the swirler-stein
- Thread starter Isayso
- Start date
hahnmeister
In Memoriam
I was just mounting a chipset cooler from an old PC on the shroud of the motor with thermal paste and a zip-tie. The hot part of the motor is the motor itself...not the gears and shaft that are on the side that attach to the acrylic. So you can still mount the bugger on acrylic, and just attach the cooling mechanism to the shroud of the motor itself.
Im got an idea though.... if you use a resistor/capacitor relay in conjunction with a small DC motor (a motor that is faster than say, 2rpm), you could run the motor on 12v or less (battery). If you size the cap right, you could make the motor rotate very fast in a timed (Im thinking 1.5 second bursts here on a 9+ rpm DC motor) second burst of electricity until the cap fills and the circuit is broken (then the resistor burns off the cap at a slow rate... Im thinking 20 seconds).
This way you could have a DC motor (more torque) at a higher speed, running on nothing but a couple D cells, that turns for 1.5 seconds resulting in a rotation of the drive disk of only 90 degrees at a time, then rests for 20, so it can easily stay cool.
The idea is to eliminate constant motion here in favor of timed bursts of movement. This would eliminate the 'constantly moving powerhead' in the tank as well as reduce heat. Any other ideas? I was thinking of trying to do caps + resistors on an AC control circuit as well for an 4rpm AC motor so it only runs 1/4 of the time..resulting in less heat as well.
Anyone else come up with something like this?
Im got an idea though.... if you use a resistor/capacitor relay in conjunction with a small DC motor (a motor that is faster than say, 2rpm), you could run the motor on 12v or less (battery). If you size the cap right, you could make the motor rotate very fast in a timed (Im thinking 1.5 second bursts here on a 9+ rpm DC motor) second burst of electricity until the cap fills and the circuit is broken (then the resistor burns off the cap at a slow rate... Im thinking 20 seconds).
This way you could have a DC motor (more torque) at a higher speed, running on nothing but a couple D cells, that turns for 1.5 seconds resulting in a rotation of the drive disk of only 90 degrees at a time, then rests for 20, so it can easily stay cool.
The idea is to eliminate constant motion here in favor of timed bursts of movement. This would eliminate the 'constantly moving powerhead' in the tank as well as reduce heat. Any other ideas? I was thinking of trying to do caps + resistors on an AC control circuit as well for an 4rpm AC motor so it only runs 1/4 of the time..resulting in less heat as well.
Anyone else come up with something like this?
Last edited:
miwoodar
Likey the bikey
Nice work. This looks like a great DIY. I think I may have to dive in and try it. I have a quick question regarding the heat issue. Has anyone tried plumbing in a line from an airpump? Am I overlooking something? Wouldn't the positive air pressure inside the box help to keep salt creep and corrosion out too?
hahnmeister
In Memoriam
Air has a very low heat capacity. Simply running cool air over the motor wont do much. Removing the shroud does allow the motor to convect, but it still remains very hot. Adding some sort of heat-dissipating material like aluminum or copper is the best way to distribute the heat where then passing air can carry it away...but the amount of heat is so great that the volume provided by just an air pump wont do... I would say that at a minimum we are talking a large heat sink, or a 40mm computer fan.
A water block with a water cooling loop would work, or some other heat exchanger. But this might be going overboard. FWIW, it doesnt seem like there is any reason why a clock/timing motor shouldnt just run hot. How I understand it, they are on 24/7 inside punch-card machines and time-clocks anyways in their stock configuration...they just run hot...so what. Its not like were running them underwater or something.
But the fact that DC timing motors run cool is the major reason I am considering one over an AC one. They provide more torque as well.
A water block with a water cooling loop would work, or some other heat exchanger. But this might be going overboard. FWIW, it doesnt seem like there is any reason why a clock/timing motor shouldnt just run hot. How I understand it, they are on 24/7 inside punch-card machines and time-clocks anyways in their stock configuration...they just run hot...so what. Its not like were running them underwater or something.
But the fact that DC timing motors run cool is the major reason I am considering one over an AC one. They provide more torque as well.
hahnmeister
In Memoriam
You still need to increase the cooling surface area if you are going to use air cooling though. Its alot like a computer cpu cooler, its too small of an area for just air to remove the heat... so you spread out the heat in aluminum or copper (or water or heatpipes or TEC or phase exchange) while expanding the cooling surface area with a heatsink.
miwoodar
Likey the bikey
Thanks Disco Stu. I think you're right about air and poor heat transfer. I'm going to start by mimicing Hmott's version (Hmott - the parts list and photos have been very helpful). The motor is already on the way. I'll keep my eye out for the heat sink though. I want this thing to be maintenance free and reliable because I travel a fair bit for work.
July has been the month of upgrades...first the skimmer, then the lighting, then the corals. The MJ mod connected to the swirler-stein takes the cake though.
July has been the month of upgrades...first the skimmer, then the lighting, then the corals. The MJ mod connected to the swirler-stein takes the cake though.
humboldt reefer
New member
Hmmm maybe I just got lucky But my motor barely gets warm.
hahnmeister
In Memoriam
some are hot, some are not.
The best HSF I can think of is a chipset cooler. There are even some that are large enough to not even need a fan...like the ones by www.zalmantech.com (the ZM-NBF47).
The best HSF I can think of is a chipset cooler. There are even some that are large enough to not even need a fan...like the ones by www.zalmantech.com (the ZM-NBF47).
miwoodar
Likey the bikey
I am wondering...I believe the strain on the motor may be causing part of the heat issue. I am wondering...wouldn't the amount of torque on the motor be dependant on the length of the arm (ie the distance from the rotational center)? A shorter arm would cause less torque on the motor and might allow it to run a little cooler, right? Has anyone with a hot motor noticed their build uses a longer arm than shown in some of the photos?
hahnmeister
In Memoriam
yeah... it was mentioned a few pages back by Paul (oceansmotions), that many timing motors just run hot... nothing to do with the load.
