I don’t know mach 3. I think we’ve gone through this on the forums before. Some are 0-12V, some are 0-5V and some are 0-10V.
Does it have a potentiometer to control it now?
I guess I should clarify, you probably won’t be able to do speed control, but you’ll be able to do power control. So you’ll be able to set it to 40%, not 5000rpms.
Yes there is a pot. I’ve been reading through the forums trying to find a completed project with control over the spindle speed. I was reading one thread where the guy had used a PWM controller with an IR LED and was able to control the speed. I don’t think it was controlled via Marlin, but it was 75% if not more complete. Of course, I can’t find the thread now. I’ll post it when I find it again.
Ryan has a PID project he was working on. The title was something like “software needed for a hardware fix”. He was eventually outoutting a signal to a triac for an AC universal motor (dw660). It’s similar, but not the same as your pwm input.
There are some off the shelf pid stuff, buts it’s pricey. IIRC, it was called super PID. Ryan’s solution is pretty elegant, and outputs a pwm, reads speeds fron Marlin, and measures the spindle speed with an ir led.
Just controlling the power would be a good step though.
Today I found an old enclosure at work that was going to the trash bin. I instantly thought it would be perfect for the RAMBO board and LCD. I CNC’d the opening for the LCD and the control knob.
Learning as I go. The big CNC I have access to at work has a vice, which when stock material is placed into, automatically squares you material to the X and Y axis. Obviously, when you randomly screw a piece of stock onto the MPCNC, it is not going to magically be squared up 
I didn’t think that far ahead, lesson learned.
I also had my spindle mount rubbing against a bolt on my Z axis preventing it from lowering consistently, which took 30 or so minutes to really figure out.
Even with the learning curves and random curve balls, I think the LCD opening came out kinda ok. I didn’t account for the round endmill not being able to cut a square corner so a recess for a hidden part of the LCD is preventing it from protruding evenly. Either way, I’m glad the board has a safer home instead of lying on the ground.
[attachment file=75541]
1 Like
Well I found the PID thread:
I’m thinking about purchasing the kit from Ryan. I’ve got about 10 Arduino Unos lying around that could fit the bill, but I do not have a Dewalt 660. Does anyone know what might need to change for my Chinese spindle?
The PID setup for the DW660 has:
Marlin -> arduino Nano -> Triac board
And there is an IR LED that also goes into the nano.
What you want is:
Marlin -> arduino (uno or nano) -> “Mach3” input.
The arduino and the triac board are both 0-5V PWM. From what I can tell, the “Mach3” input is 0-10V (analog, but PWM works, I think). So for starters, you’ll need something to level shift the 5V to 10V. Either a transistor, or an off the shelf level converter, or something like a ULN2803. That will let the arduino control the spindle to it’s full speed (I think, I’ve never seen one 
).
The other half of this is getting the IR sensor (LED and transistor, technically) mounted on a place where it’s relatively dark, but it can see something that’s spinning on the spindle, and has been painted half white, half black.
The other other half of this is that the PID isn’t 100%. There are still some gremlins. Eliminating the triac from your setup might actually be a big help. It might not be so noisy. Just guessing here though.
The triac in Ryan’s kit won’t be useful for you, but maybe you’ll want it in the future, and the rest of the stuff should be apropos.
A brush less spindle with hall effect sensors does not need this. They cost more but it is built in.
I’m back! Haven’t posted in a week or so.
Today I started to cut out some acrylic and noticed a lot of jerky movements. Here is a video:
- Passes were taking .5mm off at at time.
- Feed rate was 575 mm/min
[attachment file=“acrylic settings.png”]
I don’t think I was pushing any limits of the machine. I read a few people having issues with their baud rates to their marlin boards, but I don’t think that is what is going on here. I thought the RAMBo board had 1/32 micro stepping, but was reading the product page and it looks like its actually only 1/16 
I should have gotten the Archim board.
Can anyone advise?
The 1/16 vs 1/32 doesn’t really matter. Don’t worry about that.
