There seems to be a large number of resistor arrays pulling signals to what I assume is ground. Can anyone confirm that this is some sort of signal termination? Anyone with a slotket able to read some values?

I don't have a scope but can read whatever you want from a fluke.

the resistors in purple (with the exception of the single one horizontal on the left) are around 68 ohm.

Excellent!

Your adapter looks pretty basic, should be an excellent model to study for a minimum requirements.

If you feel super adventurous values of all the components would be handy.

And if you are really bored, you could study where everything seems to be going. That may be kinda difficult though if it is more than 2 layer. Is it? I assume it probably is.

I suspect someone may pop up here soon who has actually worked on their own design for these slotkets. (There are at least two people around here who have.) And share with us the details of those components. Just like many of my other projects I’m starting from almost scratch here. So the usual process will probably follow, many alphas as design is researched, layout is researched. And once something seemingly possibly functional is taking form I’ll start calling it a beta, at some point I or someone here will prototype one of the betas, and more changes will probably be made based on the prototype. Depending on how everything looked there may be more prototypes. If all looks good then finally pi.0 will be released.

I can probably start an alpha for the edge connector and socket at some point soon, and map out the nets. In the meantime anyone willing to help, let’s try to understand fully what all those ICs are for and what they connect to.

Perhaps a crude alpha will come first, and once a basic design is formed we can start adding advanced features like those described in the socket 370 tweaker thread.

After a lot of thought I think I will probably start work on the slotket 370 (this one) first. And later on can apply what we learn here to the 370 tweaker.

I’m also thinking a onboard voltage regulator may happen on this one using a murata, or a pair of murata modules. As apposed to the usual strategy I have been using of making the power external. The reason for this is there is a fair amount of room available on a slotket. 😀

Also,

Thanks to wiretap for forwarding a socket datasheet link:

https://www.mouser.com/ProductDetail/TE-Conne … u5qjsQLew%3D%3D

I'll do anything for a 72 pin simm gerber 🤣. I can send this to you if you'd like a closer look. not doing anything over here. But I'll get some info as time permits.

attempting to make a slot 1 edge connector footprint from scratch here but something isn't adding up the measurements are close, but not exact when using the datasheet for a pentium 2 processor. do we have a more official slot reference? here is what I got so far:

dead space between pins should be .1 but it is .95
total height should be .191 but it is .188

this after setting spacing at .039
large pad at .043 x .09
small pad to .016 x .098

not sure where the error lies.. thoughts?

perhaps a agp reference would yield clues? I think they are the same connector pattern.

Talking with oxcats on discord, apparently he agrees that the datasheet is messed up.

I also found this agp datasheet but apparently I am told it also will not be right. (On phone atm or I would check)

He said he used a microscope to measure for his footprint.

Suggests cloning a existing board. Says driver impedance matching and crosstalk elemination will be critical.

Fun times.

Gave a link to his footprints for slot 1.
https://github.com/rigred/kicad-retro

I've got my project symbols and footprints here:
https://github.com/rigred/kicad-retro

Symbol, don't panic, but yes it's cursed.

Also 370 with full pin map.

(Though I've not done the 370 symbol yet - haven't needed it yet)
Still though, just load the pins into excel, sort them and use kipart to generate the symbol
- Then spend like 3 days laying all the pins down 😁

Slot-1 is under the processors folder.

Thanks oxcats! Any conditions to using this? Or free to use?

I’ve already started on my own 370 footprint. Holes are laid out. No nets though. That’s the fun part 😳. Lol will take a look at what you have though. Might be further along, save some time. Though, I already sense this will be a very very long project.

That slot edge footprint will definitely be handy!

Sphere478 wrote on 2022-10-15, 07:16:

Thanks oxcats! Any conditions to using this? Or free to use?

It's TAPR Open Hardware License
https://tapr.org/the-tapr-open-hardware-license/
https://github.com/rigred/kicad-retro/blob/main/LICENSE
Basically free to use as long as you don't claim you made it all yourself and don't remove/change the license to be closed/something else.
And just submit improvements/changes back to upstream.

