Offizielle Stellungnahme Asetek zur LS

Das ganze bezieht sich auf Evos Thread (zu dem ich weiter unten noch was hinzufügen möchte) und auf eine Mail zum PR1:

-------------------------------------------------------------
Yes I read that review, and I'm not very impressed with it. There is a lot of errors and mistakes, and from my point of view it looks as he wants the LS to look bad with purpose. Just the fact that he didn't contact our support about the LCD/USB before posting the review shows this. Also it seems as he has maybe bad CPU contact.

Again this was his hardware. You don't actually know what your CPU/motherboard will show in a MACH II compared to the LS? Each CPU/motherboard has different read outs.

OK we have done some testing on an ABIT IC7 MAX3 with two different CPU's. One of them a Prescott 3,4G and a P4 EE 3,4G. Under full load (we have software that can actually load the CPU 100%. Prime is not 100%). We ran the P4 EE at 4,4G@1,725V. The temperature rose to +5'C on the CPU according to Motherboard monitor. As you know this CPU is FAR above 100W.

As you maybe also know the Prescott extracts 103W at default speed. We did the same test with this CPU with NO overclock. This showed +12'C. In other words with a much lower heat load the CPU had a higher temperature.

As you understand this is of course wrong. I don't think you can trust motherboard read outs at all. You wrote to me a Prescott would maybe force the VapoChill to its knees. The is not true. The system is not stressed at all. We brought the cooling unit to the lab, and it removed 240W easily as promised.

Because we optimised the system for 240W we have to charge more coolant in the cooling system, which makes a bit higher pressure, which again leads to a bit higher temperature. If we optimised for 200W like the nVENTIV units, we can remove 15Gr of R507 and gain a bit lower temperatures - we already tested that.

Try to read this thread: http://www.bleedinedge.com/forum/sh...age=1&pp=15 As you can see he gets -15'C CPU read out. Again a proof that internal CPU temperatures are different on each CPU / motherboard. Please also note that in Intel's white papers it is specified, that the CPU read out is calculated NOT measured. ASUS, ABIT etc, are doing it in different ways, and nobody really knows what is right and what is not.

Because of our cooperation with Intel we will soon get real CPU dummies, where we will be able to simulate an actual CPU under load and thereby have the chance to calibrate what is right and what is not.

When we test our systems, we have a thermal dummy, which is a copper part same size as a CPU with a variable heater inside. Inside we have a thermal probe that can read the temperature. These temperatures are what you can see on our web site. We can ONLY promise how the cooling system will perform. And in order to do that, we need to have a standardized setup. If we used a given motherboard / CPU for that, you could never trust it. However because of this you cannot compare your CPU temperature to the graph on our site. We promise our system can remove 240W, and it can.

Furthermore nVENTIVs evaporator temperatures are ridiculous. They cheat with purpose. If you want to, we can make a new software that is "calibrated for nVENTIV" standards, and you will see -70'C on the evaporator, but why?

Personally I think people are blinded a bit by nVENTIVs fake evaporator temperatures, and as I wrote before it is pure physics that an 11cm3 compressor will perform better than an 8,5cm3 compressor. However if you want to get close, get someone to charge your system with a less amount of coolant.

We have a few standard MACH II's here. We will charge one with R507 and compare it to the LS. Furthermore we will compare it a standard MACH II. However when we compare the two units in our test bench, the MACH II is way behind the LS. Apparently the difference is not so big when mounted in a real setup, and we will try to learn more about why that is.

The overall conclusion from here is:

Thank you for bringing these issues to our attention. We just wish that you buddy who made the review had the same professional approach. I guess being an nVENTIV reseller and a VapoChill reviewer at the same time isn't the best blend

However your mail certainly made us look into our products, and if they performed as promised. Our conclusion is - yes the units perform exactly like promised. We also learned that if we want to get even closer to a GT we can do that by reducing the charge of refrigerant, however we will then sacrifice on cooling performance and land on a maximum of 200W.

Furthermore we learned by sending an official request to Danfoss, that to use the compressors nVENTIV are using is suicidal behaviour, and this is not a route for asetek pursue unless Danfoss changes their mind. You saw the response from Danfoss yourself.

The remaining issues are now the strange behaviours you have noticed with the USB driver that disappear from Windows and the weird characters in the LCD display some times. The challenge is that we cannot re produce these issues internally, but we will continue to look into it. We have the best support in the industry and we intend to live up to that in this case as well. As you maybe know however software errors seen once in a while can take some time to solve, so please be patient.

---------------------------------------------------------------

Zusammenfassung:
1) Die LS Sie leistet was sie soll
2) wenn das Display nicht gleich geht den Stecker verkehrt an der Chillcontrol anstecken, dann gehts
3) es wird an einer neuen Software+Firmware gearbeitet damit der connect windowssoftware und chillcontrol unter bestimmten Umständen nicht mehr verlorengeht. die software ist wie immer gratis und kann jederzeit nach Verfügbarkeit eingespielt werden.

Und nun zu Evos Vapo, die wir gestern wieder zurückbekommen haben.

Das Display ging auf Anhieb (siehe PIC1)

Die Kühleistung IDLE ist völlig normal und wie von Asetek versprochen, auch die Anzeige und die Meßtemp passen auf unsere erwähnte Abweichung von 5-6 grad zusammen, das ist ok wenn man die Fühlerstandorte berücksichtigt. Es ist sogar der von uns gemessene Wert tiefer als die Asetek Zusage laut Chart. (sie PICS 2 bis 4)



Lasttest kommen wenn unser Heater endlich fertig ist.

Bowman ist gerade dabei die MAch GT und die Vapo LS zu testen.
Der ist garantiert neutral und weis auch was er da macht.
imo will er auch ein paar mods für die ls zeigen.

Hmm ..
Bin auch gespannt, was beim Bowman so rauskommt ...
Also wenn für die LS auch extrem krasse Mod´s verfügbar sind, dann wird sie auch wieder interessanter ...

MFG
Tscheckoff

Ja, aber 200W sind echt utopisch. In diesem Jahr zumindestens.
Die hätten ruhig mehr für die Temps machen können.
Zu laut finde ich persönlich die LS auch. Kann aber daran liegen, dass ich Parkett habe und das geräusch natürlich noch reflektiert wird, muss mal was drunterpacken...

Ach ja, die läuft bei mir nun seit einen Monat ohne aussetzer, ausser das die usb connection wirklich spinnt.

cyas

Naja, den Tejas hat Intel ja gestoppt Da hätte sich die LS gelohnt.
Und wenns wirklich auf die Pentium-M Architektur umsteigen,
dann werden wir vorerst keine 200W erreichen.

wieso kaufts dann Kompressorkühlungen die bis >160 Watt gehen wenns eh keiner braucht ?

Zitat von Joe_the_tulip

bin schon auf evo sein comment gspannt. nur irgendwie isses auch shit, wenn ich um die Kompressorkühlung bei mir ordentlich benutzen zu können gas ablassen/zuführen muss! aber so ne automatische regelung wär sehr fein!

Du kannst auch die CPU auf 240 W bringen


Würde mich aber interessieren was ein "Watt-Regler" dazu noch kosten würde, wahrscheinlich noch eine Lawine zu den üblichen Apothekenpreisen dazu

man braucht erstens das ventil was nicht gerad billig ist (find aber grad nix passendes ) und dazu einen behälter der das überflüssige gas speichert
wenn das dann noch automatisiert sein soll, kann man sicher mit 3000€ rechnen