Projects
Some projects saw the light of day again out of necessity, though not many.
memholder
This week, the memholder OpenSCAD file got major improvements. On October 8th, I officially started to produce these designs for use at 2nd Life Inc. after pointing out that the current method of storing memory modules, just tossing them in labeled bins is inefficient and most often leads to damage. This 3D printed design I have calculated to be at a significantly lower cost to produce myself than to buy equivalent memory shipping trays online. From my testing, PLA seems to be the optimal resin, further reducing costs over using PETG or ABS.
After all this time, I found out that DDR4/DDR5 SODIMMs are slightly longer than DDR/DDR2/DDR3 SODIMMs, which always had me wonder why DDR4 SODIMMs did not fit in racks that fit 68mm length modules or DDR3 modules were loose in racks that were 70mm length. I added the option to extend the edge walls a little bit to fit DDR/DDR2/DDR3 SODIMMs while being stack-compatible with the DDR4/DDR5 version. As a result of having memholders incompatible with certain SODIMM generations, for the run I am doing for 2nd Life Inc, I am going to have the racks color coded; SDRAM would be orange, DDR orange, DDR2 would be green, DDR3 would be red, DDR4 would be blue and DDR5 would be magenta (if we even have any DDR5 apart from that one SK hynix SODIMM). This color scheme is based off the former RAMList pages that I had on this website.
On October 11, I added an extension to the "front" wall brace that provides a surface for a label. Currently I have this set to 62mm wide and 29mm high to accommodate for a Brother DK-1209 label, which we have many rolls of at 2nd Life.
I am definitely going to be pushing this Creality Ender 3 V2 printer to its limits; this printer is most likely not intended for the manufacturing scale I intend for this project but it is all I have to work with for now.
On October 12, I have started to consider returning to Build, RVA to continue manufacturing these with equipment better suited for this scale as I already have the Ender 3 V2 at its limits and I need much more of these made. I have not been there since I started working on MonitorCow MNC-N156KME-GNA, that external monitor project made from MDF boards and an Innolux N156KME-GNA that I never got around to finishing with going back to college in January this year (then eventually dropping out entirely in May-June).
3D printing
Apart from what is already covered in the memholder section above, I plan to continue adding upgrades where seen fit on the Ender 3 V2.
One upgrade I considered since 2024, mentioned in the Week 51-52, 2024 blog post, was to switch out the USB micro port for an RS-232 connector and transceiver board, bypassing the CH340 USB-UART chip on the controller board. This would be a much more robust connection, going from the fragile USB micro connector that tends to disconnect at the slightest touch.
Workshop lab area
The area in the back of the garage which me and my father built in 2021, called "The Lab" has been left unused for a while now.
On October 10, in anticipation of having to stay in the lab as much as possible to 3D print more memholder racks, I brought down the VG2249 monitor and set it up in the lab. I put so much time, money and effort into finding a monitor with a decent Innolux panel, I might as well make use of it. I am practically going to be living in the lab for the time being, surrounded by the drone of Supermicro fans and stepper motors.
Steam Deck
The Steam Deck OLED finally started to see some use again after getting back onto the device to gift an online friend Hiveswap Friendsim (I know I was going to stop mentioning Homestuck in this blog but I just have so much respect for this game) back on September 27. I noticed that Terraria was quite playable on the controller. Ever since I stopped reading Homestuck on September 20th and stopped watching television on August 16th, I have not really had any form of entertainment, forcing myself to work during those times, which has shown to be unsustainable.
DDR3 Memory Overclocking System
On October 10, I decided to get a motherboard that was recommended to me for experimenting with seeing how far certain DDR3 DRAM ICs can be undervolted, specifically Qimonda DDR3. The board being used is an ASRock Fatal1ty Z97X Killer. Supposedly this board can allow VDIMM to be set somewhere close to 1.0v. Specifics about the DRAM voltage range is seemingly nonexistent online or in the manual so I guess I would find out when I receive the board. My guess it should be able to go to at least 1.2v. I later found out this board has a VDIMM range of 1.165-1.8v.
2nd Life Inc.
Monday, October 6
I continued to test DDR4 SODIMMs using the four-micro-desktop setup.
Being the first time in likely over a year, I set back up the 3D printer in the lab to print out prototype versions of the new memholder design.
Tuesday, October 7
At some point I was tasked with upgrading the memory and storage in a Dell Latitude 3520 for resale. I tried multiple SSDs, with none of them being recognized by the Linux kernel, I noticed in the BIOS setup that Intel Rapid Storage Technology was enabled, which would cause this to happen. For a single SSD in a client system like this, RST has no benefit and acts as yet another obstacle for those trying to use anything other than Windows.
Wednesday, October 8
I was given back an HP Elite X2 2-in-1 that I inspected previously to add an SSD and install Windows 11 for a customer. I spent a lot of time just trying to find the correct M.2 retaining screw where after digging through a bin of many tiny screws, I found something that worked. I had difficulty with the display eDP connector becoming disconnected every time I tried to put the screen back on. A quick solution I used was to put some tape over the connector to hold the display cable in.
Thursday, October 9
I was tasked to take apart over a dozen iMacs, pulling any RAM and HDD/SSDs still in them. As displays are useful, I set them aside instead of putting them back in, which does not do much as these units are going to scrap anyway. I put all of the RAM into a couple of the earlier SODIMM memholders, one for DDR2 and one for DDR3. Before I started to work on these iMacs, I made sure to get all of the inventory IDs for them, instead of using a laptop or taking pictures of all of the tags, I used a USB C-to-A adapter and connected a USB scanner directly to the Zenfone 9 which actually worked.
As described in the memholder section above, I found out that DDR4 and DDR5 SODIMMs are slightly longer than DDR3/DDR2/DDR SODIMMs, requiring me to make some changes to the design.
Friday, October 10
I do not have much from this day other than that I continued to test RAM.