I see, thinking it may be less trouble to just finish my own footprint than to tie the project to a license. I basically public domain all my work. Total open source, no strings. Lemme see what I come up with. May reconsider. 😀 appreciate it none the less.

Sphere478 wrote on 2022-10-15, 07:40:

I see, thinking it may be less trouble to just finish my own footprint than to tie the project to a license. I basically public domain all my work. Total open source, no strings. Lemme see what I come up with. May reconsider. 😀 appreciate it none the less.

That's the strangest response I've seen yet to the TAPR Hardware License.
Have you read it?
What in specific do you object to?

You should strongly consider putting a license on open source (free) hardware projects - precisely to prevent it from being abused in the public domain.
It ensures it (remains) free and cannot be 'claimed' by anyone as a private/copyright/etc and taken out of the public domain.
It also makes it obvious what someone can do with the design, and removes ambiguity.
Whereas if they find it in the public domain they may be unsure if it's perhaps leaked/stolen design where they might be infringing on some businesses rights/license.

For example if someone wanted to be evil:
They could apply for a patent on your existing released projects, copyright them and then claim you infringed on their patent and rights.
It sounds nuts for something this niche - but there are all too many grifters out there. I've had personal experience with this.

I just don’t want to deal with making sure I am not violating it. I’m not a lawer. If I don’t use it then I have peace of mind. I’m not upset about it, just making a decision for my project. I don’t see any problem with you doing it that way. 😀 I express my appreciation.

People are free to do what they want with my projects. I won’t stress myself trying to control software on the internet that I don’t make money on anyway. I give it to the community freely.

In that case sounds like TAPR is perfect fit for you.
It's a copyleft license that ensures such freedoms

https://en.wikipedia.org/wiki/Copyleft

Quick blurb

Licensing – Hardware
(OSHW): Best Practices for Open-Source Hardware: OSHW is sharing the design files for a piece of hardware for others to modify or make hardware from (including for commercial purposes). [www.oshwa.org]

Popular copyleft licenses include “free-software”: require that derivatives be licensed under the same terms
(CERN OHL) OPEN HARDWARE LICENCE, (Products). [www.ohwr.org]
(TAPR-OHL) Open Hardware License, (Products) TAPR grants permission for anyone to use the OHL as the license for their hardware project, provided only that it is used in unaltered form. [www.tapr.org]
(TAPR-OHL) Although it (OHL) does not prohibit anyone from patenting inventions embodied in an Open Hardware design, and of course cannot prevent a third party from enforcing their patent rights, those who benefit from an OHL design may not bring lawsuits claiming that design infringes their patents or other intellectual property. [www.tapr.org]

Basically it's really easy not to violate the TAPR
- you keep the used parts under TAPR
- you do whatever the fuck with it
- you don't sue me
- nobody who uses this sues anyone 😁

The alternative is to put everything under CC-BY-SA 4.0 which is how KiCad's built in footprints and symbols are licensed.
https://creativecommons.org/licenses/by-sa/4.0/
https://www.kicad.org/about/licenses/

I remember having the same kind of slotket as mr-spain, just modified to run Coppermines.

The reason why - it's because the slotket itself was made with Mendocino/66FSB Celeron only in mind. It did run P3 as high as 1000B from what I can remember - including a 1.1GHz CuMine out of a IPC Starnote laptop - but the main issue is it was very crash-prone on almost all of my three 440BX mainboards I've tested it with - ABIT BE6-II, ASUS P2B Rev1.02 (fyi, the Coppermines I used seem to not give a flying damn about 1.8v VCORE being the lowest on that board, lmao) and a Soyo 6BA +IV.

I'd love to find out what exactly is the culprit causing so much instability with both 100 and 133FSB chips across three mainboards when using the slotket.
That, and if I can find out what causes this instability, I can then fix the other slotket I have, some "PII CPU CARD REV1.1" that uses DIP switches for the Vcore. It's also been modded since it wasn't Coppermine compatible either.

http://www.thg.ru/cpu/19980514/print.html

B21 pin is high/low for selecting 66/100 fsb