non-socketed CPU

[IngoQ]

Yes, I understand. I hoped to find a way to make the current STE booster work. But all remaining options sound too risky and maybe it's not worth it.

Maybe someone else has some good ideas to add, and in the meantime my STE remains stock.

Thanks a lot nevertheless for all the input, and at least I now know exactly why it does not work as I intended

[exxos]

Its a good topic. Well worth talking about. I don't think there is any real solution to this problem though If anything it still needs a new booster PCB even if a socket is somehow fitted.. Just not enough hours in the day.

Only thing you could try is a PLCC socket upside down on the CPU , see how it fits ?

[Maeke]

Its a good topic. Well worth talking about. I don't think there is any real solution to this problem though If anything it still needs a new booster PCB even if a socket is somehow fitted.. Just not enough hours in the day.

Only thing you could try is a PLCC socket upside down on the CPU , see how it fits ?

I'm a bit concerned by the height of this solution, it may not fit under the shield, since almost all sockets he could find today are your funky plcc adapter, it means the booster will be higher than normal.

[exxos]

I'm a bit concerned by the height of this solution, it may not fit under the shield, since almost all sockets he could find today are your funky plcc adapter, it means the booster will be higher than normal.

Yes, but if he want to upgrade that style, he has no choice

Its just another reason why we need a new motherboard with proper space to install upgrades. Not STE of course, but have to start somewhere

[IngoQ]

I don't think ribbon cable is a good idea either.

For the sake of completeness: Why is a ribbon cable a bad idea, besides of being ugly? Signal degradation?

[Maeke]

you risk electromagnetic interferences inside the computer, as well as cutted wires overtime.

[exxos]

For the sake of completeness: Why is a ribbon cable a bad idea, besides of being ugly? Signal degradation?

I would say, it would be hard to do, look ugly. There could well be issues with gnd bounce as well.

I'm not saying it wouldn't work, Though its one of those things I wouldn't attempt to even try myself.

My MMU RAM upgrade uses a ribbon cable, it works well, but its only the databus there. Likely a better more direct route than on the motherboard.

I was mostly put off ribbon cables when I found my boosters wouldn't work with Marpet RAM upgrades, due to the huge long cables everywhere. The machine just wasn't stable.

Just a few milimeters of "track length" in the wrong place, and all chaos breaks lose.

Knowing how fickle the Atari circuit is, I would just generally advise not doing it. This isn't to say it won't work, just I would expect it wouldn't. Or knowing how difficult it would be to solder in the first place is a challenge in itself.

Though if you wanted to try, I have no objection

[IngoQ]

Would using a shielded ribbon cable help?

[exxos]

Would using a shielded ribbon cable help?

Probably not no. Problem with thin cables (IE screened) is the capacitance between the signal and shielding. You only need a few Pico Farads and it can kill a 32mhz signal for example.

I did a bit of work into that as "someone" recommended screened cable for the 16mhz clock for the falcon. May work with huge thick coax with a large gap between the signal and sheiling, but such a large cable on small pads..is a no no. I tried thin screened cabled, it crushed the 16MHz line (5volts) down to about half a volt! Granted this cable is several inches long. But I think it will cause more issues than it would solve.

Overall, you need thick cables or solder pins for the 5V and gnd above all else. If you keep everything short as possible, then it stands the best chance of working.

[IngoQ]

Okay thanks. For me not a way to go anyways, since it would most likely be too ugly, not to mention the soldering issues. And I surely don't want to do a lottery, since this would be most likely irreversible.

So to sum it up:

If you have a STE with SMT mount CPU, you are pretty much screwed.

If you remove your CPU and replace it with a PLCC socket, you COULD make a adapter PCB for the STE booster. You would need to adapt a PLCC plug to the "funky" staggered PGA layout originally used on the STEs CPU-socket (and therefore mirrored on the current STE booster). PLCC plugs are around 30-40 EUR, so this solution will be expensive. In addition the height of PLCC-socket + PLCC-plug + adapter PCB + Booster might not fit under the STEs shielding.

If one day there is a pluggable STE booster, you could just use that after socketing your CPU of course.

In any way, there is no no-soldering-solution, since the original CPU has to be removed.

Did I forget anything?