Nebby
Headphoneus Supremus
- Joined
- Aug 31, 2004
- Posts
- 2,201
- Likes
- 19
Quote:
I'm not too worried about the absolute temperature value as I am about the relative temperature values. Some transistor devices change characteristics depending on the temperature they are operating at so if you have mismatched temperatures it can cause issues. I can't say whether or not it's a factor with the beta22, but I think it's at least worth keeping in mind.
For your simulation did you factor in the thermal resistances of each thermal junction? Did you also factor in the resistance of your electrical isolator between the mosfet and the aluminum angle?
I just ran an example of the B22 waterblock system. Two B22 boards share one waterblock in my design. Each MOSFET is rated for 175deg F according to the data sheet. They are attached to a 3/16" angle beam attached to a 3/16" plate.
According to amb's site, on a typical setup the outboard amplifier MOSFETS put out 8V*0.16A and the inboard cascode MOSFETS put out 22V*0.16A (correct me if I'm wrong, Ti). This puts 4 whole B22 boards at 38.4W. That sounds about right given its 100W power supply. For simulation purposes, let's double that for a factor of safety. That's 2.56W for each outboard MOSFET and 7.04W for each inboard MOSFET.
At 100% load my 125W CPU cooled by a Corsair H50 (single radiator cooler) shows 133deg F at an ambient temperature of 75deg F. Let's assume somehow we'll need to also remove 125W from the entire system and that the waterblocks are at this same temperature. When a steady-state simulation is run, the maximum indicated MOSFET temperature is 142deg F, safely below the manufacturer posed limit.
Take note that this temperature value is for a hypothetical extreme example. In actual use, much less power will need to be dissipated and the initial waterblock temperature will be much lower. The MOSFET temperatures will probably hover around 120deg F at most.
Logically speaking, if those little passive bolt-on heatsinks can adequately cool the MOSFETs, an active water loop should do the job much better.
I'm not too worried about the absolute temperature value as I am about the relative temperature values. Some transistor devices change characteristics depending on the temperature they are operating at so if you have mismatched temperatures it can cause issues. I can't say whether or not it's a factor with the beta22, but I think it's at least worth keeping in mind.
For your simulation did you factor in the thermal resistances of each thermal junction? Did you also factor in the resistance of your electrical isolator between the mosfet and the aluminum angle?