Draytology's Experiment: Heatsink vs No Heatsink!

Hello.
I recently bought 10 100k thermistors. It wasn't until I received them that I realized they were only rated up to 125 C. But, although I really can't use them for my heated bed or hot end, I still managed to find something useful to use them for. I have heard many different opinions on whether or not a heat sink should be used. Of course we all agree a fan is without question beneficial. So I figured I would put together a little scientific experiment for all of our beneift.
I wired up one of the thermistors to T1 and since I am not currently using a heated bed I can use the bed temp readings in Repetier to monitor the temp of my thermistor. I removed a driver and carefully placed the thermistor right under the little PCB, underneath the driver. After letting it settle into the temperature of the driver with no power to it I got my notepad out and got started. FYI before I began I checked the accuracy of the thermistor just in case I got the settings wrong in the firmware, and after comparing it to my hot end temp and a control (me) I verified it to be accurate.

Here is the set up that will be tested.

Arduino Mega/Ramps 1.4
Drv8825 @ 1/16 set to 1.0v
PSU @ 12v
X axis > Nema 23(Rated 2.5a per/phase)

I took readings every 10 minutes for 30 minutes total with a complete cool down period in between each configuration.

*All Temperatures are in Celsius*

No Heatsink (No cooling Fan):

Temp @ 0min - 26
Temp @ 10min - 86
Temp @ 20min - 90
Temp @ 30min - 91.2

Heatsink (No cooling Fan):

Temp @ 0min - 26.9
Temp @ 10min - 85
Temp @ 20min - 91.4
Temp @ 30min - 91.8

No Heatsink (Cooling Fan ON):

Temp @ 0min - 26.3
Temp @ 10min - 58.9
Temp @ 20min - 61.7
Temp @ 30min - 62.5

Heatsink (Cooling Fan ON):

Temp @ 0min - 26.4
Temp @ 10min - 53
Temp @ 20min - 54
Temp @ 30min - 53.5

I plan on expanding on this experiment over the next few days. So, if anyone has any suggestions, questions, or requests please feel free to let me know.

Edited 4 time(s). Last edit at 02/13/2017 07:40PM by Draytology.

After finishing the experiment I paused the print and put the heatsink back on the driver. I was curious to see if the temp would actually fall back down to the range I recorded with the heatsink on. So I gave it another 10 minutes and recorded a temperature of 55.9 down from 62.5.

It's nice to see someone taking measurements!

May I suggest testing with different fan arrangements, in particular a fan above the drivers blowing straight down on them, vs. a fan blowing across the drivers, vs. a fan blowing along the drivers so that some of the air can flow underneath them.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

You could glue a heatsink with CA glue instead of using the thermal pad it comes with.
I made some experiments with raised drivers and added a heatsink underneath, but it wasn't worth it for the steppers I used to use.
Maybe for the z-axis of the D-Bot I'll be using a single NEMA23, then it is getting important again.

That is a very interesting idea O lampe! I have actually been thinking about putting together some kind of set up that would allow better cooling of the underside of the driver so I might try a variation of your set up. I soldered two drivers together a few weeks ago for fun, so I can definitely appreciate the precision soldering I see in the pic.

Right now I have the thermistor positioned right underneath the pad on the underside of the driver. Luckily it fit perfectly right in between the driver header and the male jumper plugs. Before I start testing configurations like yours I will have to figure out thermistor placement. That might be tricky. I am also trying to avoid direct fanning of the thermistor because I feel like that will skew the results.

Dc 42 I think I am going to try that thanks for the suggestion. For the above tests I kept the fan arrangment like I always have it which is air blowing on the Ramps from the side(the long side where the endstops are) but with the board positioned at close to a 45 degree angle. I set it like this just to increase the turbulence of the air. So, I am curious to see how this turns out. Also I think it is worth mentioning that I use two heatsinks, one stacked on the other.

O lampe that is an awesome idea moving the stepper plugs to the top which allows direct airflow on the heatsink, it just hit me why you had it on the top.

One question, I am pretty new to 3d printing and have 0 experience with deltas, but you mentioned using a Nema 23 on the Z axis. I was under the impression that delta's use three identical motors and set ups for X,Y, and Z?

Also, I am using a Nema 23 right now on one of my axis. I kept the Drv8825 as my driver. I have it set to 1.0v which seems to be adequate as long as I keep it cooled. What driver set up are you planning on using for your Nema 23? I was looking into a TB6600 but I don't know if that is even needed considering the DRV8825 seems to be handling it just fine. I plan on increasing my voltage soon so I will have to see how everything holds up at 24v.

Edited 1 time(s). Last edit at 02/14/2017 12:33PM by Draytology.

The NEMA23 would move a CoreXY Z-axis. I guess, I'll drive it with a RAPS128.

The reason to move the stepper connector to the top of the driver was also to eliminate another source of failure: Why would you feed the stepper through the Ramps? It's a second set of connectors and PCB traces, that aren't needed.

That's a good idea. I am actually having an issue with sudden voltage drops and interference noises. It doesn't seem to happen when I have my fan on but I noticed it did happen once when the temp was not very high at all so I'm not sure if it's actually a temp issue. It is very similar to the problem I was having before I upgraded my wire gauge. So I think your idea might help me. It could be the small PCB traces on the Ramps board meant for a Nema 17 are having trouble with my Nema 23. I think I will give it a shot, thanks for sharing I probably would have never thought of that otherwise.

The traces on Ramps are strong enough for NEMA 23@2.5A, but the connectors are often cheap and unreliable. ( also between Arduino and Ramps )
If you are sure about your wiring, you can solder the stepper wires directly to the driver.

So without active cooling of the board (which neither of my kits came with), heatsinks do essentially nothing. A fan, even without a heatsink, if far, far, far more important. Very interesting. I guess I should add a fan!

I didn't see that you mentioned what fan, how far away, etc. Any idea of the airflow you are getting?