[pHaestus]

Bill's questions are more or less the reason that I haven't actually DONE anything with my "new" test setup.

OK so I like CPU as heat source not because they are accurate but because they give me a chance to look at mounting/real world usage issues of blocks while I am testing. And also because the secondary heat sources and dissipation paths on the real PC system can do some interesting things when you get to either very high performance or very poor performance. But they are NOT good heat sources from a bench testing/modeling standpoint. OK that's fine still I state the system I use and I try to frame my results (when possible) vs. other people and I go on with life.

But my dP xmitter just won't seem to "go" like it should. When everyone is so particular then this shuts me down. I just don't feel like I can do the kind of testing I did on SocketA now. I'll be producing the numbers noone really wants to see :/

What we (as a community) need are some high quality test results on a quantified bench testing platform (for modelers and for pushing our understanding of the theories we blabber about on this forums) AND more good results of the type I produced with a real CPU to verify our bench testing is simulating real world performance.

The TTV would be a perfectly fine choice for heat source if one has the equation which relates the C/W for that source to the C/W obtained with different dies/heat sources. I don't think anyone whose reviews we read is making up numbers (me, Robotech, JoeC, Swiftech, who else?). They just require correlation with one another and there's not much common ground for us to do so when insulation of the die/die size/probe placement/W(in my case)/other factors aren't characterized.