I am absolutely unsure wether my concern is warranted here, but afaik all K6 CPUs use aluminium interconnects. If the diffusion barrier on those chips is not 100% liquid metal TIM may quickly become the end of those chips.

0.18 micron is AMD's "coppermine" but they went back to aluminum on Duron Morgan just to see if they could.

Afaik they made 0.18 micron Athlon Thunderbirds with both aluminim and copper interconnects. They started using copper in Dresden and continued aluminium in Austin.

Sphere478 wrote on 2022-03-17, 08:43:

I wonder how hard it would be to make a copper heat spreader and use liquid metal

Too bad I don’t have access to the dies used to make the old spreaders.

I wanted to make a copper spreader when I started with my K6 overclocking. I am lucky to have my own milling machine 😁

Sphere478 wrote on 2022-03-17, 08:43:

I wonder how hard it would be to make a copper heat spreader and use liquid metal

Too bad I don’t have access to the dies used to make the old spreaders.

Annealed copper is pretty soft, the thickness isn't great, so .... https://hackaday.com/2020/03/10/forming-sheet … d-printed-dies/

Nemo1985 wrote on 2022-03-17, 08:06:

Skalabala wrote on 2022-03-17, 07:46:

This is interesting, my one best friend is the owner of Thermal Grizzly 😁 Now I know what paste I am going to use 😜

Keep in mind that my thermal grizzly became dry in something like 2 years (I used it on my current pcs and when changed the cpu on my 10 years old laptop), while Artic Silver 5 was soft after like 6 years.

AS5 that was put on a CPU when it first came out would most likely still be fine today. I've never once seen it dry out

MX-5 is also not likely to ever dry out from what I can tell.

I once used some copper electrically conductive paste on a CPU and when I went to remove the heatsink after a few months, the liquid was all gone but the copper had stayed and it had formed a bond between the cooler and the heatsink. It came off ok, but it was a bugger to get the copper layer cleaned off.

Pretty sure it was the one made by MG Chemicals.

Skalabala wrote on 2022-03-17, 16:04:

Sphere478 wrote on 2022-03-17, 08:43:

I wonder how hard it would be to make a copper heat spreader and use liquid metal

Too bad I don’t have access to the dies used to make the old spreaders.

I wanted to make a copper spreader when I started with my K6 overclocking. I am lucky to have my own milling machine 😁

Oh cool, can you make us copper spreaders with custom etching?

cyclone3d wrote on 2022-03-17, 19:40:

I once used some copper electrically conductive paste on a CPU and when I went to remove the heatsink after a few months, the liquid was all gone but the copper had stayed and it had formed a bond between the cooler and the heatsink. It came off ok, but it was a bugger to get the copper layer cleaned off.

Pretty sure it was the one made by MG Chemicals.

Maybe they repacked copper antiseize compound 🤣

BitWrangler wrote on 2022-03-17, 20:23:

cyclone3d wrote on 2022-03-17, 19:40:

I once used some copper electrically conductive paste on a CPU and when I went to remove the heatsink after a few months, the liquid was all gone but the copper had stayed and it had formed a bond between the cooler and the heatsink. It came off ok, but it was a bugger to get the copper layer cleaned off.

Pretty sure it was the one made by MG Chemicals.

Maybe they repacked copper antiseize compound 🤣

I had the same thought 🤣

Has anyone tried it?

I'm going to try and put the IHS back on my 570. Here's how it performed with a bare core with a very basic Socket 7 heat sink with an upgraded fan and Arctic Silver 5 running Prime 95 for about an hour:

Sphere478 wrote on 2022-03-17, 22:16:

BitWrangler wrote on 2022-03-17, 20:23:

cyclone3d wrote on 2022-03-17, 19:40:

I once used some copper electrically conductive paste on a CPU and when I went to remove the heatsink after a few months, the liquid was all gone but the copper had stayed and it had formed a bond between the cooler and the heatsink. It came off ok, but it was a bugger to get the copper layer cleaned off.

Pretty sure it was the one made by MG Chemicals.

Maybe they repacked copper antiseize compound 🤣

I had the same thought 🤣

Has anyone tried it?

I have tried the copper anti-seize as well quite a few years ago. I think it performed ok.

I kinda wanna try the silvery anti-seize at some point.

I was thinking, anyone who is playing with the modified version has the thing to measure the pc power consumption?
It will measure the whole pc pwer consumption but will also give an idea of the watt (and then heat) used from cpu.
That would allow us to understand if the thermal paste is really useful or not.

Nemo1985 wrote on 2022-03-17, 23:48:

I was thinking, anyone who is playing with the modified version has the thing to measure the pc power consumption?
It will measure the whole pc pwer consumption but will also give an idea of the watt (and then heat) used from cpu.
That would allow us to understand if the thermal paste is really useful or not.

