I was poking around the firmware on Ryan’s github page and noticed something peculiar. I’m looking at Configuration.h in the MPCNC_Rambo_T8_16T_LCD branch.
Has this recently changed? If we’re using 200 steps/rev steppers, 16 microsteps, 2mm pitch belts, and 16 teeth pulleys shouldn’t that work out to 100 steps per mm for X, Y? And shouldn’t Z be 400? Or am I missing something?
Also, the drivers are set to 16 microsteps correct?
I would except I’m not actually using a rambo right now. I’m setting up linuxcnc for lowrirder plasma. I’ve been poking around your configurations to get some initial settings for the steppers. (and converting them to barbari… err um imperial units.)
Looking at your new testing config I see you really dialed back velocity and acceleration. I’m afraid that’ll be too slow for my application. I’ve upped the juice to my motors with a new power supply and drivers so hopefully they will handle quicker speeds smoothly as well *** Tim The Toolman Grunt ***
When I get back to working with my mpcnc router I will definitely be testing out your new config.
I do not think this has ever really been tested much so if you run into issue start here.
I did a lot of messing with this I suggest leaving it and only upping it if needed, it is just on the cusp of noticeable. Have a look at my latest YT vid. And if you use ARCS (you should)…it is a must to use this one. The stock accels are kinda ridiculous if you ask me even for a light printer (3000mm/s/s get you to 40mm/s in .2mm)
I have set them to what I tested will not soften PLA after extended use. (on this new firmware…70-75%)
Are you referring to Marlin? Linuxcnc should have it covered.
I will need to up it though. Looks like your new max velocity 118 ipm. I’ll need to exceed that to hit the book value speeds on some of the thinner materials.
Good point. What does 70-75% work out to?
I’m currently setup at 24v/1.5a for my nema 17 84ozin steppers.
You should change the pulleys to 20 teeth or larger then as well. This is built for low speed power, if you need high speed it is best to keep the steppers in there powerful RPM range. Just because they can turn faster does not mean they should.
At 24V I am not sure the setting is the same never tried to look into it…But I have them now so maybe I should just check myself…2 minute check (lazy). Just power it up and let it run at a suitable ambient temp for at least 30-45 minutes to verify they do not melt your mounts.
The thought was to reduce the speed to help people on the learning curve. If you want to increase them, no one will stop you. Rather than leave it on the advanced settings by default and let people do crazy stuff in cam.
Be careful with “book” values. I don’t think in inches or minutes, so 50mm/s seems awfully fast for milling. I do not think you’ll be happy at those speeds, but feel free to prove me wrong. There are so many ways to set up those calculators and they are designed on huge, big money machines.
Will do. But just for reference can you tell me what you’ve set the drivers to in amps? I tried looking through the code and couldn’t find it. I feel like github used to have better code searching capabilities…
But we’re not milling, we’re melting steel!! As I understand it the plasma machine manufacturers dial in they’re optimal cut speeds quite well. Most if not all comments I’ve read say follow the mfr speed charts if they’re available. But all that said I imagine most of my cutting will be around the 50 to 100 ipm mark.
Oh, I forgot you were plasma cutting. Maybe that will be light enough to reach those speeds. Sounds like fun.
The 24V shouldn’t change the current limits. The current is the real value. the voltage is just overhead and as long as you don’t need more, it will be the same, so I suspect the same values will be useful for 24V
Yup, each little bit higher the steppers get a lot more heat. Tested with a flir, but I also have to worry about people in Arizona in the summer, in a garage next to a window. so there is room for more. The board is fine to way higher if mounting it vertical with room to breath.
Sweet, thanks I was not sure if I was over simplifying anything in my calc, but yup no voltage in there at all. Speaking of I should make a jscalc page for them.
In summary, more voltage = faster, more torque…and more vibration/noise, and heat.
Have you put a IR thermometer on the motors to find the temp limits? (EDIT strike that question. flir = thermal imaging. got it) Have an idea of a specific temperature you’d recommend to stay under to prevent deforming the mounts?
The problem with that analysis is, once the set current is reached, the power is turned off to the motor for the rest of the duty cycle.
If the motor is stationary, and it needs 10% duty cycle at 12V to reach the target current, it will need 5% duty cycle at 24V. The extra voltage doesn’t give you any more torque. When you go faster, the resistance is higher, and the duty cycle needed to reach the current set point increases.
At some point, it will take 100% duty cycle at 12V to reach the current and only 50% at 24V. Below that speed, the two power supplies act exactly the same. Same current, same heat, same torque.
Above that point, the 12V will not be able to provide enough current to reach the set point, and the 24V will have more torque. I am sure the 24V will have more temperature than the 12V after that, but I am not sure if it will be hotter than the 12V at lower speeds. I assume the temperature is proportional to the current, which is lower past that speed.
FWIW, that speed is pretty high, 60mm/s or so. It will also be lower for serial wiring, but still above the typical milling speeds. You are right to want a 24V supply in your speed daemon, but you won’t get higher torque until you are going very fast.
Something I found out, and should have thought of it myself. If you’re using a flir or other ir camera, or any ir thermometer, paint a flat black spot on the object you’re measuring. Metal reflects heat, so your measurements can be off without it. I could see a reflection of my hot end on my heated aluminum build plate on my printer.
It is crazy to see a heat reflection! I tried a spots of masking tape and it seems to work on things that do not get too hot. There are settings to check reflective surfaces but it even warns it could be off.
Getting a thermal camera really ruins movies for you.
FBI Guy: “The drone sees the bad guys in the building with it’s thermal camera.”
You can’t even see through a tent wall, let alone an insulated building wall. Hell, regular glass blocks thermal cameras, unless someone is touching it.
HAHAAA but it can show you footprints on a cold floor.
It was cool to even be able to see which wires heated up a bit. Not something I think much about but seeing it reminds you they really are tiny resistors. Seeing how fast the drivers cool down when powered off is crazy. It would be awesome to be able to see the airflow somehow, I understand how impossible that is but it would be awesome.
I saw a video somewhere showing someone running some powertools through extension chords. The thermal camera showed a tangled bundle of chords getting quite hot but when the extension chords neatly strung out they didn’t heat up much at all. That’s why they recommend if you’re using one of those retractable extension chord reels to always pull all the chord out of the spool… i still never do
