did that for you. Attached is a board with 192
instances
of a resistor-cap-sot23. Does this show the problem?
Roger.
--
** R.E.Wolff@BitWizard.nl ** https://www.BitWizard.nl/
<https://www.BitWizard.nl/> ** +31-15-2049110 **
** Delftechpark 11 2628 XJ Delft, The Netherlands. KVK:
27239233 **
f equals m times a. When your f is steady, and your m is going down
your a is going up. -- Chris Hadfield about flying up the space
shuttle.
_______________________________________________
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
Also components and their positions are normally part of a PCB definition,
so they get placed automatically.
On Sun, 14 Jul 2024 at 16:35, nop head nop.head@gmail.com wrote:
You are missing a semicolon after the green_terminal module.
On Sun, 14 Jul 2024, 16:29 Harvey white via Discuss, <
discuss@lists.openscad.org> wrote:
OK, I think that establishes that there isn't a memory limit, it's about
250 or so parts, but they are a bit complex. So let's not think that
it's a memory limit since the demo of the openSCADlib uses much the same
parts, and more of them
I think I have something, not related to parts limits.
the following two lines:
//translate([6.191,68.262,1.600]) rotate([0,0,270])
rotate([0,0,-90]) green_terminal(gt_5p08, 3) //X2 0
translate([40.640,70.802,1.600]) rotate([0,0,90])
smd_resistor(RES0603,"22"); //R2 22
when put in a module will demonstrate the problem. The code shows one
part, the terminal block. Comment out the terminal block and the
resistor pops in.
I've done the same thing with the original code that I made the board
with, and the next part past the green terminal doesn't get on the
board. No error messages, thank you.
The way the code is formatted (as an explanation) is that the position
from the board is the first translate, the rotation is from the board.
Another translation/rotation pair may happen to adjust each part
depending on how it was modeled.
I suspect that I'm not using the terminal block correctly.
I do have some requests for nopSCADlib, but perhaps those may be better
taken off list.
I too do the parts that I need, but I'm doing more microprocessor stuff
on custom boards. I'm also doing a PNP machine which got modeled in
OpenSCAD first, I have an algorithm that is particularly useful for
moving stepper stage.
Harvey
On 7/14/2024 4:16 AM, nop head via Discuss wrote:
I have never hit any memory limits with NopSCADlib and many parts. One
of its tests has a single instance of all the parts in the library. My
biggest design has about 2500 parts, maily fasteners, rather than
electronic components but I don't see why they would be different.
Regarding the components available I only model what I need for my own
designs. Pull requests are welcome.
On Sun, 14 Jul 2024 at 08:02, Rogier Wolff via Discuss
discuss@lists.openscad.org wrote:
On Sat, Jul 13, 2024 at 10:24:22AM -0400, Harvey white via Discuss
wrote:
I’d love to give you the source files, but it’s an entire
project
with subdirectories, referring to other libraries of parts I’ve
designed. Since I think it’s a memory limit, a partial file
would
not help all that much.
How many parts are there? Eagle has a board-stats ULP. Then we know
how many parts we need to "simulate" to try and reproduce this.
How about you put say a resistor, led, and an sot23 transistor in a
schematic, switch-to-board, place them compactly, in schematic view
load another design, but DO NOT load the board. Now back in board
save-as with a new name, and then copy-paste your three components.
select everything, repeat doubling the number of components until
you've filled your allowed design area. Now run it through your
rendering pipeline. Now there are only three models that need to be
included.
OK. F*** it. I did that for you. Attached is a board with 192
instances
of a resistor-cap-sot23. Does this show the problem?
Roger.
--
** R.E.Wolff@BitWizard.nl ** https://www.BitWizard.nl/
<https://www.BitWizard.nl/> ** +31-15-2049110 **
** Delftechpark 11 2628 XJ Delft, The Netherlands. KVK:
27239233 **
f equals m times a. When your f is steady, and your m is going down
your a is going up. -- Chris Hadfield about flying up the space
shuttle.
_______________________________________________
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
The reason there is no error message is because you are passing the
resistor as a child of the green terminal but the green terminal module
doesn't do anything with children.
On Sun, 14 Jul 2024 at 17:05, nop head nop.head@gmail.com wrote:
Also components and their positions are normally part of a PCB definition,
so they get placed automatically.
On Sun, 14 Jul 2024 at 16:35, nop head nop.head@gmail.com wrote:
You are missing a semicolon after the green_terminal module.
