Note: known bugs, forgot to update the resistor values to their new locations after moving them from previous project. also noted one pad trace not routed. will follow up with update momentairily.

Alpha v3

Alpha v3.1

-Adjusted traces.
-Fixed non connected pad
-Updated resistor values

Edit: known bug need to flip resistor footprint to take vss off island

Alpha 3.14

-Found island on vcc breaking connection to resistor. fixed.
-Tweaked a few other things. minor stuff

Edit: known bugs
-Via not on VSS net
-Trace in pad on VSS net

might prototype this...

beta 3.141

-Went through and painstakingly aligned the holes best I could.

-Made a best guess for passive hole size and interference fit hole size.

- Adjusted traces and footprints

-Connected "tillamook mods" switch pads to resistors.

-Enlarged writing.

Known bugs:
Tillamook bf2 silkscreen label

Re: Idt Winchip 1, 2, 2a and 2b differences
some better pics of the winchip version of this concept. There is also a version with multiplier switches but I can't find a pic. but same idea with the pcb sandwich.

My digital caliper bit the dust a while back, I found a mechanical caliper though and took some measurements on my socket 5/7 collection the pins seem to universally measure about 0.017-0.018” The latest beta still is set to .020

I’m debating on wether or not the tinning process will add enough thickness to the hole to make it a good press fit. I think It just needs to be ordered to find out, here is to hoping all the holes are aligned to one another. In the schematic they seem to be right on the money. But how that translates to a real world product we’ll see.

Having a hell of a time trying to order these.

Here is how it looks on my end:

And here is what jlcpcb sees:

Idk what the heck is going on… and their “support” just closed after helping not a bit.

idk what is going on over at jlcpcb because the gerbers and pcb file look fine on my end... as you can see above^

A few very minor tweaks here are the latest files. didn't update version number but this is newer than last 3.141

fixed tilamook silk

adjusted traces for clearance

Beta 3.1415

Still can't figure out why JLCPCB is glitching out the render.

-Went through and solved some DRC complaints, in hopes it would solve the glitch. no luck.

-Trace routing changed a little

-Moved tillamook dip switch a tad.

Still no help from jlcpcb. They can't figure anything out. EVERYTHING looks fine on my end but as soon as I upload the file they say it's all messed up on their end. Days wasted dealing with this and no help from them. their website keeps glitching out also, causing many timed out chats and wasted orders. I try to just send the pcb file but they refuse to generate the gerber themselves, They would rather waste everyone's time than do something that makes sense. I also keep having to explain myself to the next agent about every ten minutes. Beyond infuriating.

Few minor tweaks to last file:

Edit:

did A experiment:

when I paste all into the same pcb file and upload it, they seem to upload fine.

when pasting these I moved them around on the grid. maybe that's all I need to do? move it to realign it?

still can't figure out why it shows drill holes in the wrong place at jlcpcb
I've gone through this with a fine tooth comb and nothing. either there is something I can't find or jlcpcb is glitching out.

If I paste one of the other project boards into the file and upload it, the drill holes go to the right spot, but transplanting this pcb into another file doesn't fix it. I'm really starting to think that this is a jlcpcb glitch.

In any case new version, -Minor changes made to mask for holes, moved to footprint, converted edge cuts to polygon, and did a fab outline

Okay, apparently either they fixed it in review or it’s just a glitch when uploading. Becuse once it goes into production all is fine.

wow, I need such a cool thing 😉

GREAT work! Looking forward to see first results in real 😀

Prototypes ordered should arrive soon! :p

The fit on this is pretty good, but I don’t think it’s tight enough for press fit. (Press fit may not be possible.) for now, I think it should be assumed that to install these you need to solder them on

I’m gonna at this point say that the latest version can probably be considered usable. And out of beta.

It occurs to me that I could probably make a tillamook variant that solders to the top so that you can still see the markings below if you choose to solder it

The plastic pentiums are unique in that they have a ridge at the bottom of the pin, which makes the tweaker float a bit.

Currently I don’t have the equipment to solder these on. So aside from inspecting and fitment checking that is all I will do at the moment. Also, I have hangups in regards to soldering to the pins of my cpus as a collector. But from what I can see I think this project is ready for members to start ordering if they wish.

Be sure to order .6 or .4 thickness.

Sphere478 wrote on 2022-05-22, 22:33:

Also, I have hangups in regards to soldering to the pins of my cpus as a collector.

That has always been my biggest concern with this project. It makes it much more work to reuse as well. Good to see the time and effort you have been putting in to learn what is needed for these projects though. I would have lost enthusiasm by now 😀

Good fitment

I've been wondering if, one could use something like this to actually lower the final clock of a CPU. Like adding a 2 and 4 divider for the FSB just before the clock hits the CPU. The benefit of this would be to make a Socket 7 CPU much better suited for older games. For instance, setting the FSB on the board to 50Mhz and then setting a divider of 2 between them to lower the clock to 25Mhz. Then a 2x CPU multiplier would give a 50Mhz final CPU speed. This on a Pentium CPU + disabling Brancg prediction and v pipeline would properly give a good match for a 486DX2-66. Taking it further, a divider of 4 would lower the CPU to 25MHz. Lower CPU speeds opposed to just disabling the cache has a smoother slowdown in my experience. Would be nice to have this in a way that 486 speeds would be achievable without the disadvantages of disabling the cache.