"Spicy O'clock" project has started

alxlab

Active Tinkerer
Sep 23, 2021
287
305
63
www.alxlab.com
Hey JDW,

Sorry for the late response. Haven't really been on the forums that much of late.

My initial thoughts like you said is that PETG should be fine. Most sites quote anywhere from 70C-85C for the temperature where it will start to deform.

ASA and ABS would give a big margin at the ~100C mark before it softens.

1679550633506.png


I'd probably consider getting one done using SLS and 3201PA-F Nylon. It's good up to 147C and it only costs $1.48 USD + shipping on JLCPCB for 1 clip.

I probably order one at sometime in the future just to see what the tolerances are like compared to my Prusa i3 MK3S which is a FDM type printer. It probably won't be anytime soon though since it's not a priority.
 
  • Love
Reactions: retr01 and JDW

alxlab

Active Tinkerer
Sep 23, 2021
287
305
63
www.alxlab.com
Actually a quick follow up on FDM printing and why you see mostly PLA and PETG and not ABS, ASA or Nylon. It's mainly because of ease of printing. The different plastic tend to shrink more or less when they cool. If a plastic shrinks a lot then it can cause the part to curl at the edges when printing or even worse detach completely from the build plate. Also different materials tend to stick to a build platform better than other.

PLA has very little to 0 shrinkage as it cools and has good adhesion to most build platforms.
PETG has low shrinkage and pretty good adhesion to most build platform types. It's can actually stick so well to a glass build plate that it take bit of glass off when detaching a part.
ABS and ASA tend to shrinks a lot when cooling you need an enclosure for the 3D printer to keep the ambient temperature high enough to print with it.
Nylon is one of the trickiest since it shrinks a lot when cooled and tends to not stick to build plates very well so enclosure is a must.

FDM has it's own limitations / quirks when printing objects since builds from the bottom up in layers. Any overhangs grater that 45 to 60 degrees, depending on the material, will need addition support material to make sure the overhangs don't sag. Since I use an FDM printer I try to optimize the geometry of my 3D objects so they will not require any supports. This somwwhat limits the possible shapes that can be printed on an FDM printer with good results.

The SLS type printer JLCPCB has actually print the object in a bed of powder plastic and uses a laser to fuse the power together. Since the object is always supported the geometry limitations of FDM and even SLA printing are almost mostly gone. The surface does have a kinda grainy texture though compared to SLA.
 
  • Like
Reactions: JDW

JDW

Administrator
Staff member
Founder
Sep 2, 2021
1,290
1,142
113
53
Japan
youtube.com
The only other thing I noticed about the PLA-printed sample I received (again, see my earlier post for photos), was that the little "lip" that I believe is supposed to go below the edges of the 040 CPU is very small. I've colored that in red below...

1679554548986.png


I've not tested anything yet, but I must assume that "lip" to be the weakest part of the entire clip. All the force of the clip used to press the heatsink down atop the CPU will rely in the integrity of those two tiny lips. With a vinyl part, I would not have any doubt it would hold well over time. With lesser plastic, I am not so sure. Again, you need a fair amount of force to press down the heatsink hard against the CPU (with thermal paste in between), plus we must consider the impact of heat coming from the heatsink, which I believe will be close to 60°C even with a fan in some cases.
 

retr01

Senior Tinkerer
Jun 6, 2022
2,469
1
780
113
Utah, USA
retr01.com
Another thought...

After heat treatment, Carbon Fiber HTPLA (High-Temperature PLA) can withstand and maintain structure up to approximately 170 degrees Celsius since it is semi-crystalline (part amorphous and part crystalline).

So where does HTPLA's heat resistance come from? It comes from its semi-crystalline structure. Amorphous materials like PETG and ABS surrender their stiffness (and thus form) at glass transition (Tg), about 80 and 105 deg C, respectively. HTPLA's Tg is lower at about 55 deg C, but HTPLAs are semi-crystalline (part amorphous and part crystalline).
When amorphous, HTPLA loses stiffness at about 55 deg C, but heat-treating HTPLA transforms the material structure from mostly amorphous to partially crystalline. Crystalline structures retain useful stiffness up toward their melting temperature, which for HTPLA is about 170 deg C.
When heat-treating HTPLA, people are generally nervous that a part changes size during the heat-treating process and can even distort it. Scaling or size change is minimized significantly from around 2-3% with translucent HTPLA down to about 0.6% with Carbon Fiber HTPLA. Additionally, Carbon Fiber HTPLA experiences little to no distortion during heat treatment alleviating the concern of warping.
And...

The major advantages of carbon fiber over traditional materials like steel, aluminum, wood, and plastics include the following:
  • High stiffness and strength.
  • Lightweight.
  • Corrosion resistance.
  • X-ray transparency.
  • Low CTE (Coefficient of Thermal Expansion)
  • Chemical resistivity.
  • Thermal and electrical conductivity.
 