On Sun, 14 Jul 2024, 16:29 Harvey white via Discuss, <
discuss@lists.openscad.org> wrote:
OK, I think that establishes that there isn't a memory limit, it's about
250 or so parts, but they are a bit complex. So let's not think that
it's a memory limit since the demo of the openSCADlib uses much the same
parts, and more of them
I think I have something, not related to parts limits.
the following two lines:
//translate([6.191,68.262,1.600]) rotate([0,0,270])
rotate([0,0,-90]) green_terminal(gt_5p08, 3) //X2 0
translate([40.640,70.802,1.600]) rotate([0,0,90])
smd_resistor(RES0603,"22"); //R2 22
when put in a module will demonstrate the problem. The code shows one
part, the terminal block. Comment out the terminal block and the
resistor pops in.
I've done the same thing with the original code that I made the board
with, and the next part past the green terminal doesn't get on the
board. No error messages, thank you.
The way the code is formatted (as an explanation) is that the position
from the board is the first translate, the rotation is from the board.
Another translation/rotation pair may happen to adjust each part
depending on how it was modeled.
I suspect that I'm not using the terminal block correctly.
I do have some requests for nopSCADlib, but perhaps those may be better
taken off list.
I too do the parts that I need, but I'm doing more microprocessor stuff
on custom boards. I'm also doing a PNP machine which got modeled in
OpenSCAD first, I have an algorithm that is particularly useful for
moving stepper stage.
Harvey
On 7/14/2024 4:16 AM, nop head via Discuss wrote:
I have never hit any memory limits with NopSCADlib and many parts. One
of its tests has a single instance of all the parts in the library. My
biggest design has about 2500 parts, maily fasteners, rather than
electronic components but I don't see why they would be different.
Regarding the components available I only model what I need for my own
designs. Pull requests are welcome.
On Sun, 14 Jul 2024 at 08:02, Rogier Wolff via Discuss
discuss@lists.openscad.org wrote:
On Sat, Jul 13, 2024 at 10:24:22AM -0400, Harvey white via Discuss
wrote:
I’d love to give you the source files, but it’s an entire
project
with subdirectories, referring to other libraries of parts
I’ve
designed. Since I think it’s a memory limit, a partial file
would
not help all that much.
How many parts are there? Eagle has a board-stats ULP. Then we know
how many parts we need to "simulate" to try and reproduce this.
How about you put say a resistor, led, and an sot23 transistor in a
schematic, switch-to-board, place them compactly, in schematic view
load another design, but DO NOT load the board. Now back in board
save-as with a new name, and then copy-paste your three components.
select everything, repeat doubling the number of components until
you've filled your allowed design area. Now run it through your
rendering pipeline. Now there are only three models that need to
be
included.
OK. F*** it. I did that for you. Attached is a board with 192
instances
of a resistor-cap-sot23. Does this show the problem?
Roger.
--
** R.E.Wolff@BitWizard.nl ** https://www.BitWizard.nl/
<https://www.BitWizard.nl/> ** +31-15-2049110 **
** Delftechpark 11 2628 XJ Delft, The Netherlands. KVK:
27239233 **
f equals m times a. When your f is steady, and your m is going down
your a is going up. -- Chris Hadfield about flying up the space
shuttle.
_______________________________________________
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
The board I have designed has rounded corners, but a different radius
(including none) on each corner. I could not figure out how to generate
a board using your library.
This gives rise to the situation I used.
I could not figure out how to do all the parts using your library,
specifically, a 5 pin chip SOT23_5 (74LVC1G125DBG) It's a single
section of a 74LS125.
Still work to be done, I need to put in the milling layer.
Harvey
On 7/14/2024 12:05 PM, nop head via Discuss wrote:
Also components and their positions are normally part of a PCB
definition, so they get placed automatically.
On Sun, 14 Jul 2024 at 16:35, nop head nop.head@gmail.com wrote:
You are missing a semicolon after the green_terminal module.
On Sun, 14 Jul 2024, 16:29 Harvey white via Discuss,
<discuss@lists.openscad.org> wrote:
OK, I think that establishes that there isn't a memory limit,
it's about
250 or so parts, but they are a bit complex. So let's not
think that
it's a memory limit since the demo of the openSCADlib uses
much the same
parts, and more of them
I think I have something, not related to parts limits.
the following two lines:
//translate([6.191,68.262,1.600]) rotate([0,0,270])
rotate([0,0,-90]) green_terminal(gt_5p08, 3) //X2 0
translate([40.640,70.802,1.600]) rotate([0,0,90])
smd_resistor(RES0603,"22"); //R2 22
when put in a module will demonstrate the problem. The code
shows one
part, the terminal block. Comment out the terminal block and the
resistor pops in.
