VOGONS


First post, by kaputnik

User metadata
Rank Oldbie
Rank
Oldbie

Been modeling some ATX I/O shields for 3D printing the last couple of days, and posted about it in the "What retro activity did you get up to today?" thread. Was kind of surprised by the interest for the models amongst other members, assumed that everyone here owning a 3D printer already would have done something similar, but apparently not. Figured it might be an idea to post the models in a separate thread, might be useful to more of you guys 😀

Test printed the shields in ABS. PET should be good too. Haven't tried PLA, but suspecting it might be too brittle, the latching edges might break when installing/removing the shield.

There's been some valid concern about electromagnetic interference in the thread. The simplest way I can think of to combat that is to apply aluminum foil tape to the rear side of the shield, and fold the tape around the edges to make electrical contact with the case, if the contact made by connector shieldings etc isn't enough. Aluminum foil tape is really easy to work with, and readily available at any well stocked hardware store. Copper foil tape would probably be even better, since it's designed for this kind of purposes, but it's quite expensive, and probably not trivial to get hold of everywhere. There's also copper foil tape used to keep slugs out of flower pots etc that might work, which probably is cheaper and more readily available.

If you got any suggestions on how to improve the shields, please tell me 😀

Last edited by kaputnik on 2019-02-15, 11:17. Edited 6 times in total.

Reply 1 of 57, by kaputnik

User metadata
Rank Oldbie
Rank
Oldbie

The first model is a blank shield.

Printed it in ABS with good results. Had however to compensate for thermal expansion, did that by simply scaling the model in the slicer app, 100.3% worked well with my printer. Strange that the slicer software won't do this kind of compensation automatically by the way. Or maybe I just didn't let it cool down sufficiently before removing it from the heat bed.

Print the shield with the front side up, and let it slowly cool down to room temperature on the print bed before removing it, to minimize warping and thermal contraction. Used 0.15 mm layer thickness. Set infilling to solid.

Attached .stl file ready for slicing/printing, and the original Autodesk Inventor 2018 model for own modifications.

Edit: also added a blank version of the thicker and generally sturdier AT shield in the next post. It's 2 mm instead of 1. The modified latching edges works better in my opinion too, even if they're not as elegant. It might be better suited for machining cutouts. Would however still recommend the 1mm one if doing it by hand with a knife etc 😀

Thermalwrong did a variant optimized for PLA, can be found here.

blank1.PNG
Filename
blank1.PNG
File size
353.28 KiB
Views
3674 views
File license
Fair use/fair dealing exception
blank2.PNG
Filename
blank2.PNG
File size
331.33 KiB
Views
3674 views
File license
Fair use/fair dealing exception
blank3.jpg
Filename
blank3.jpg
File size
66.16 KiB
Views
3674 views
File license
Fair use/fair dealing exception

Attachments

  • Filename
    blank2mm0.1.zip
    File size
    165.97 KiB
    Downloads
    109 downloads
    File license
    Fair use/fair dealing exception
  • Filename
    blank0.3.zip
    File size
    156.24 KiB
    Downloads
    91 downloads
    File license
    Fair use/fair dealing exception
Last edited by kaputnik on 2019-02-19, 10:34. Edited 3 times in total.

Reply 2 of 57, by kaputnik

User metadata
Rank Oldbie
Rank
Oldbie

The second shield is customized for AT boards in ATX cases. Besides the premade cutout for the AT keyboard connector, there are cutout guides for two DB25 and two DB9 ports. There's also a guide for a PS/2 connector of the kind that's mounted in an expansion slot bracket. The connectors seems to have a somewhat standardized form factor. Nice to not have to waste an expansion card slot on that. Had to keep the guides quite narrow to avoid roughness on the front surface, but it's still not too hard to do the cutouts with an X-acto knife or similar.

Otherwise compared to the blank shield, I adjusted/improved the latching mechanism and upped the thickness from 1 to 2 mm to make the construction sturdier, since D-subs and PS/2 connector will be mounted directly on it.

The location of the PS/2 cutout might not be the most aesthetically pleasing. Wanted to keep it as far away from the mobo and anything it could collide with as possible. Tried putting it right above the AT connector first, but after finding an actual PS/2 connector in my stashes, I realized it would never fit there. Didn't want to remove a DB port cutout to make room for it either. Might however make an alternative version with only one DB25 cutout after doing some other stuff I got in mind, if there's interest 😀

The same printing suggestions applies here too, print the shield with the front side up, and let it slowly cool down to room temperature on the print bed before removing it, to minimize warping and thermal contraction. Used 0.15 mm layer thickness. Set infilling to solid. This model did however not need any scaling, dunno why.

