Using CAT6 cables for 3D printer motor / sensors / fans

Question

I'm considering using CAT6 cables to connect my printer's extruder assembly to the control board. They seem like an elegant solution, but I've read conflicting opinions online on whether or not this would be feasible.

I would like to know if CAT6 cables can handle the required current, whether I should be worried about electromagnetic interference or other problems, and how I should pair up the wires. Cable length would be 30cm max.

Here are the relevant parts:

1. E3D heater cartridge (2 wires)
2. E3D thermistor cartridge (2 wires)
3. 30mm hotend fan (2 wires)
4. Z-axis auto-leveling probe (3 wires)
5. NEMA 17 extruder motor (4 wires)
6. 50mm part cooling fan (2 wires)

[cable A] I imagine I would use one CAT6 cable for parts 1-4, which form a logical unit (and in the future I might combine them into a removable module). I've been given to understand that power for the fan can be spliced from the z-probe or heater cartridge, so 8 wires should be enough.

[cable B] I would use a second CAT6 cable for parts 5 and 6. There will be two spare wires, so I could potentially double the bandwidth for the motor.

Answer 1

The ampacity question is not completely answerable because CAT6 does not specify wire gauge, so the current limit will depend on the specific gauge you get. CAT6 can be anywhere from 22 AWG to 24 AWG, and depending on who you ask this can be good for as much as 7A or as little as 0.5A. Given that you will have a bunch of wires in a bundle, this may cause them to heat up more than if they were in free air. For the steppers (1-2A) a single wire should suffice, but for the heater (around 3-4A) you might want to double up.

EMI will likely not cause any problems regardless of how you wire things up. CAT6 cables have the wires twisted in pairs of 2. Some people recommend to take advantage of these pairs: the +12V and GND of the heater should use a pair, each of the two coils of the steppers should have their own separate pairs. The reasoning behind this is that with equal current flowing in opposite directions in each wire of the pair, the generated electromagnetic fields will cancel out.

Twisted pairs are usually used when dealing with multiple pairs of wires that are carrying high frequency signals that might affect each other. The main concern for crosstalk in this application is if the stepper motor might cause the endstop to be erroneously triggered, but this is only a concern during homing when the feedrate (and thus frequency of the signal) is low anyways.

Answer 2

I did this for some of the wiring on my printer, and it's working fine so far. The two cautions are:

• At @tom pointed out, the heater is the high current item, so be careful of the wire gauge, and avoid running the wire where air can't circulate well to cool it. Wire ratings differ greatly depending on whether they're in a bundle (poor air circulation) or free.

• For the most part I agree that EMI shouldn't be a problem -- but if you switch to thermocouples it might become a problem -- they're much more sensitive, and this might have been part of the problem I described at How to get consistent and accurate readings from thermocouples? (though alternate wiring didn't completely solve the problem in that case).

Answer 3

CAT6 cable by itself is not a problem, it is typically 23 AWG solid core wire which can take you to 4A just fine. The real problem comes from the connectors you use. CAT6 usually goes hand in hand with 8p8c ethernet connectors which only have contacts rated to 500mA.

Also typically CAT6 cable is meant to be stationary (hence the solid core wires), so I'd go for something stranded. McMaster sells some nice cheap cabling that fits your needs, and it's actually meant for moving platforms like a CNC machine.

Answer 4

I've used some cat 6 in one of my printers and just to be safe I used 4 wires for the extruder heat block, 2+, 2-

On the thermistors 1 is more than enough for each +/-.

I also stripped the thick shielding off and used some 'curly' cable organizer in it's place, not the shielding on the wires but the cord. So i could fit 2 cables in slightly more than the same space as one. You are probably doing this to keep it organized too but this is an option if you run out of space in your wire runs.

Answer 5

I was using network cables, for a new tool head (a stealth burner). I needed 14 wires, so two cables would be amazing, as I would have access to 16 wires, at 23 AWG, with an amazing rate of current for everything I needed.

Observation: Some network cables are 24 AWG, so be aware.

I remade the cable five times, and I am going for the sixth. It always broke something, and I thought it was due to so many mistakes that I made myself. It looks like my big mistake was using these cables for moving parts - they are not suitable, for moving, and will break overtime.

One of the mistakes I thought what that I was using solid core, and then I switched to flexible network cable, but same thing - it just broke on me, for the fifth time.

So this time I will waste a few hours researching for a good flexible cable.

I might be wrong, and I might have made a mistake in another area. I will come back and tell you guys if it's the case. However, I truly believe the cable that I used was my biggest mistake.

Network cables are amazing - just don't put them on anything that moves.

Answer 6

It would be best to use stranded wire, since the cables will be subjected to constant motion. The service life will be higher.

Connector current is a issue. My experiments in years past with connectors is that up to 2 amps it is unlikely to be a problem. But pin materials and formulations change.

Were this my printer, I would look for commonalities. For instance, does the thermistor have any direct signals in common with the Z sensor? Avoiding loops is good, so don't have the wires tied together on each end. But, if, for instance, the thermistor was connected to GND or VCC on one of the wires, and the Z-sensor had the same direct connections, I would share the line between the sensors. This saves a wire.

Ideally, the rapidly switching lines would be separated in a different cable from the more static lines. This requires knowing how the lines are driven. For instance, the heater could be driven with a 40 KHz PWM signal, or it could be derived with a 1/2 Hz bang-bang signal. The fans are listed as 2 wire rather than 3 or 4 wire, so rotation feedback may not be used. The fans are likely to he a high-frequence PWM, which is more likely to interfere with the sensors.

The motors are definitely driven with a high-frequence PWM signal.

If you could consider using 3 cat6 cables, you might use:

Cable 1, pair 1: motor A winding Cable 1, pair 2: motor B winding Cable 1, pair 3: hotend fan Cable 1, pair 4: part fan

Cable 2, pairs 1, 2, 3, and 4: heater, each pair carries the two lines so that everything remains balanced. One pair could be reallocated in the future if needed, but best to have the current capacity if available.

Cable 3: pair 1: thermistor Cable 3: pairs 2 and 3: Z-sensor. Double the GND wire, one per pair Cable 3: pair 4: spare