Hi All,
I am replacing my router with a quieter 52mm spindle, so I designed my own mounts that are pretty cool. I am yet to try out the dust shoe.
52mm Spindle Mount (www.thingiverse.com/thing:4189934)
Also this Ramps case which is a little better than some when it comes to cooling, wiring and maintenance.
Ramps Case (www.thingiverse.com/thing:4170854)
You don’t have any problems using the individual connectors of the CAT5 cables? I’ve always been of the mind that you’d have to use the full pair for each connector to the stepper. I’d be concerned that you’re pushing quite a bit of current through those tiny wires. CAT5 specifies 24AWG, Ryan specifies 22AWG. Based on the Wikipedia charts, that’s almost twice the mΩ/m, and half the ampacity. Although, at the levels used by the stock components, you’d still be within the limits… I think.
My stepper drivers don’t pull that much current, even with the pots turned up to 80%. I have a 3A supply and it is not anywhere near it’s limit. You can put a tonne of current through a copper cable if it is pure because it has no resistance. Before I started the build, I tried a little experiment and shorted out a bench top power supply with one wire from the Cat 5 cable. It wasn’t a problem up to a stupid number of amps. So I was satisfied!
There are several things to think about when choosing a wire size. In general, thicker wire has less resistance. Longer wire has more resistance. So a short thin wire could have the same resistance as a long thick wire.
If your wire has more resistance, it will have two effects: 1) the wire will get warmer, and 2) there is a lower voltage at the other end of the wire (in this case at the stepper).
If the resistance gets too high, the wire can get so hot that it melts the plastic around it. But even before you are melting wires, the voltage at the stepper will be less, so you start losing torque at higher speeds.
In a 3D printer, the wires are generally pretty short. But my Z axis stepper wire is 2.5m for example and the second X and Y stepper cables are around 2m. That said, I haven’t tried CAT5 cable and I haven’t tried to measure the voltage drop.
Also, btw, RJ45 connectors are supposedly not built to withstand a lot of movement of the connector. So you’ll want to secure the cable such that it doesn’t move near the connector all the time.
Lastly, there is solid core and stranded CAT5 cable. You absolutely need the stranded version. Solid core will break quickly apparently. Also, even for stranded wire, the bending radius should not be very small.
I’m not trying to shut down the idea of using CAT5 for stepper+endstop, heck I’d love it if it’s a reasonable choice. It just makes me nervous. What I’d love is a more definitive answer from someone with the bona-fides to back it up.
Nothing personal, Peter, I just don’t recall you claiming experience as an electrical engineer or electronics specialist. And most people I’ve read about using twisted pair double-up the connectors (the others don’t mention it).
No worries, I’m not a specialist at all. But I did consider using CAT5 cable for a similar project, and researched it on the internet. I also tried some wire calculators etc. My post summarized what I learned from that.
Also, btw, RJ45 connectors are supposedly not built to withstand a lot of movement of the connector. So you’ll want to secure the cable such that it doesn’t move near the connector all the time.
To minimize stress at the RJ45 connector and make it more secure you can use a Neutrik / etherCON connector, looks a lot like a XLR plug (used for microphones and powered speakers).
The only problem on our application is that they ar bulkie !
Mine have both been using cat5/6 since I built them. No issues with the connectors. I am doubling the wires though, each pair is a single stepper color.
I’d just try it and feel the temps of the cable during prolonged use. As long as you are present, the worst that happens is the wire acts as a fuse.
TE Connectivity (http://te/com) bought us lunch the other day and they have some nice industrial RJ45 connectors available. I’d go with T12 personally, which can come in 22 AWG.
I don’t want to start hysteria. But you can have enough voltage drop over a cable to effectively reduce the voltage, which can lead to reduced torque, or reduced torque at higher speeds. That can be caused by wires that are too small.