Are the things that connect the motor to the belt tight? I can’t remember what they’re called. Not pulleys, not couplers… Brain fart.
Maybe post a picture of you zip ties? Can you cut a square and see if it’s only in one direction? Do you have an LCD? Can you “print from sd” just to eliminate any computer/usb/whatever issues?
They are pulleys! I was thinking that was the idler. If it’s slipping on the flat part of the motor, but not completely slipping, that could explain it.
The pulleys are secure, and the set screw is aligned perfectly with the flat portion of the shaft. I don’t think it’s slipping there.
Yes, it is only in one direction, I meant to include this piece of info in my initial post, only the X axis, that’s why I don’t think it’s a baud rate problem.
I’ve attached some photos of my zip ties. Also a picture of how badly the cuts are coming out.
[attachment file=76842]
Acrylic is hard to cut, it’s “grabby”. Try trichodial full depth of cuts and see if that makes it better. Also, are you still using that straight flute bit you posted a picture of earlier? If so, that could also be the issue.
I’ll try to make fusion 360 do trichodial cuts, I’m not really clearing pockets, just tracing. Although I’d think it would do it in both directions, x and y, if it were a problem related to the type of cutting (conventional), still worth trying though
I ordered a few endmills from the shop. In particular I’m using a 2 flute 1/8” carbide.
I think fusion calls it adaptive.
I’ve been spending a lot of time trying to fine tune my MPCNC. I’ve had issues with electronics, hardware, endmills, materials, clamping systems, my table, pretty much everything. The forums are amazing! If it weren’t for that, I’d have spent a lot more time getting it all dialed in.
I’ve been working on a spindle speed controller, and have had a weird issue where my RAMBo board will output the correct PWM signal, then I power on my spindle, and there is no telling what signal I’d get from the RAMBo board. I hooked up a scope to see what the Rambo board was outputting when the spindle was on and when it was off.
Spindle off with a good PWM signal coming out:
[attachment file=77980]
Spindle on, clearly a lot of noise on the PWM line:
[attachment file=77981]
So my spindle is causing an insane amount of noise on the RAMBo board. The RAMBo board and the spindle are not connected in any way except through the 110VAC wall outlet.
Now with the inconsistent cutting motion of the axes. I haven’t tested this yet, but my guess is the interference from the spindle is causing the stuttering of my axes. Tonight I will do a run with the spindle turned off and see if my axes stutter across the work space.
I’ve found this other thread that has some similar issues.
At this point, personally, I would not recommend the secondary spindle option until I find what is causing the issue and a solution.
Are you sure that noise is real? Maybe its noise getting in on the scope probe and its ground connection. Motors are quite noisy and any long cables can pick up the noise.
It is not uncommon depending on where you connect the ground lead of the scope probe to see all kinds of rubbish on the scope that isnt actually on the signal you are probing.
I would be really surprised for the pwm output to have that kind of noise signal on it. Do you have a pic of how you connected the scope to the board?
Which output are you using in the Rambo? If it’s one of the 12V ones, maybe it needs a pull up resistor.
What spindle? The import ones are notorious for this issue.
I’m never 100%, with this particular situation I’m about 90% sure this noise is real. I’d love to hear how I can ensure this noise is a real thing.
I do not have pictures right now of how the scope is connected, but when I get home this evening, I’d be glad to take some pictures and post them.
A little more detail, I have an Arduino Uno that I am using to detect this same PWM signal (its been disconnected for the test I ran above). The Uno board does begin to act oddly when the spindle is turned on. It is connected in the same manner that my probe is, so the connection method could be the issue.
I’m using the pins from the pins_RAMBO.h file in the marlin firmware.
//
// M3/M4/M5 - Spindle/Laser Control
//
#define SPINDLE_LASER_PWM_PIN 45
#define SPINDLE_LASER_ENABLE_PIN 31
#define SPINDLE_DIR_PIN 32
I using the import spindle that is recommended in the “Specifications” section of the MPCNC build, under “Tool”. Do you know of a solution to reduce or eliminate this noise?