Last edited:
  • Like
Reactions: JDW

alxlab

Active Tinkerer
Sep 23, 2021
287
305
63
www.alxlab.com
Yeah HTPLA is kinda interesting. For me personally the 2x the price of PETG and ABS and the need to do the extra step post heat treating are not necessary for this application. If I needed something over 100C it would probably be the best choice for ease of printing.

Carbon fiber adds even more to the price of the plastic. It's also is highly recommended to use a wear resistant nozzle (additional cost and installation) to prevent destroying your printer's nozzle over time.

For printing the bracket on a FDM printer I'd still suggest printing with ABS/ASA if possible or use PETG. At a glass transition temperature of around ~80C, PETG should be more than enough for the clip.

If you're just going to use a third party to print it for you then might as well go with an superior industrial SLS printer and material if the price is reasonable.

For FDM printing PETG has other properties that make it goo for the forces involved in the application of a heatsink clip. PETG has excellent strength and flexibility without breaking or deforming. When I designed the heating clip I into account. Let's take a look at this picture:
1679597728131.png


The red circles are providing downward force onto the heatsink and CPU is providing and upward force against the clips circles in green.

The torque applied to clips in the green circles don't have much leverage so are unlikely to break. The walls of the clip are also thick to prevent them from bowing (that's not the only reason they won't bow easily), which if they could, could also allow the green clips to detach from the CPU. The wall thickness of the clip also helps prevent permanent deformation of the clip when installing and removing the clip.

Now the the downward force of the red point result in torque being applied at the where these levers attach to the heatsink represented by the small blue circles. These points are the most likely to deform/break. PETG being strong and flexible will prevent this from happening as long as there's continuous strands of plastic perpendicular to the force. So the orientation of printing the part on an FDM printer is for just for ease of printing, It's maximize the strength at points of stress.

Now you have noticed that the middle has a continuous strip of plastic unlike the other 4 smaller levers. This is to provide additional downward force in the middle of the heatsink is (since the die is there) through tension along with torque resistance. This also makes it very hard to snap at the ends.

You may be asking why I didn't do continuous strips for all of them if it's superior. Well, plastic is also an insulator so I'd like to cover as little as possible of the heatsink's surface to allow better heat dissipation.

Regarding heat concerns, the hottest part of the CPU will at the center of the CPU. The points of contact there are the red dots in picture which are small to help prevent direct heat transfer. If 60C is the max temperature when overclocking with a fan it should be fine.

Well that ended up being rather long, but I felt like talking a bit about the design decisions I make when I design a 3D printable part.
 
Last edited:
  • Like
Reactions: JDW and retr01

alxlab

Active Tinkerer
Sep 23, 2021
287
305
63
www.alxlab.com
The only other thing I noticed about the PLA-printed sample I received (again, see my earlier post for photos), was that the little "lip" that I believe is supposed to go below the edges of the 040 CPU is very small. I've colored that in red below...

I've not tested anything yet, but I must assume that "lip" to be the weakest part of the entire clip. All the force of the clip used to press the heatsink down atop the CPU will rely in the integrity of those two tiny lips. With a vinyl part, I would not have any doubt it would hold well over time. With lesser plastic, I am not so sure. Again, you need a fair amount of force to press down the heatsink hard against the CPU (with thermal paste in between), plus we must consider the impact of heat coming from the heatsink, which I believe will be close to 60°C even with a fan in some cases.

I wouldn't consider those clips marked in red as the weakest point. If it's PLA I'd be more worried about the little arms pushing down on the heatsink snapping off.

Regarding installation, you're not meant to forcefully snap on the clip with no tools. You're supposed to clip on one edge of the clip to the cpu and then bow outward the other with a screw driver so it can clip on the other edge. This is similar to how those old metal heatsink clips used to be installed. If the clip is made of PETG then it shouldn't deform the clip. PLA is more brittle but I did make the plastic pretty thick so maybe you'll have no problem bending it without breaking. You may mark the slot for the screw drive but that's cosmetic.

Now, will you be able to install and remove the clip hundreds of times without damaging it? Probably not, but these types of plastic clips aren't meant to be installed and removed that much.
 

nottomhanks

Tinkerer
Oct 31, 2021
92
34
18
Hey, just wanted to send shoutouts to @Kay K.M.Mods , @Drake @jcm-1 and @mrfahrenheit. I had Drake install OS Cons on my 575 board, Joe installed the Spicy and I’m running the 60ns VRAM. I have a heat sink and fan, running steady at 44 Mhz. Apple IIe card installed, Total Replay, BlueSCSI with 7.1, 7.5.5, 7.6.1 and 8.1 partitions. Prior to the OS-Cons and some cleanup, the machine would overheat and lock up after 30 minutes. I’ve been beating up on this for several hours and it’s rock solid. Will check serial ports/MIDI and floppy reliability shortly. I really liked @Action Retro ’s PPC card video. Very curious if you can run the IIe card and that PPC card at the same time, and also would OS9 run on the 601 or does it need the 604?
IMG_3277.jpeg
 

Attachments

  • IMG_3279.jpeg
    IMG_3279.jpeg
    1.7 MB · Views: 37
  • Love
Reactions: jcm-1 and Drake