Not gonna use this one myself, so I'm especially interested in feedback from actual users here 😀

Attached .stl file ready for slicing/printing, and the original Autodesk Inventor 2018 model for own modifications.

Edit: did a small cosmetic fix, moved the PS2 cutout into line with the upper row of D-sub ones. Updated the Inventor file and added an updated .stl. Left the old .stl since I haven't test printed the new one.

Edit2: Also uploaded a version with a bit higher and thus more flexible edges, that should work better with PLA, or if you generally have problems with edges breaking when installing. See this post.

Edit3: kuenzign added optional gameport and onboard PS/2 port cutout guides to the original design, clicky.

atio1.PNG
Filename
atio1.PNG
File size
379.89 KiB
Views
3674 views
File license
Fair use/fair dealing exception
atio2.PNG
Filename
atio2.PNG
File size
292.99 KiB
Views
3674 views
File license
Fair use/fair dealing exception
atio3.jpg
Filename
atio3.jpg
File size
108.54 KiB
Views
3674 views
File license
Fair use/fair dealing exception
atio4.jpg
Filename
atio4.jpg
File size
80.37 KiB
Views
3674 views
File license
Fair use/fair dealing exception

Attachments

  • Filename
    atio0.4.zip
    File size
    384.41 KiB
    Downloads
    222 downloads
    File license
    Fair use/fair dealing exception
Last edited by kaputnik on 2021-02-04, 21:29. Edited 3 times in total.

Reply 3 of 57, by kaputnik

User metadata
Rank Oldbie
Rank
Oldbie

Noticed that many 440BX boards and their counterparts with other chipsets from the same era seems to share a common connector layout. Would guess it was standardized in the beginning, but haven't really looked into it. Did a shield model with cutouts for this layout, found on two different 440BX and a 440LX board of mine. If I'm right, there should be a lot of boards out there using it.

Oddly enough, there seems to be a slight height difference between the PS/2 connectors on my different boards. Check photo, it's the board that deviates the most. Could move the holes if the pictured height is the more common.

Did variants with both the 1mm and 2mm base shields. Up to you to decide which one you like best 😀

Attached .stl files ready for slicing/printing, and the original Autodesk Inventor 2018 model.

440bx2.jpg
Filename
440bx2.jpg
File size
87.87 KiB
Views
3432 views
File license
Fair use/fair dealing exception
440bx1.PNG
Filename
440bx1.PNG
File size
369.89 KiB
Views
3432 views
File license
Fair use/fair dealing exception

Attachments

  • Filename
    440bx.zip
    File size
    491.91 KiB
    Downloads
    103 downloads
    File license
    Fair use/fair dealing exception
Last edited by kaputnik on 2019-02-17, 11:52. Edited 3 times in total.

Reply 4 of 57, by kaputnik

User metadata
Rank Oldbie
Rank
Oldbie

When the manufacturers started to include onboard sound, it seems like many of them kept the old common layout, and just added 3.5mm jacks and a game port to it. Even seen a board with a network connector added to the same layout, above the USB ports.

All cutouts except for the network one are premade on this model. Added cutout guides for a network connector if you got one, should be easy enough to open up cleanly, since all needed cuts are straight. Just do the cut closest to the USB ports first.

1mm and 2mm variants available here too.

Attached .stl files ready for slicing/printing, and the original Autodesk Inventor 2018 model.

intsound1.PNG
Filename
intsound1.PNG
File size
472.04 KiB
Views
3432 views
File license
Fair use/fair dealing exception
intsound2.PNG
Filename
intsound2.PNG
File size
377.69 KiB
Views
3432 views
File license
Fair use/fair dealing exception
intsound3.jpg
Filename
intsound3.jpg
File size
85.35 KiB
Views
3432 views
File license
Fair use/fair dealing exception

Attachments

  • Filename
    intsound.zip
    File size
    609.12 KiB
    Downloads
    88 downloads
    File license
    Fair use/fair dealing exception
Last edited by kaputnik on 2019-02-16, 15:58. Edited 1 time in total.

Reply 6 of 57, by brostenen

User metadata
Rank l33t++
Rank
l33t++

This is pure awesomeness.... Thanks for sharing. Love it

retardware wrote:

But: How to ensure electromagnetic shielding?

