Brief tour of my Portable Primo

Last night I was featured on the YouTube livestream of TOG the Dublin Hackerspace, and briefly showed my MPCNC and a couple of things I have made with it. There were some technical problems (my first time livestreaming with Google Meet) but I think the key info is there. Here is the link:
And in case this build is new to you, here is my growing set of build photos and videos:

You can learn more about TOG at

I will be having a Virtual Booth about my MPCNC and related things at this year’s Dublin Maker fair, June 19-20. More info:


Super awesome build! You have so much going on here. Really impressive stuff. I would love to know, are you sensing temp data from the motors themselves or the motor drivers on the board?

Thanks, Matthew! I am sensing temps from the stepper driver boards to see if my enclosure/fan were sufficient to keep them cool. Here you can see the DS18B20 digital temp sensors I glued to the heatsinks. (and you can see the endstops/Zprobe connector). The highest they have gone is about 35C.

1 Like

I told myself not to be that guy, but let’s be real. I am that guy. Your heat sinks, specifically, the outer fins of your heat sinks only got to 35C. Now, for the degree of accuracy you’re aiming for, that may be fine. (See what I did there?) But if you really want to know how hot the driver chips are getting, this ain’t it.

Again, this is not a bad way to see if things are getting hot, and certainly falls within the purview of the maker ethic/standard. And it will reliably tell you when the entire system is starting to heat up. But don’t expect to use those temps to try to ride the razor’s edge of thermal shutdown/failure.

This is not intended as a critique of your methods or what you’re doing, just pointing out what you’re actually measuring, so if anyone else wants to do the same thing, they too understand what they are actually measuring, and don’t mistake SWAG for precise measurement.

And yes, I am that guy… :shame:

edit: Forgot to mention that I absolutely love your setup. Both mechanically and in life. Love to quit my day job and do travelling maker events… If only I had the skills, talent, equipment, money, and disposition…

KV, you are absolutely right! I have been pondering this issue much because I am designing the next version with much bigger stepper controllers. I am wondering if I should put the temp sensors right where the chips are, to get more accurate temp measurement, or a bit further away because they would block airflow to the most crucial part of the heatsink if I put them on the chips. I don’t really plan on “riding the edge of thermal shutdown” anyway, but more sensors is always better in my book.

Hmm… I’d say there are two ways to go. First is to look at trying to put a probe on the underside of the driver board. Yes, it will be insulated from the chip by the board itself, but the chips are usually designed to transfer heat through the bottom anyway. The other thing to do would be to try to source some sort of wafer/membrane probe that you could mount in between the chip and the heat sinks (using a thermal adhesive of some sort, of course). This is assuming you are sticking with the stepper-stick style driver, and not moving to the big, monolithic drivers.

1 Like

Great idea! I opened a stepper driver box and there is plenty of room in there for a sensor. The chip is actually already on the underside, so the sensor will go on the top of the PCB right on the vias that lead to the chip.

Just don’t accidentally make them hotter by covering them up with a temperature probe. :slight_smile:.

Measuring the top is fine, as long as you think of it as 35 degrees Stevenheight, which is a slightly different scale than celsius of the chip. The drivers will shut down to protect themselves. You just want to look at see, “oops, we are at 50 degrees Stevenheight, that must be the problem, I will clean out the case”.


Have you considered switching to 2209 drivers as they have an internal thermostat and settings for current fallback if temps reach dangerous levels.

Yes, those would be a good plug-in replacement for the RAMPS board and DRV8825s I am using. But as I mentioned above, I will be moving to bigger and better electronics soon, and using TB6600 drivers. I already glued the DS18B20 digital temp sensors on them, as KV suggested.

If the goal was to simply see if your enclosure cooling is sufficient. Then wouldn’t a single probe in the airstream near your hot air exhaust be sufficient? Don’t get me wrong, what your doing would work, just seems like overkill for your stated goal. After all if your case temp never gets above 40C than you should have lots of headroom.

When I built it, I did not know what the various thermal limitations of my MPCNC would be, not having built my own CNC before. So I installed more sensors to educate me, e.g., how hard the different axes work when doing a job, how shall I set the currents, how much does wires blocking the heatsinks affect their efficiency (it is very crowded in the RAMPS enclosure I used), etc. If one stepper driver were to burn out, I would want to know if that was due to overworking it or poor cooling of it. I already have a warning beeper installed in the spindle cooler if it gets too hot; I plan to do that for the stepper drivers, too, and I can look at the display to see which one is cooking. The fan is blowing on the board and leaking out of the RAMPS enclosure in many places, so there’s no one “hot air exhaust”. All of these sensors are also useful in the educational context I will be using this portable CNC for, maker workshops.

1 Like

I don’t recall ever hearing of a stepper driver “burning out” because of over heating. Just so you know. They all have temp safties on them. So if they get too hot they simply shut down and ruin your project. At which point it is easy to see what driver over heated because it corresponds to the motor that is no longer moving. Once the driver cools down it will work again. Just wanted you to know in case you wanted to test/share that knowledge given you are using the machine for educational purposes.

That is good news, thanks, Atom.