Have you ever had a temperature sensor in your PC which was highly and immediately susceptible to what you did with the fans? This is most likely cause by a heat soaking area that is almost immediately clearer out by the fans.
Obviously if you are trying to get a handle on how hot things actually are this effect can be a nuisance. Obviously the results are impaired by bias somehwere.
This happened to me recently. A temperature sensor on the base of my radiator would drop rapidly when a bottom intake fan was spun up. As soon as it spun down the apparent radiator temp would rise back up again. Grrr. Pointless radiator sensor.
But... it got me thinking. A long time ago I wanted to make an airflow sensor for a PC case (electronics hobby). When I asked the 'grey beards' on the electronic forums one of the most commonly supported suggestions was using 2 thermocouples. One "hanging out the in the breeze" the other shrouded in something to keep the airflow off it. Like a bit of cardboard, lets not get technical.
The concept is simple. The one hanging out in free air is going to get cooled by the air flowing through the area. Where as the probe wrapped in a nice insulating bubble of cardboard won't. It's going to sit in it's own little heat soaked bubble of "average ambient" temp.
The difference between the two (mostly meaningless on their own) temperatures, is your "air cooling factor", or basically how well you airflow fans are working for that area.
If you have zero airflow the sensor should read the same, give or take a smidge. Turn all your case fans on full and the free air sensor will be cooled rapidly by the moving air, the cardboard wrapped one will cool much, much slower.
You can even put units on it. Granted it would seem odd, but it's "air cooling effect" in degrees C delta. Delta-T just.
My original idea was that this could be tidied up into a small package with an LED or RGB header so LEDs could be set to reflect this delta. RED if it's minimal, GREEN if it's mid ground and BLUE if it's high.
Obviously if you are trying to get a handle on how hot things actually are this effect can be a nuisance. Obviously the results are impaired by bias somehwere.
This happened to me recently. A temperature sensor on the base of my radiator would drop rapidly when a bottom intake fan was spun up. As soon as it spun down the apparent radiator temp would rise back up again. Grrr. Pointless radiator sensor.
But... it got me thinking. A long time ago I wanted to make an airflow sensor for a PC case (electronics hobby). When I asked the 'grey beards' on the electronic forums one of the most commonly supported suggestions was using 2 thermocouples. One "hanging out the in the breeze" the other shrouded in something to keep the airflow off it. Like a bit of cardboard, lets not get technical.
The concept is simple. The one hanging out in free air is going to get cooled by the air flowing through the area. Where as the probe wrapped in a nice insulating bubble of cardboard won't. It's going to sit in it's own little heat soaked bubble of "average ambient" temp.
The difference between the two (mostly meaningless on their own) temperatures, is your "air cooling factor", or basically how well you airflow fans are working for that area.
If you have zero airflow the sensor should read the same, give or take a smidge. Turn all your case fans on full and the free air sensor will be cooled rapidly by the moving air, the cardboard wrapped one will cool much, much slower.
You can even put units on it. Granted it would seem odd, but it's "air cooling effect" in degrees C delta. Delta-T just.
My original idea was that this could be tidied up into a small package with an LED or RGB header so LEDs could be set to reflect this delta. RED if it's minimal, GREEN if it's mid ground and BLUE if it's high.