That is a thing of the past. Like before 1993 to 1996. Phones and anything else modern are sturdy enough. Unless of course, you have a first generation pacemaker or you are in a coma, hooked up to a respirator. I am always running my AT cases open, whenever I play. And have done that since like 1995, and I have never experienced any issues. Well... GSM cell/mobile phone signals went into any electronics back then. Even computers that had their cases closed. Ohhhh.... That tick-tick noise, whenever you were about to recieve a text message on an Nokia 3310. 🤣

Don't eat stuff off a 15 year old never cleaned cpu cooler.
Those cakes make you sick....

My blog: http://to9xct.blogspot.dk
My YouTube: https://www.youtube.com/user/brostenen

001100 010010 011110 100001 101101 110011

Reply 7 of 57, by kaputnik

User metadata
Rank Oldbie
Rank
Oldbie
retardware wrote:

Very nice and convenient idea!
But: How to ensure electromagnetic shielding? Metallic spray?

Seen plastic electronics enclosures coated on the inside with what I believe is some kind of copper or nickel paint. Can't imagine other than it's for RF shielding purposes. Might be possible to buy somewhere?

gdajacobs posted an interesting idea here too.

As mentioned in the OP, there's also conductive filament. It's kind of resistive though, so can't say for sure if it'll work, don't have sufficient knowledge about RF interference for that.

brostenen wrote:

This is pure awesomeness.... Thanks for sharing. Love it

My pleasure 😀 Oh, and stay tuned, got more coming, hopefully the next couple of days.

Reply 8 of 57, by appiah4

User metadata
Rank l33t++
Rank
l33t++

Already sent the AT io shield to a local 3d printer for five copies (four of which I will likely never use but why not..)

Retronautics: A digital gallery of my retro computers, hardware and projects.

Reply 9 of 57, by kaputnik

User metadata
Rank Oldbie
Rank
Oldbie
appiah4 wrote:

Already sent the AT io shield to a local 3d printer for five copies (four of which I will likely never use but why not..)

Did you send the updated .stl? Added it just before you posted, a minor cosmetic fix. If not, there might still be time to do it before he starts printing.

Reply 11 of 57, by appiah4

User metadata
Rank l33t++
Rank
l33t++
kaputnik wrote:
appiah4 wrote:

Already sent the AT io shield to a local 3d printer for five copies (four of which I will likely never use but why not..)

Did you send the updated .stl? Added it just before you posted, a minor cosmetic fix. If not, there might still be time to do it before he starts printing.

Yeah I sent the v0.4 STL; I should get it in a few days I think.

Retronautics: A digital gallery of my retro computers, hardware and projects.

Reply 12 of 57, by kaputnik

User metadata
Rank Oldbie
Rank
Oldbie
appiah4 wrote:
kaputnik wrote:
appiah4 wrote:

Already sent the AT io shield to a local 3d printer for five copies (four of which I will likely never use but why not..)

Did you send the updated .stl? Added it just before you posted, a minor cosmetic fix. If not, there might still be time to do it before he starts printing.

Yeah I sent the v0.4 STL; I should get it in a few days I think.

Updated the .zip with a 0.4fix.stl an hour ago. Try clearing the browser cache and redownload the .zip 😀

Maybe no biggie if you get the old version printed though, it's purely cosmetical, just moved the PS/2 cutout downwards a millimeter or two, into line with the upper row of D-sub cutouts.

Reply 13 of 57, by appiah4

User metadata
Rank l33t++
Rank
l33t++

Oh! I will send the update now.

Last edited by appiah4 on 2019-02-15, 04:49. Edited 1 time in total.

Retronautics: A digital gallery of my retro computers, hardware and projects.

Reply 14 of 57, by retardware

User metadata
Rank Oldbie
Rank
Oldbie

Regarding EMF, I would prefer not to gamble.
In all legislations I know of the EMF problem is regulated the way, that if anybody complains about it and the oversight authority finds that the EMF pollution at your place is excessive so it causes disturbances, they have the right to locate the source, find the culprit, and force him to stop their excessive emissions. In some jurisdictions the offending equipment even can be confiscated and its owner be fined. But in general, there is the rule that the offender has to pay the costs for locating the source. And these usually are in the 4-digit dollar/lbs/euro range.

It was 1987 or 1988 in Berlin (Germany). I was used to have my PC running open, like @brostensen described. One day when I came back home, my neighbor greeted me and told me that the German Bundespost (which did the radio oversight back then) was there and asked about PCs in the house, as a radio amateur living about 300 meters farther away had complained about the radio noise my computer caused.
My neigbor showed them his PC (closed case) and they found no emissions.
If I had been at home or my PC running that time, it would have been a nasty and expensive surprise for me.