Amd appears to have used the same adhesive for both the 2+/3+

I have a at2atx

Nemo1985 wrote on 2022-03-17, 23:48:

I was thinking, anyone who is playing with the modified version has the thing to measure the pc power consumption?
It will measure the whole pc pwer consumption but will also give an idea of the watt (and then heat) used from cpu.
That would allow us to understand if the thermal paste is really useful or not.

Only if the better paste allows the voltage to be lowered.

I’ve been testing with the at2atx device and haven’t been able to get anything conclusive between a 2+ 570 at 400 2.15v and a 3+ modded 570 at 400 2.15v

The amperage is just jumping all over the place

I think variations in heatsink temperature are causing more of a fluctuation in amperage than the cache being enabled or disabled is.

It may be that the cache is always on, but just ignored?

cyclone3d wrote on 2022-03-18, 00:37:

Nemo1985 wrote on 2022-03-17, 23:48:

I was thinking, anyone who is playing with the modified version has the thing to measure the pc power consumption?
It will measure the whole pc pwer consumption but will also give an idea of the watt (and then heat) used from cpu.
That would allow us to understand if the thermal paste is really useful or not.

Only if the better paste allows the voltage to be lowered.

Well you've got a few million of what are effectively thermistors in the thing with an inverse temperature resistance relationship, heat go up, resistance go down, resistance go down, current go up, current go up, power consumption go up, heat get embiggened extra more largely.

BitWrangler wrote on 2022-03-18, 01:11:

cyclone3d wrote on 2022-03-18, 00:37:

Nemo1985 wrote on 2022-03-17, 23:48:

I was thinking, anyone who is playing with the modified version has the thing to measure the pc power consumption?
It will measure the whole pc pwer consumption but will also give an idea of the watt (and then heat) used from cpu.
That would allow us to understand if the thermal paste is really useful or not.

Only if the better paste allows the voltage to be lowered.

Well you've got a few million of what are effectively thermistors in the thing with an inverse temperature resistance relationship, heat go up, resistance go down, resistance go down, current go up, current go up, power consumption go up, heat get embiggened extra more largely.

So what you is saysing… is moarrrrr coooolder?

It would be pretty awesome if someone with a cnc machine could start cranking out copper IHS. Replacments we could make a limited edition K6-III+ 600 chip!! Liquid metal, big ole heatpipe heatsinks. Maybe even one of us could go nuts with some phase change.

I put the shield back on with Arctic Silver 5 and four dabs of super glue. The result was counterintuitive; all else was the same, but the temp was noticeably cooler. I then decided to swap to the best Socket 7 heat sink I have. Unfortunately, the super glue didn't hold (the shield stuck to the heat sink), so I'll eventually have to try something else. Since right now I only have super glue and JB Weld on hand (JB Weld is forever) I used super glue again for now. The larger heat sink I have cannot be used if the IHS is off since it has ridges that prevent it from being able to make contact with the CPU's core with the shield removed. No surprise that it ran cooler still with a larger heat sink and fan.

Back in about 2000, when internet commerce still sucked, local outlets just kept the more expensive and latest stuff, and I was particularly broke, I had to put a heatsink on a gold top CPU with no clips available. What I used then and which lasted for several years until decommissioned was using a "household cement" type glue. That stuff which is similar to rubber cement, and brands include UHU, Bostick etc. Not really found a great generic name for it where everyone knows what I'm talking about... Anyway, reason I picked that is I wanted some shrinkage when it set up. So I put the regular zinc oxide goop on the gold cap and "cement" all round the ceramic part and left it with weight piled on it overnight to "set up". So it worked, heatsink stayed on, but after a month or two the adhesive had turned an horrible brown color... but it held up, can't say if temps were great because it was socket 7, no thermistor, but didn't crash from heat that I know of. Then when I wanted the heatsink off again I scored the glue and twisted it off.... cleanup was horrible though, stuff made a mess, think I needed nail polish remover, but more than 10 years ago it had a decent amount of acetone.

Repo Man11 wrote on 2022-03-19, 02:42:

I put the shield back on with Arctic Silver 5 and four dabs of super glue. The result was counterintuitive; all else was the same, but the temp was noticeably cooler. I then decided to swap to the best Socket 7 heat sink I have. Unfortunately, the super glue didn't hold (the shield stuck to the heat sink), so I'll eventually have to try something else. Since right now I only have super glue and JB Weld on hand (JB Weld is forever) I used super glue again for now. The larger heat sink I have cannot be used if the IHS is off since it has ridges that prevent it from being able to make contact with the CPU's core with the shield removed. No surprise that it ran cooler still with a larger heat sink and fan.

J&b weld would probably work great for the corners. I can try it.