I've done the same thing with the original code that I made
the board
with, and the next part past the green terminal doesn't get on
the
board. No error messages, thank you.
The way the code is formatted (as an explanation) is that the
position
from the board is the first translate, the rotation is from
the board.
Another translation/rotation pair may happen to adjust each part
depending on how it was modeled.
I suspect that I'm not using the terminal block correctly.
I do have some requests for nopSCADlib, but perhaps those may
be better
taken off list.
I too do the parts that I need, but I'm doing more
microprocessor stuff
on custom boards. I'm also doing a PNP machine which got
modeled in
OpenSCAD first, I have an algorithm that is particularly
useful for
moving stepper stage.
Harvey
On 7/14/2024 4:16 AM, nop head via Discuss wrote:
I have never hit any memory limits with NopSCADlib and many
parts. One
of its tests has a single instance of all the parts in the
library. My
biggest design has about 2500 parts, maily fasteners, rather
than
electronic components but I don't see why they would be
different.
Regarding the components available I only model what I need
for my own
designs. Pull requests are welcome.
On Sun, 14 Jul 2024 at 08:02, Rogier Wolff via Discuss
discuss@lists.openscad.org wrote:
On Sat, Jul 13, 2024 at 10:24:22AM -0400, Harvey white
via Discuss
wrote:
> 8.
>
> I’d love to give you the source files, but it’s an
entire project
> with subdirectories, referring to other libraries
of parts I’ve
> designed. Since I think it’s a memory limit, a
partial file would
> not help all that much.
How many parts are there? Eagle has a board-stats ULP.
Then we know
how many parts we need to "simulate" to try and
reproduce this.
How about you put say a resistor, led, and an sot23
transistor in a
schematic, switch-to-board, place them compactly, in
schematic view
load another design, but DO NOT load the board. Now back
in board
save-as with a new name, and then copy-paste your three
components.
select everything, repeat doubling the number of
components until
you've filled your allowed design area. Now run it
through your
rendering pipeline. Now there are only three models
that need to be
included.
OK. F*** it. I did that for you. Attached is a board
with 192
instances
of a resistor-cap-sot23. Does this show the problem?
Roger.
<https://www.BitWizard.nl/>
<https://www.BitWizard.nl/ https://www.BitWizard.nl/>
** +31-15-2049110 **
** Delftechpark 11 2628 XJ Delft, The Netherlands. KVK:
27239233 **
f equals m times a. When your f is steady, and your m is
going down
your a is going up. -- Chris Hadfield about flying up
the space
shuttle.
_______________________________________________
OpenSCAD mailing list
To unsubscribe send an email to
discuss-leave@lists.openscad.org
_______________________________________________
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
You can model any shape board by specifying a polygon.
[image: image.png]
This one was done like this:
Blinder = pcb("Blinder", "PCB Blinder", [42.0, 45.0, 1.6], colour =
grey(90), hole_d = 3.2, holes = [[28.00096, 8.001]],
parts_on_bom = true,
polygon = rounded_polygon([
[-21.0, -22.5, 0],
[ 11.0, -22.5, 0],
[ 13.0, -13.5, 2],
[ 21, -11.5, 0],
[ 19, 20.5, -2],
[-19, 20.5, -2],
]),
components = [
[21.5, 34.9625, 90, "smd_cap", CAP0805, 0.8, "100n"],
[6.513, 14.6185, 90, "smd_cap", CAP0805, 0.8, "220n"],
[23.15, 4.0775, -90, "smd_cap", CAP0805, 0.8, "100n"],
[13.625, 14.47, 90, "smd_cap", CAP0805, 0.8, "15n"],
[23.15, 9.1575, -90, "smd_cap", CAP0805, 0.8, "15n"],
[40.0 - led_pitch(LED3mm) / 2, 40.0, 180, "led", LED3mm, "green",
2.0, 5],
[11.72, 1.35, 180, "multiwatt11", "L6203", 3],
[5.37, 19.55, 0, "molex_hdr", 02, undef],
[27.0, 15.02, 180, "jst_ph", 3, false],
[37.0, 15.0, 180, "jst_ph", 3, false],
[18.34, 15.74, 180, "jst_ph", 2, false],
[15.022, 30.583, 90, "smd_res", RES0805, "10K"],
[15.0, 34.79, 90, "smd_res", RES0805, "10K"],
[15.022, 25.757, -90, "smd_res", RES0805, "10K"],
[12.72, 20.82, 180, "smd_res", RES0805, "10K"],
[21.5, 20.0, 90, "smd_res", RES0805, "10K"],
[35.0, 41.0, 90, "smd_res", RES0805, "2K2"],
[11.339, 14.597, 90, "smd_res", RES0805, "1K"],
[8.545, 14.581, 90, "smd_res", RES0805, "1K"],
[28.6442, 40.9983, 0, "button", button_6mm_7, true],
[18.45, 41.0, 0, "button", button_6mm_7, true],
[7.91, 25.9 + pcb_length(tiny_buck) / 2 -
pcb_holes(tiny_buck)[0].x, 90, "pcb", 1.2, tiny_buck],
[36.668, 27.277, 0, "pcb", 0, ESP_12F],
],
grid = [18.0, 20.82, 1, 6, silver, inch(0.1), inch(0.1)
]);
To model an SOT23_5 I would copy the SOT23_6 and add an extra parameter to
allow missing legs to be listed.
