TR Forums

This is the most horrible design EVER conceived. I was building a system for work and it was HELL getting it secured and then off so i could use a screw mounted. one.

I had a $30 or $40 tower heatsink that I was trying to get on. Ended up totally destroying one corner which = expensive heatsink collecting dust. Ageed, whoever designed that should be shot. Nahhhhh, tar'd, feathered, drawn and quarted

Defnitely a textbook case of how not to design a fastening system. Right up there with the Socket A clip that chipped CPU cores.

Glad to hear I'm not a complete dolt for having needed at least 15 minutes to install one.

Yeah , I think everyone can agree the 775 push pins are a bad design.

The actual pushpin process sucks, to be sure, and I can't believe they continue to use it. OTOH, having the holes in a perfect square is great because that means any heatsink can be oriented in any direction the builder desires.

When most good cases these days now come with a cutout, I wouldn't waste my time with any heatsink that didn't use a real backplate. Pulling out a board isn't that big a deal for a regular DIYer for the cases that don't.

Backplates ensure a proper fit and the heatsink is lot less likely to become a metal slug of death ruining the insides when you transport the computer.

OTOH, having the holes in a perfect square is great because that means any heatsink can be oriented in any direction the builder desires.

Hole positioning is independent of push pins and you're still limited to right angles. Truly good coolers work with many sockets and can be mounted in all four 90 degree positions anyways, even on rectangular mounts like AM2/3/etc.

When it works it's not so bad actually. Pleasantly easy even. The problem lies in the materials and precision. It fatigues and breaks easily which is where you run into issues. Try applying the same cooler a few times and it will get progressively harder. I wonder how much better it would be if it was an all metal design, rounded off so not to gouge the pcb.

I actually just built my first S775 computer (putting one together out of cheap used parts), so I got my first experience with the push-pin cooler. Didn't have much trouble installing it, but it's frightening how much it bends the motherboard. It's also very loud. Think I'll be replacing it when I can.

Yeah so I was at work when I posted this. That's how pissed I was. I mostly have done AMD systems for home use. My last build was an i7 which didn't have them. For this latest build I decided (since it was for work) I would go Intel ( and that's pretty much what I always do for enterprise) the difference was this wasn't a server MB (in the 2U sense) that used passive mounts with active fans in the case. So this was my first "desktop" 775. My view is that reviewers really should have hit Intel on this. I'm sorry but that mounting mechanism for ANY CPU is the worse thing I have EVER seen and I've been building systems since I was 16. Using pure plastic for a CPU fan is utter BS. Period.

Basically (I know I know) ended up using a screw driver on the back of the MB to get the fan off and chipped one of the jumpers on the back at one of the soldering points. That b##$h is gone (ON A F@#@#@# SUPER MICRO BOARD). But it's still working. It's now in test so if the NB doesn't catch fire I'll be OK. I'm hoping the thing I chipped off was for PCIE since the onboard GPU seems to be fine.

Oh well we'll see.

bdwilcox wrote:

Defnitely a textbook case of how not to design a fastening system. Right up there with the Socket A clip that chipped CPU cores.

I've done Socket A without the heat speader before. The deal with that is simple. Make sure the heatsink is in the exact spot you need it to be before you latch it down. It can be a pain, but in relation to the 775 it's cake comparatively. I really would prefer that crap to what Intel did with 775. I was pissed because I've read every CPU article from Anand to Tom's and no one has mentioned it. It wasn't until I had problems that I decided to google it and there they were, many many people just as pissed as I was. That fastening mechanism is the worse thing EVER!!!!!!

Bauxite wrote:

Hole positioning is independent of push pins and you're still limited to right angles.

Well, yes. On both counts. Just trying to give poor Intel a morale boost, to say "see, your mounting system isn't ALL bad".

Truly good coolers work with many sockets and can be mounted in all four 90 degree positions anyways, even on rectangular mounts like AM2/3/etc.

Not my HDT-S1283. And in my experience, though I guess the mounting hardware could have changed in the last year or so, not the Hyper 212+, both of which get frequent recommendations around here.

It works out in my system, because I have a top-mounted exhaust fan, so the vertical air direction works out fine.

edit: whoa! quote tags auto-close if you screw up. Nice, but I nested instead of split.

derFunkenstein wrote:

, not the Hyper 212+, both of which get frequent recommendations around here.