For this reason I would advise to take the RF emissions lightly.

The copper/nicel plastics coatings found in many equipment's casing inner sides mitigate the emissions to some degree, but far less effectively as actual metal foil does.
For this reason I believe the most sensible and responsible way would be to design the inner side of the shield in a way that makes it easy to glue a copper foil to it and ensure conductive contact with the case all around. If the latter condition is not given, the shield's conductive coating could even make it effect as antenna.

Reply 15 of 57, by kaputnik

User metadata
Rank Oldbie
Rank
Oldbie
retardware wrote:
Regarding EMF, I would prefer not to gamble. In all legislations I know of the EMF problem is regulated the way, that if anybody […]
Show full quote

Regarding EMF, I would prefer not to gamble.
In all legislations I know of the EMF problem is regulated the way, that if anybody complains about it and the oversight authority finds that the EMF pollution at your place is excessive so it causes disturbances, they have the right to locate the source, find the culprit, and force him to stop their excessive emissions. In some jurisdictions the offending equipment even can be confiscated and its owner be fined. But in general, there is the rule that the offender has to pay the costs for locating the source. And these usually are in the 4-digit dollar/lbs/euro range.

It was 1987 or 1988 in Berlin (Germany). I was used to have my PC running open, like @brostensen described. One day when I came back home, my neighbor greeted me and told me that the German Bundespost (which did the radio oversight back then) was there and asked about PCs in the house, as a radio amateur living about 300 meters farther away had complained about the radio noise my computer caused.
My neigbor showed them his PC (closed case) and they found no emissions.
If I had been at home or my PC running that time, it would have been a nasty and expensive surprise for me.

For this reason I would advise to take the RF emissions lightly.

The copper/nicel plastics coatings found in many equipment's casing inner sides mitigate the emissions to some degree, but far less effectively as actual metal foil does.
For this reason I believe the most sensible and responsible way would be to design the inner side of the shield in a way that makes it easy to glue a copper foil to it and ensure conductive contact with the case all around. If the latter condition is not given, the shield's conductive coating could even make it effect as antenna.

Interesting that they managed to determine the interference source before even seeing it. Computers weren't really commonplace in every home back then either, PCs running unshielded could hardly have been a prime suspect in any such case.

Aluminum foil tape would probably work very well with the current design. The back side is absolutely flat, especially if printed on a smooth print bed. Aluminum foil tape is easy to apply without getting any creases, can be cut in a neat manner at connector cutouts etc, and should be easy enough to fold over the edge to establish an electrical connection with the computer case, if connector mantles etc won't do the trick alone. There's play enough in the recesses for it, they're 1 mm wide, plenty of space left over for the tape, most cases I've seen are made from 0.75 mm sheet metal. Since modern cases - the one used for testing in my pics for example - usually are fully painted, and must have been built with RF emissions in mind, my guess would be that it's not really needed though.

Reply 17 of 57, by gdjacobs

User metadata
Rank l33t++
Rank
l33t++

I asked earlier about conductive paint. Carbon is apparently used in various binders for commercial EMI suppression coatings. Someone wrote up an instructable where they MacGyvered their own versions and put them to the test.
https://www.instructables.com/id/Conductive-Paint/

Apparently mixing graphite and acrylic works okay and is extremely inexpensive.

All hail the Great Capacitor Brand Finder

Reply 18 of 57, by kaputnik

User metadata
Rank Oldbie
Rank
Oldbie

Added new models with cutouts for a couple of common connector configurations 😀

gdjacobs wrote:

I asked earlier about conductive paint. Carbon is apparently used in various binders for commercial EMI suppression coatings. Someone wrote up an instructable where they MacGyvered their own versions and put them to the test.
https://www.instructables.com/id/Conductive-Paint/

Apparently mixing graphite and acrylic works okay and is extremely inexpensive.

Would be interesting with some performance data and an expert opinion on it. I really don't know enough about RF shielding to assess how well this would work 😀

Reply 19 of 57, by gdjacobs

User metadata
Rank l33t++
Rank
l33t++

His testing is not really robust enough to know if volume resistivity is in the range required for EMI shielding, although testing would certainly be inexpensive to complete as the materials required are inexpensive. Based on his relative testing I suspect the home made stuff is somewhat effective for frequency ranges < 10mhz. It might require a more generous application as it's not optimized at all.

If you want a fully engineered and qualified product or something with a wider range of shielding frequency, it's probably worth looking into a commercial coating.

All hail the Great Capacitor Brand Finder