On Sun, 14 Jul 2024 at 17:49, Harvey white via Discuss <
discuss@lists.openscad.org> wrote:
The board I have designed has rounded corners, but a different radius
(including none) on each corner. I could not figure out how to generate
a board using your library.
This gives rise to the situation I used.
I could not figure out how to do all the parts using your library,
specifically, a 5 pin chip SOT23_5 (74LVC1G125DBG) It's a single
section of a 74LS125.
Still work to be done, I need to put in the milling layer.
Harvey
On 7/14/2024 12:05 PM, nop head via Discuss wrote:
Also components and their positions are normally part of a PCB
definition, so they get placed automatically.
On Sun, 14 Jul 2024 at 16:35, nop head nop.head@gmail.com wrote:
You are missing a semicolon after the green_terminal module.
On Sun, 14 Jul 2024, 16:29 Harvey white via Discuss,
<discuss@lists.openscad.org> wrote:
OK, I think that establishes that there isn't a memory limit,
it's about
250 or so parts, but they are a bit complex. So let's not
think that
it's a memory limit since the demo of the openSCADlib uses
much the same
parts, and more of them
I think I have something, not related to parts limits.
the following two lines:
//translate([6.191,68.262,1.600]) rotate([0,0,270])
rotate([0,0,-90]) green_terminal(gt_5p08, 3) //X2 0
translate([40.640,70.802,1.600]) rotate([0,0,90])
smd_resistor(RES0603,"22"); //R2 22
when put in a module will demonstrate the problem. The code
shows one
part, the terminal block. Comment out the terminal block and the
resistor pops in.
I've done the same thing with the original code that I made
the board
with, and the next part past the green terminal doesn't get on
the
board. No error messages, thank you.
The way the code is formatted (as an explanation) is that the
position
from the board is the first translate, the rotation is from
the board.
Another translation/rotation pair may happen to adjust each part
depending on how it was modeled.
I suspect that I'm not using the terminal block correctly.
I do have some requests for nopSCADlib, but perhaps those may
be better
taken off list.
I too do the parts that I need, but I'm doing more
microprocessor stuff
on custom boards. I'm also doing a PNP machine which got
modeled in
OpenSCAD first, I have an algorithm that is particularly
useful for
moving stepper stage.
Harvey
On 7/14/2024 4:16 AM, nop head via Discuss wrote:
I have never hit any memory limits with NopSCADlib and many
parts. One
of its tests has a single instance of all the parts in the
library. My
biggest design has about 2500 parts, maily fasteners, rather
than
electronic components but I don't see why they would be
different.
Regarding the components available I only model what I need
for my own
designs. Pull requests are welcome.
On Sun, 14 Jul 2024 at 08:02, Rogier Wolff via Discuss
discuss@lists.openscad.org wrote:
On Sat, Jul 13, 2024 at 10:24:22AM -0400, Harvey white
via Discuss
wrote:
I’d love to give you the source files, but it’s an
entire project
with subdirectories, referring to other libraries
of parts I’ve
designed. Since I think it’s a memory limit, a
partial file would
not help all that much.
How many parts are there? Eagle has a board-stats ULP.
Then we know
how many parts we need to "simulate" to try and
reproduce this.
How about you put say a resistor, led, and an sot23
transistor in a
schematic, switch-to-board, place them compactly, in
schematic view
load another design, but DO NOT load the board. Now back
in board
save-as with a new name, and then copy-paste your three
components.
select everything, repeat doubling the number of
components until
you've filled your allowed design area. Now run it
through your
rendering pipeline. Now there are only three models
that need to be
included.
