Question Installing Intel i9-10900K in FGA1200 socket: How much lever pressure is too much?

[LilAbner_45]

Hi all, I'm a newcomer to your forum. Thanx for being here!

My son and I are building up a gaming system. The motherboard has an FGA1200 socket and the processor is an Intel i9-10900K, 3.7GHz and all the rest.

We laid the CPU in the socket. Then closed the socket-frame over it, the front edge of the frame slid under the screw closest to us, and we began to push the lever down. Pushed harder and harder until I started to get nervous over the amount of pressure we were using. Never got closer than 1/2" from the lever head to the circuit board, then chickened out and let it up.

How much pressure should we be using? I don't want to bend something expensive and hard-to-find (the i9-10900K was both). Also don't want to bend anything on the socket itself.

We've looked at several Youtube videos on building up a gaming system, especially concentrating on the part where the guy lays the CPU in the socket and flips the lever. In most, it seems trivial: He just moves the lever down to the flat position, doesn't seem to be using much pressure, and it's done.

Am I doing something wrong? Or should I just take a deep breath, use both feet on the lever (exaggerating) and let what happens happen?

 

[LilAbner_45]

Well, my son just discovered in the little booklet that came with the motherboard, that we're supposed to leave the little plastic cover that came with the socket, ON the socket when we close it on the CPU and start pushing the lever down. I'm 66 years old, been doing digital hardware most of my life, and I had forgotten for the 1,437,539th time "When all else fails, read the directions".

So we did that, and the lever force seemed to be a little less, down in the acceptable range for a 1200-pin socket, so I'd guess. Pushed the lever all the way down, latched it under the other lever that was there for the purpose, and (I think) we're done. We were able to just lift the plastic cap off afterward, it was sitting there loosely.

I have to wonder how this IC/Socket combination works. The IC (the CPU) has no pins or solder balls, just a zillion tiny gold-plated pads, flat on the bottom. The socket seems to have an equal number of finger-like projections, at least that's what I think they are. My old eyes are having a tough time with them, a magnifying glass helps but still not enough.

Do those finger-like projections yield slightly, each individually, when a CPU pad presses down on them? And that's how the socket manages to make contact with 1,200 pads all at once and reliably?

Does the socket have no hinged, moving parts at all (except the lever and frame), nothing that moves except those fingerlike projection that bend slightly when the CPU is pressed onto them?

Sorry for the dumb newbie questions. I started doing digital circuits just before the 74LS series first came out (early 1970s), a vast improvement over the 7400 series, and my first processor was a Moto 6800, with a whopping 40 pins, and clock speeds measured in kilohertz. I've since used Pin-grid arrays, Ball-grid arrays, surface mounts, and ZIF sockets that closed little jaws on each pin. But 1,200 pins (pads, whatever) is new to me. The first computer I built had fewer than 1,200 pins in the entire computer.

Thanks for bearing with me, and letting a garrulous old man ramble while I come up to speed, however slowly.