just borrowed a laser thermometer and checked the temperature of my Block Erupter.
I put a tiny heatsink with some Arctic MX-4 thermal paste onto the ASIC and measured the following at room temp of about 26°C
- tiny heatsink top 51°C
- tiny heatsink bottom/side of ASIC 61°C
- side of pcb/heatspreader next to ASIC 60°C
- heatspreader below ASIC 38°C
- D-Link hub 35°C
- RPi CPU (also with a tiny heatsink on top) 37°C
HW error around 1%
I'm getting a USB fan tomorrow and will measure again
I've also measured mine with a 'laser' (IR) 'thermometer' (pyrometer), but stopped when the top of the ASIC and the heatspreader on the bottom both exceeded 70 deg C (ambient ~26 deg C too and without any fans). The operating temp limit of some of the components on the PCB is just 85 deg C, getting too close for my comfort. On that topic, does anyone know if AM have released technical information about their chips with electrical data and some recommended temp limits? I fear that this information may be out there but in Chinese.
It is actually pretty hard to use a IR thermometer at close range because the laser beam (which has nothing to do with the temperature measurement and is intended as a crude indication of the measurement position) and the location that the thermopile sensor 'sees' and takes its measurement on are several centimeters different at close range (<6", usually), because they are designed to operate usually at distances of 6" or more and measure relatively large surface areas. For example my 12:1 optics model reads an average from a 1.5" diameter spot size at 12" distance. Most have fixed optics too so the best you can do is guesstimate what component is being measured at any time and if you end up measuring a site 50% over the top of the chip and 50% over the PCB, the sensor is giving an average reading which is going to mislead you and the actual component temps could be much higher than the averaged value shown on the display.
A while ago when I got my erupters I had planned to borrow one of my work's colour FLIR thermal imaging cameras but I have found that one has been taken offsite for unknown period of time and other fails cal and shuts itself down.