OK. F*** it. I did that for you. Attached is a board
with 192
instances
of a resistor-cap-sot23. Does this show the problem?
Roger.
--
** R.E.Wolff@BitWizard.nl ** https://www.BitWizard.nl/
<https://www.BitWizard.nl/>
<https://www.BitWizard.nl/ <https://www.BitWizard.nl/>>
** +31-15-2049110 **
** Delftechpark 11 2628 XJ Delft, The Netherlands.
KVK:
27239233 **
f equals m times a. When your f is steady, and your m is
going down
your a is going up. -- Chris Hadfield about flying up
the space
shuttle.
_______________________________________________
OpenSCAD mailing list
To unsubscribe send an email to
discuss-leave@lists.openscad.org
OpenSCAD mailing list
To unsubscribe send an email to
_______________________________________________
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
Thanks, didn't know how that worked.
Yet another rewrite coming. That'll handle milling layer stuff as soon
as I parse it out.
I will have to dig into your code and trap out the legs (for the missing
legs).
Harvey
On 7/14/2024 1:04 PM, nop head via Discuss wrote:
You can model any shape board by specifying a polygon.
image.png
This one was done like this:
Blinder = pcb("Blinder", "PCB Blinder", [42.0, 45.0, 1.6], colour =
grey(90), hole_d = 3.2, holes = [[28.00096, 8.001]],
parts_on_bom = true,
polygon = rounded_polygon([
[-21.0, -22.5, 0],
[ 11.0, -22.5, 0],
[ 13.0, -13.5, 2],
[ 21, -11.5, 0],
[ 19, 20.5, -2],
[-19, 20.5, -2],
]),
components = [
[21.5, 34.9625, 90, "smd_cap", CAP0805, 0.8, "100n"],
[6.513, 14.6185, 90, "smd_cap", CAP0805, 0.8, "220n"],
[23.15, 4.0775, -90, "smd_cap", CAP0805, 0.8, "100n"],
[13.625, 14.47, 90, "smd_cap", CAP0805, 0.8, "15n"],
[23.15, 9.1575, -90, "smd_cap", CAP0805, 0.8, "15n"],
[40.0 - led_pitch(LED3mm) / 2, 40.0, 180, "led", LED3mm,
"green", 2.0, 5],
[11.72, 1.35, 180, "multiwatt11", "L6203", 3],
[5.37, 19.55, 0, "molex_hdr", 02, undef],
[27.0, 15.02, 180, "jst_ph", 3, false],
[37.0, 15.0, 180, "jst_ph", 3, false],
[18.34, 15.74, 180, "jst_ph", 2, false],
[15.022, 30.583, 90, "smd_res", RES0805, "10K"],
[15.0, 34.79, 90, "smd_res", RES0805, "10K"],
[15.022, 25.757, -90, "smd_res", RES0805, "10K"],
[12.72, 20.82, 180, "smd_res", RES0805, "10K"],
[21.5, 20.0, 90, "smd_res", RES0805, "10K"],
[35.0, 41.0, 90, "smd_res", RES0805, "2K2"],
[11.339, 14.597, 90, "smd_res", RES0805, "1K"],
[8.545, 14.581, 90, "smd_res", RES0805, "1K"],
[28.6442, 40.9983, 0, "button", button_6mm_7, true],
[18.45, 41.0, 0, "button", button_6mm_7, true],
[7.91, 25.9 + pcb_length(tiny_buck) / 2 -
pcb_holes(tiny_buck)[0].x, 90, "pcb", 1.2, tiny_buck],
[36.668, 27.277, 0, "pcb", 0, ESP_12F],
],
grid = [18.0, 20.82, 1, 6, silver, inch(0.1), inch(0.1)
]);
To model an SOT23_5 I would copy the SOT23_6 and add an extra
parameter to allow missing legs to be listed.
On Sun, 14 Jul 2024 at 17:49, Harvey white via Discuss
discuss@lists.openscad.org wrote:
The board I have designed has rounded corners, but a different radius
(including none) on each corner. I could not figure out how to
generate
a board using your library.
This gives rise to the situation I used.
I could not figure out how to do all the parts using your library,
specifically, a 5 pin chip SOT23_5 (74LVC1G125DBG) It's a single
section of a 74LS125.
Still work to be done, I need to put in the milling layer.