I’m interested in this since I had a pretty awful time with interference between my stepper and my thermistor on my 3d printer. I replaced half the printer before I realized what was going on. The twisted pairs in cat5 should prevent this nicely… after reading @jeffeb3 post about voltage drop I found this… think it might be OK ? https://www.calculator.net/voltage-drop-calculator.html
The wires definitely don’t get hot, but I’d be interested to know the current you are all running. During my testing the whole MPCNC was pulling a maximum of around 1.5amps at 12V. So each stepper maybe 300mA peak. For me I’m just not worried about burning a wire with 300mA. Now I don’t know the resistance of the wire - I don’t have a meter sensitive enough, but I reason that it’s pretty good if it can carry a 10 nanosecond pulse down 100m! I think from memory my stepper drivers output 12V? So let’s say the resistance Of my 5m loop is 1 Ohm (which is fair) -I’d drop 0.3V from 12V to 11.7 - it’s hardly going to be a problem. Remember that the stepper drivers themselves will have an output resistance -possibly more significant.
Also, I’m not convinced the stepper motors pull more current when they are driven faster. They are driven faster simply with quicker alternations of the steps - they are not supplied more voltage for faster travel.
I’ve never thought to put my thermal camera on the wires to see if they’re heating up.
Btw, I really like the fact that the RAMPS case has plenty of space! I hate things that are designed so small that it’s difficult to put together and maintain afterwards (unless there is a very good reason to make something as small as possible). I might give your design a try in the near future.
I’ll just point out that ganging the wires in a twisted pair will defeat the interference rejecting property of the twist. The two wires in the pair must have opposing polarity to benefit.
Sine the stepper drivers are constant current devices, the motors should see no difference at low speeds. Depending on the length (and current) it would have the same effect as lowering the supply voltage, so lower top speeds and lost torque at high speeds. You could calculate the effect by multiplying the resistance by the current. I don’t know if it is significant but if you upped the supply to 24V I’m pretty sure you would more than compensate.
There a possibility of the wires getting hot but I really doubt that would be an issue unless they were in some thermal insulation that allowed the temperature to run away.
I manually twisted my existing wires and it fixed my issue after reading it had resolved it for others.
Honestly it’s not the 24GA wire that’s the weak point in using data cables, it’s the jacks which are typically only rated for 500mA unless you buy industrial spec ones, and the jacks are typically designed for punching down solid cables into them, and the motor leads are stranded. Too many use case exceptions for me to be comfortable with it.
**A lesson in noise **(To whoever wants to 
):
Listen guys, the twists are for cancelling noise yes. But we are not going to have a problem with noise. A stepper motor is a low resistance (almost a short circuit). Noise is a problem with a high impedance input (like an oscilloscope input (10MOhm) or an amplifier input. Or a sensitive receiver like a radio or the cable that carries the weak signal from your antenna. Our stepper signal by contrast is not weak at all. It’s driven by an amplifier (the stepper driver) to rail voltage (12V) along 5m of copper at low frequencies - not GHz or MHz. So yes there will be noise on the line, but the noise has no power and will never compete with the stepper driver. Never.
Ever notice that the noise on your Hifi is loud when you touch the central conductor of the RCA plug (high impedance input that picks up anything including low power noise on your body) but then it all but disappears when you plug in a source properly - because the input is now being driven and the noise has no power. In comms circuits where we have to detect signals that have fallen in power to below the noise power then it’s a problem.
Think about this: do you ever have a problem with all the noise on the electrical grid causing your toaster to burn your bread? (don’t say yes). Noise doesn’t have the power to burn bread or drive a stepper. But you will notice it being a problem with your weak television signal when there is a heavy mist in the air and your TV has to pick out a signal that has travelled 60miles and landed on your antenna then run down 20 m of coax before hitting a boost. Now the noise power is something to worry about.
Now, do we have any high impedance inputs on the MPCNC? Yes we do. The step inputs on the stepper drivers themselves - they could be affected by noise, but they never will because they don’t use the cat 5 cables connecting them, they are directly connected to arduino but most importantly they are driven by the arduino, so again the noise cannot compete.
If your system isn’t behaving itself I would imagine it is because of missed steps caused by too much friction, poor power supply, too much mechanical load, or improper setting of your stepper driver pots.
This should probably have been a seperate post.
By the way, this MPCNC works a treat!
Cheers All
