maxtherabbit wrote on 2022-09-16, 15:35:

Sphere I have no doubt you will be able to route this board - but how will you assemble it? Two overlapping through hole sockets seem impossible to hand solder...

I agree about this consern.

Feipoa is figuring single row pins may be the answer to this. Could work? 🤔 let’s see what Feipoa can do.

feipoa wrote on 2022-09-16, 11:24:

In terms of the current between VRM's Vout and GND thru R1/R2, expect around 10 mA. With Vout = 3.6 volts, and Iadj negligible, […]
Show full quote

In terms of the current between VRM's Vout and GND thru R1/R2, expect around 10 mA. With Vout = 3.6 volts, and Iadj negligible, then V=RI ==> I = V/R = 3.6/(R1+R2) = 3.6/(121+230) =~ 10 mA. There are caveats with going too low with these, as well as too high. Min load currents being one. I cannot remember them all, but from what I recall, it is best to use values in the approx. range of what is in the datasheet. Each manufacturer is slightly different here.

It has been many years since I built my prototype and I had tested several linear and switching regulators and landed on this MIC29302WT. I did some load tests with what might be a typical current to the CPU, and at 730 mA, VRM does not get hot, just a little warm. At 360 mA, was at about room temperature. I think that was after 10 minutes.

The trim pot should not be connected to Vcc5. The Input of the VRM should be connected to VCC5.

Dang, there's a lot of arrows. Isn't there an auto route feature?

It seems there is a auto routing feature. But I’ve always done it manually. I will see if it can help. I expect I will still have to tune the positions though. Which is probably 2/3 of the work.

I have finished manually checking each trace route. A real eye exercise! Things seem to be in order, however I didn't see a trace going to MEMW#. I also cannot verify GND, VCC, and VCC5.

For reasons discussed in PM, we will need to rotate the sockets 90 degrees counter-clockwise.

EDIT: Forgot to add this. There is one section of a trace that seems to go nowhere. Not sure what this little stub is for. See circled region attached.

That trace is fine. I was wondering if you were going to notice that 🤣.

It goes to a via. I had to move it away from the capacitor pad.

I’ll post pics of specific layers so you can see gnd, vcc3 and vcc5 later on.

Middle layers are gnd

And outer layers are the two vccs

I haven’t routed memw# yet. Where is it going? I can check the nets for it later but last time I looked I didn’t see it.

As you suggested I will also be rotating the overhang. (By rotating both sockets)

Yeah, if you don't want to rotate both sockets 90 degrees counter-clockwise, you can rotate the overhang 90 degrees clockwise.

Where does MEMW# go? It goes from the MEMW# header on the overhang to the MEMW# pin on PGA168 (grid location B16).

feipoa wrote on 2022-09-17, 20:09:

Yeah, if you don't want to rotate both sockets 90 degrees counter-clockwise, you can rotate the overhang 90 degrees clockwise.

Where does MEMW# go? It goes from the MEMW# header on the overhang to the MEMW# pin on PGA168 (grid location B16).

168, b16 is what I was wondering 😀 thanks!

I’ll see if I can get some more time in on this soon. Be patient, I’ll get it done 😀

I've been following this thread with interest. Hope you eventually get it fully working. I'd love to test my PGA168 SXL2 chips, and possibly even use one of them in a long term build 😀

No rush on this. We've all waited 5 years, so what's another few months or years. haha...

hehe, as long as it happens before I die of old age (or ill health!) I'll be happy 😀

Actually, I'd like to be able to run one in a 486 board as well, as I have more of them than I do 386 boards 😀

I forget things, but it does look like sphere is trying to run one on a 486 board.

I don’t have much to do with these as I don’t own one. I do have a board that may be able to run it though. 🤔

I’m just lending my kicad help to the project. Feiopa/other owners are making the calls/much of the design direction. 😀

Okay, aside from fixing the DRCs remaining (arrows) and cleaning up the track routing, is there anything else that needs changed that you see?

I assume that first image is bottom view and the middle image is top view? If so, could you point out pin 1 on PGA-168 and pin 1 on PGA-132 ?

feipoa wrote on 2022-09-18, 19:37:

I assume that first image is bottom view and the middle image is top view? If so, could you point out pin 1 on PGA-168 and pin 1 on PGA-132 ?

yep

Way ahead of ya, alread put a arrow to it about a hour ago :p haha

(Not in front of computer now)

I also changed up the way I was wiring the center caps. Gonna try floods.

Ahh, looks like you did the opposite rotation from what I suggested. It won't matter functionally, but if looks matter to anyone, best to have pin 1 on the top-right of your middle image (top down view). This is because you can see the markings on the trimmer, VRM, and TI-SXL2-66 in the same orientation. I'll let you decide.

feipoa wrote on 2022-09-18, 21:15:

Ahh, looks like you did the opposite rotation from what I suggested. It won't matter functionally, but if looks matter to anyone, best to have pin 1 on the top-right of your middle image (top down view). This is because you can see the markings on the trimmer, VRM, and TI-SXL2-66 in the same orientation. I'll let you decide.

See memw# trace? That’s why 😀 (confirm that is the correct pin btw. It should be it is the pin number you specified)

Looks like you have MEMW# on the PCB going to PGA-168 grid location B17, which is N/C. MEMW# on PCB should be going to PGA-168 grid location B16. You are one pin off.

So you weren't able to put pin one on the top right? In the current orientation, it will look like this: download/file.php?id=145322&mode=view

Idk when that happened but I have distinct memory of running that to b16…

Weird.. anyway. This is why Tripple checking is handy 😀

If I flip, it then memw# is going to have to run all the way to the other corner. Is this okay?

Yes, I think it running the length of the PCB is OK. The vast majority of people won't need MEMW#, except for some really unique boards. Of the dozen or so I've tested, I haven't needed MEMW#, nor needed create an external flush circuit (which is what MEMW# does inside the SXL2-G chips).

EDIT: Keep in mind that whoever does use MEMW# will be dangling a wire from the MEMW# header across the motherboard to their ISA slot.

Flipped it, and ran memw# all the way around.

Here is a preview of what I have been working on with the center caps also. (Not done)

It’s kinda funny, as much of a bowl of spaghettini as this board is in reality, there will actually only be one visible trace to the user. 😆

So I’m gathering that this is the last layout detail that you wanted added/fixed? If anything else speak now! :p no matter how small, speak up!

If that’s it though, only thing left is to clean everything up. Not much will change visually. Just slightly moving traces here and there one by one. Then going back and doing it again. And again until I’m happy with it all.

Incase my computer crashes here is a backup:

Sphere478 wrote on 2022-09-19, 05:04:

So I’m gathering that this is the last layout detail that you wanted added/fixed? If anything else speak now! :p no matter how small, speak up!

If that’s it though, only thing left is to clean everything up. Not much will change visually. Just slightly moving traces here and there one by one. Then going back and doing it again. And again until I’m happy with it all.

Incase my computer crashes here is a backup:

Downloaded the project - muuuuuuch easier to review directly in KiCAD than from images 🤣 (I have 15 years of designing PCBs as a profession). Few comments, that pops in my eye as per attached image;

1: GND pin in the middle has connections to the side, not leading anywhere. This can act as an antenna in high speed applications and is not recommended. Unless necessary for cooling, copper should only exist between 2 points (as a rule of thumb, at least). Somewhere you may want to keep it for shielding purposes, though. But if removed, it also makes the PCB more easy to understand 😀
2. the 90 degree angle on the trace next to pin Q6; should of course be 2x45 degrees as all the others are 😀