Harvey
On 7/14/2024 12:05 PM, nop head via Discuss wrote:
Also components and their positions are normally part of a PCB
definition, so they get placed automatically.
On Sun, 14 Jul 2024 at 16:35, nop head nop.head@gmail.com wrote:
You are missing a semicolon after the green_terminal module.
On Sun, 14 Jul 2024, 16:29 Harvey white via Discuss,
discuss@lists.openscad.org wrote:
OK, I think that establishes that there isn't a memory
limit,
it's about
250 or so parts, but they are a bit complex. So let's not
think that
it's a memory limit since the demo of the openSCADlib uses
much the same
parts, and more of them
I think I have something, not related to parts limits.
the following two lines:
//translate([6.191,68.262,1.600]) rotate([0,0,270])
rotate([0,0,-90]) green_terminal(gt_5p08, 3) //X2 0
translate([40.640,70.802,1.600]) rotate([0,0,90])
smd_resistor(RES0603,"22"); //R2 22
when put in a module will demonstrate the problem. The code
shows one
part, the terminal block. Comment out the terminal
block and the
resistor pops in.
I've done the same thing with the original code that I made
the board
with, and the next part past the green terminal doesn't
get on
the
board. No error messages, thank you.
The way the code is formatted (as an explanation) is
that the
position
from the board is the first translate, the rotation is from
the board.
Another translation/rotation pair may happen to adjust
each part
depending on how it was modeled.
I suspect that I'm not using the terminal block correctly.
I do have some requests for nopSCADlib, but perhaps
those may
be better
taken off list.
I too do the parts that I need, but I'm doing more
microprocessor stuff
on custom boards. I'm also doing a PNP machine which got
modeled in
OpenSCAD first, I have an algorithm that is particularly
useful for
moving stepper stage.
Harvey
On 7/14/2024 4:16 AM, nop head via Discuss wrote:
> I have never hit any memory limits with NopSCADlib and
many
parts. One
> of its tests has a single instance of all the parts in the
library. My
> biggest design has about 2500 parts, maily fasteners,
rather
than
> electronic components but I don't see why they would be
different.
>
> Regarding the components available I only model what I
need
for my own
> designs. Pull requests are welcome.
>
>
> On Sun, 14 Jul 2024 at 08:02, Rogier Wolff via Discuss
> discuss@lists.openscad.org wrote:
>
> On Sat, Jul 13, 2024 at 10:24:22AM -0400, Harvey white
via Discuss
> wrote:
>
> > 8.
> >
> > I’d love to give you the source files, but
it’s an
entire project
> > with subdirectories, referring to other libraries
of parts I’ve
> > designed. Since I think it’s a memory limit, a
partial file would
> > not help all that much.
>
> How many parts are there? Eagle has a board-stats ULP.
Then we know
> how many parts we need to "simulate" to try and
reproduce this.
>
> How about you put say a resistor, led, and an sot23
transistor in a
> schematic, switch-to-board, place them compactly, in
schematic view
> load another design, but DO NOT load the board.
Now back
in board
> save-as with a new name, and then copy-paste your
three
components.
> select everything, repeat doubling the number of
components until
> you've filled your allowed design area. Now run it
through your
> rendering pipeline. Now there are only three models
that need to be
> included.
>
> OK. F*** it. I did that for you. Attached is a board
with 192
> instances
> of a resistor-cap-sot23. Does this show the problem?
>
> Roger.
>
> --
> ** R.E.Wolff@BitWizard.nl **
https://www.BitWizard.nl/ <https://www.BitWizard.nl/>
<https://www.BitWizard.nl/ https://www.BitWizard.nl/>
> <https://www.BitWizard.nl/
<https://www.BitWizard.nl/> <https://www.BitWizard.nl/
<https://www.BitWizard.nl/>>>
** +31-15-2049110 **
> ** Delftechpark 11 2628 XJ Delft, The
Netherlands. KVK:
> 27239233 **
> f equals m times a. When your f is steady, and
your m is
going down
> your a is going up. -- Chris Hadfield about flying up
the space
> shuttle.
> _______________________________________________
> OpenSCAD mailing list
> To unsubscribe send an email to
discuss-leave@lists.openscad.org
>
>
> _______________________________________________
> OpenSCAD mailing list
> To unsubscribe send an email to
discuss-leave@lists.openscad.org
_______________________________________________
OpenSCAD mailing list
To unsubscribe send an email to
discuss-leave@lists.openscad.org
OpenSCAD mailing list
To unsubscribe send an email to
discuss-leave@lists.openscad.org_______________________________________________
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org