3D-Viewer: Soldermask-free regions should not be rendered as bare copper

Description

#3862 (closed) (probably) introduced unrealistic 3D preview as can be seen in the attachment. In reality, all soldermask-free copper regions comes plated, I haven't seen bare copper exposed on any simple pcbs.

Steps to reproduce

1. Create any PCB and observe bare copper.

KiCad Version

Application: Pcbnew

Version: (5.99.0-3256-gd52c475b2), release build

Libraries:
	wxWidgets 3.0.5
	libcurl/7.71.0 OpenSSL/1.1.1g (Schannel) zlib/1.2.11 brotli/1.0.7 libidn2/2.3.0 libpsl/0.21.0 (+libidn2/2.3.0) libssh2/1.9.0 nghttp2/1.41.0

Platform: Windows 8 (build 9200), 64-bit edition, 64 bit, Little endian, wxMSW

Build Info:
	Date: Sep  8 2020 21:05:27
	wxWidgets: 3.0.5 (wchar_t,wx containers,compatible with 2.8)
	Boost: 1.73.0
	OCE: 6.9.1
	Curl: 7.71.0
	ngspice: 32
	Compiler: GCC 10.2.0 with C++ ABI 1014

Build settings:
	KICAD_SCRIPTING=ON
	KICAD_SCRIPTING_MODULES=ON
	KICAD_SCRIPTING_PYTHON3=OFF
	KICAD_SCRIPTING_WXPYTHON=ON
	KICAD_SCRIPTING_WXPYTHON_PHOENIX=OFF
	KICAD_SCRIPTING_ACTION_MENU=ON
	BUILD_GITHUB_PLUGIN=ON
	KICAD_USE_OCE=ON
	KICAD_SPICE=ON

added priorityundecided statusnew labels

But that's usually what you find if you scratch off the soldermask ;-)

Aren't we going too far with the "realistic" vs "unrealistic" ?

This with is not easily implemented on the current 3D-Viewer or it would be so time consuming, both for development and as CPU calculations, that wont bring any joy.

Maybe it is easier to implement an option to disable non-plated/plated pad distinction.

I mean realistic rendering existed in all builds before today's build. v5.1 branch renders solder mask openings correctly.

Before today's build, there was no difference between plated and unplated copper.

A workaround would be to set the plated and unplated copper both to be the gold color

added 3d-viewer prioritywishlist labels and removed priorityundecided label

How does kicad decides which copper is plated or not?

From my experience:

1. "plated copper" = "any copper with void of soldermask";
2. "unplated copper" = "any copper covered by soldermask".

PCB manufacturers use this logic, and I think one would need to contact PCB manufacturer to explicitly tell which additional areas should be voided from plating process, and this will cost more. So "unplated copper" is a truly non standard option, and should not be seen on 3D viewer as default.

@KammutierSpule I think kicad decision of which copper should be treated as "plated" is wrong in recent build and needs to be fixed.

@jon Evans - I agree, there should be an additional color setting "Bare copper color" next to existing "Surface finish color..." inside 3D viewer settings.

Having both colors will enable us to have renderings of:

1. "Exotic PCB" (#3862 (closed) case with transparent soldermask): Bare copper traces rendered under clear soldermask + bare copper pads;
2. "Usual PCB": Bare color traces rendered under soldermask + gold-finished copper areas under soldermask openings;
3. "Homemade PCB": Bare copper traces + bare copper pads, without any soldermask rendering;

How does kicad decides which copper is plated or not?

At this moment, only plated pads are rendered. It was made that way because it was the easiest way (and faster to compute) to render plated/non-plated pads.

So this option should be called "render only plated pads as plated copper"

mentioned in merge request !407 (merged)

mentioned in issue #5582 (closed)

Just a side note about word meanings to be used:

1. Plated pad (or plated hole) - it is an internally metalized hole, having electrical contact between two layers, like a VIA;
2. Unplated pad (or unplated hole) - it is a hole without internal metalization, without contacts between layers, usually used as pcb mounting pads;
3. Plated copper (general term) - it is copper, which surface is treated in some way, like golden pads in the screenshots of this issue;
4. Unplated copper (or bare copper) - copper tracks on PCB right after etching process;

I am sure we are meaning the same, but these terms should not be mixed :)

yeah! Actually, what I mean with "plated pads" is: pads that has copper and are marked (checkbox on the layer option) to remove the soldermask (so they will get plated copper)

I agree with @WhiteChairFromIkea. I think it's too limiting to only have pads plated and not other exposed copper (traces, fills, etc.). Isn't the calculation needed for this very similar to the "Subtract soldermask from silkscreen" function (well, maybe union instead of subtraction)?

As it was implemented, it gets the information directly from the pads. Pads that has copper and are using SolderMask are considered to be plated (plated copper).

mentioned in issue #5627 (closed)

marked #5627 (closed) as a duplicate of this issue

marked this issue as related to #5627 (closed)

mentioned in issue #5716 (closed)

marked #5716 (closed) as a duplicate of this issue

marked this issue as related to #5716 (closed)

mentioned in issue #5639 (closed)

Hey, I'm not totally sure I follow some of the above discussion, but this behavior does seem to be pretty unpredictable in general. Look at this screenshot of a connector - some pads are rendered in the copper color and others in the surface finish...

Could you check the layer options on those pads if they are equal? Does it render same thing on different zooms or if you place the same component on a new (eg smaller) board?

@alexisvl @KammutierSpule It also happened to me because you optimized the pads stack before filling the zones. Try to see if it returns to normal, on Pcbnew at the menu

Tools/Remove Unused Pads...

Enable Pads and Reset unused

Obviously something is not working properly

!407 (merged) fixed this issue, thank you @KammutierSpule :)

closed

No! .. sorry, !407 (merged) didn't fixed the issue.. it just added an option to control it and now by default it will render as before this plated feature was implemented.

So now, you will need to tick the checkbox to enable this issue :)

changed milestone to %6.0 Feature Freeze

mentioned in issue #9637 (closed)

changed milestone to %7.0

reopened

Since this is not addressed, we'll re-open and target to v7

marked #9637 (closed) as a duplicate of this issue

marked this issue as related to #9637 (closed)

This will be difficult to implement/full fix this issue.

As in the original example image it would need to get all solder mask layer and calculate the intersection with the copper layer. This will also result on long board CPU calculations.

The pads are easy to display because there is plated/unplated information on pad design, so the pad can be created directly with the appropriate render material.

I believe there is already code to generate the full solder mask polygon set which should result in the exact solder mask coverage in all cases. The tricky part is going to be calculating the exposed copper and applying the plating material color to it. I agree that this is most likely going to be painfully slow so it probably should be a preference setting so that it can be disabled.

There are currently some issues with pads and vias depending on the solder mask settings that can result in incorrect copper plating rendering so this solution would resolve those as well.

We do have commercial users who design boards for their customers who would like to be able to provide accurate 3D images of the final product.

Note that for true accuracy you have to post-process to the soldermask layer as a whole to remove areas that are less than min-web-width. You can find code that does this in PlotSolderMaskLayer() and in DRC_TEST_PROVIDER_SOLDER_MASK::buildRTrees().

added statusaccepted label and removed statusnew label

removed milestone %7.0

It looks like this would require significant work, and there has only been a single up-vote in 3 years. Closing for now....

closed

reopened

Meh... I ended up implementing it after all....

assigned to @jeffyoung

changed milestone to %8.0

closed with commit c1a988e3

added statusfix-committed label and removed statusaccepted label