Madshrimps Forum Madness - belgiums first solid gpu tube's / blind 1.1

Madshrimps Forum Madness (https://www.madshrimps.be/vbulletin/)

- Hardware Overclocking and Case Modding (https://www.madshrimps.be/vbulletin/f10/)

- - belgiums first solid gpu tube's / blind 1.1 (https://www.madshrimps.be/vbulletin/f10/belgiums-first-solid-gpu-tube-s-blind-1-1-a-32959/)

Quote:

Originally Posted by blind_ripper (Post 144105)

keep on sleeping witte ;) , ones u start LN² il tell u my secerets :D .
water buffer is not know to most ppl but my cola trick is new and it works for me ^^

Does it keep the difference between idle and load even smaller?

I thought Hipro did an insane job keeping the delta at 5°C and lower :o

Quote:

Originally Posted by geoffrey (Post 144107)

Teach me master

A water buffer is used to lower the delta between idle and load temps when using ln² :)

Quote:

Originally Posted by Massman (Post 144110)

A water buffer is used to lower the delta between idle and load temps when using ln² :)

hhhmmm ask jort start bench @ -146°C when done -148°C or lower with the coca cola trick , whet the dragon evo the water buffer is not needed becaus the center pole is one off the best buffers .
witte when u start Ln² il teatch u , have a other studient ready for Ln² .
soon will strik with spi :) , we have new design for a solid copper cpu tube .
oyes the student is troman ^^

Hehe ... Steven really wants to get OPB, doesn't he? :D

Anyway, I wish both of you succes. What hardware are you going to use?

BTW: Isn't it normal that temps drop again at the end of the bench as the load decreases?

Quote:

Originally Posted by Massman (Post 144184)

euhm NO! only when bench is done ,
when the bench start u want to get 1°C off the coldbug , when bench is running first sec add ln² till u are @ 10°C under coldbug .
this hass to be done fast and dont any more so when the bench is done ul have ure starting temps or little lower .
this is where the cola trick dos the work , it lets the ln² boil so fast the temps drop 5°C in 2sec's ;) .
so yes if been doing some test :p, lol at the meeting when we had this i whase like omg . becaus normal ln² ddos not boils *** fast and temps dont drop so good , goal on my quad is running 06 @ 4.9ghz :)

will have all sort's off hw , P35 will be a good chipset and if DDR3 gives that nice boost will go for it with a E6700 and my old X6800 , the guy i sold it to will lend it to me ;) .
u see i learned more about ln² and stuff :p , QX6700 benches easy @ 5ghz with 1.65vcore for pi 1m all 4 core's unable . oyes that whas on the brokken P5W64 ws pro so thinking that il hgit more with new mobo !

Quote:

Originally Posted by blind_ripper (Post 144193)

The first part is what I meant: when the bench has ended, the temps start to drop again. I never knew you had to add ln² directly after the bench start to really control the bug :)

Quote:

Originally Posted by Massman (Post 144204)

The first part is what I meant: when the bench has ended, the temps start to drop again. I never knew you had to add ln² directly after the bench start to really control the bug :)

dont think , u will find out when u bench Ln² :)

Quote:

Originally Posted by blind_ripper (Post 144074)

hehe u think so ^^ , 1 ure design will fail to keep control off the temps and will use to mutch LN² .
also the load temps with controlled Ln² will suck so hard !
2 yes the slab off copper on the back will me removed , and steps around 2mm that makes 18 steps total .
now if u are sure ure design will do beter , just do it becaus im sure it will suck hard .
ln² containers are build to make the Ln² FLOW to where its needed and in this case the base , whet the steppers and maybe center pole this will be the only thing it needs .
been studie'ing other design's , exept kingpins . all have a flat butom !
there's wher im sure vince his gpu's have also a stepper design , becaus its the only way u can let the Ln² flow where its needed .

any other things u wanne know :p .

Ah wait, I think I finally managed to understand what you ment, it's not about the center pole, it's about the flat horizontal bottom of the tube itself you're talking about. It's no good, I agree, the approach MM uses to have more LN2 flow to the baseplate will certainly be better. Though, with your approach it's inpossible to get a center pole drilled out off a massive piece of copper, that's why I think a combo of both our drawings might be the best solution, though it's really hard for me to draw that in 3 dimensions :-p

Quote:

Originally Posted by geoffrey (Post 144244)

After multiple hours of benching you'll end up with a tube filled with Cola :D

For real now, what do you do with the Cola you just poored inside your tube? Does it float (it's iced after few seconds right?) on the LN2 surface untill you pick it out?

Ah wait, I think I finally managed to understand what you ment, it's not about the center pole, it's about the flat horizontal bottom of the tube itself you're talking about. It's no good, I agree, the approach MM uses to have more LN2 flow to the baseplate will certainly be better. Though, with your approach it's inpossible to get a center pole drilled out off a massive piece of copper, that's why I think a combo of both our drawings might be the best solution, though it's really hard for me to draw that in 3 dimensions :-p

cola = full of gas and stuff , this will flowed around the ln² and wil give it a boiling boost !

And besides the Cola trick, are there more things that we should have in mind when benching ln²?