That looks to be a screw down, which is what I got when I went to Microcenter. They had a Thermaltake at $9.99 but I should have gotten it for 6.99 according to the sticker, but I wasn't pissed because I needed that box up for testing on Sat and Sun. So 9.99 it was.

derFunkenstein wrote:

Bauxite wrote:

Hole positioning is independent of push pins and you're still limited to right angles.

Well, yes. On both counts. Just trying to give poor Intel a morale boost, to say "see, your mounting system isn't ALL bad".

Truly good coolers work with many sockets and can be mounted in all four 90 degree positions anyways, even on rectangular mounts like AM2/3/etc.

Not my HDT-S1283. And in my experience, though I guess the mounting hardware could have changed in the last year or so, not the Hyper 212+, both of which get frequent recommendations around here.

What? The Hyper 212+ can absolutely be mounted vertically or horizontally on AMD systems. I know it can. I did it. I'll take a picture of it if you want.

I never had any problems with Socket 775 push pins. For installation, you just have to put even pressure on pins that are diagonal from each other. For uninstallation, just use a regular screwdraw to draw out the plastic push pins. The heatsink should then come off effortlessly.

I think Socket A's level mechnism was poorly designed. It was never meant to deal with oversized after-market HSFs. To make matters worse, core on Socket CPU was completely exposed. It didn't take much to crack/clip the core. I have seen and heard too many horrors with Socket A HSF installations. The worse you can get with an improper Socket 775 installation is a broken push-pin on the HSF ($20-30 to replace). The motherboard ($100-150) and CPU ($100-$999) usually surivie. The thermal protection on Intel CPUs ensure that having HSF poor contact will not fry the chip.

Yeah when I got my 775 board I spent a good 20 mins trying to get the dang thing on. I would get 2 corners on, do the third and then one would pop out. After much cursing and screaming I finally got it on...replaced it at the first chance I got with a tower heatsink with a backplate, I just did not trust the pushpin heatsink to stay on at all. Only heatsink I hated more than that was the 3Dcpu crossdrilled aluminum heatsink that looked cool, but didn't cool worth a damn. The worst thing I id with that was get a 9k rpm 38X60mm delta screamer to try and make it not suck by ramming tons of air through it...all it did was make it suck slightly less and give me massive headaches after 30 mins on the computer.

derFunkenstein wrote:

Bauxite wrote:

Truly good coolers work with many sockets and can be mounted in all four 90 degree positions anyways, even on rectangular mounts like AM2/3/etc.

Not my HDT-S1283. And in my experience, though I guess the mounting hardware could have changed in the last year or so, not the Hyper 212+, both of which get frequent recommendations around here.

It works out in my system, because I have a top-mounted exhaust fan, so the vertical air direction works out fine.

There are good coolers, then there are truly good coolers Anything that presumes to dictate what direction you want the air to flow in your case is an incomplete design IMO.

212+ is a vertical or horizontal mount regardless of socket, older versions of the 212 (non plus or whatever) might not be.

If something that is often on sale around $20 can do it right (and still not suck as a HSF) then all the fancy boutique coolers should take note to at least meet that bar.

I did had to send off to get a part for some coolers I have (noctua) but it was free.

Both mounting mechanisms suck to be honest. The push pins are obviously horrible on your first go around, but I didn't think they were that big of a deal. You just have to pay attention, maybe even try glancing at the instructions for the recommendation technique. AMD still not using a square design drives me insane. Yes, a quality tower style heatsink should provide a mounting mechanism that will go in any direction. But, that's not always the case and a square design would remove the ability for a HSF manufacturer to screw up in that regard.

It all gets piled into my long list of the horrendous usability issues in computer component design. I'm talking stuff like front panel connectors still using pins, expansion card security latches, SATA connectors, I/O port shields, etc. Stuff that could be improves so easily but has just been neglected for years or clearly not done for the customers sake.

Skrying wrote:

... but has just been neglected for years or clearly not done for the customers sake.

It's for backward compatibility sake. I do wholly agree with you, though.

morphine wrote:

It's for backward compatibility sake. I do wholly agree with you, though.

They managed to include both an AC97 and HD Audio connector.