Drag Chain and wire gauge

I looked at those links but I can’t make heads or tails of current capabilities. I am used to 14, 16, 18AWG and so on.

Also, is it the 60Hz from the AC or the noise from the router that is the problem.

Wouldn’t it just be simpler to shield the control cables?

Use this for the control wires instead?

Hey Ziggy,

you find the definitions and wire sizes in American Wire Gauge, it has a chart for comparison of AWG to cross-section area.

As a quick reminder,

6 mm² corresponds to 9 AWG
5.3 mm² corresponds to 10 AWG
2.5 mm² corresponds to 13 AWG
1.5 mm² corresponds to 15 AWG
0.75 mm² corresponds to 18 AWG
0.5 mm² corresponds to 20 AWG

In engineering, wires are not measured in wire diameter but in cross-section area, because the cross section is directly proportional to its strength and weight, and inversely proportional to its resistance. The cross-sectional area is also related to the maximum current that a wire can carry safely. Wire diameter is not.

There is a nice comparison chart for American Wire Gauge to IEC 60228 in this file:

60 Hz is EMI dirt and it is a high current, since it is a .75 kW motor, with 120 V there flows 6.25 A AC. Control signals for the stepper motors are also very dirty and the currents are high, up to 6 A, stepper motors are driven by power electronics. There is nothing that should not be shielded.

You mean EMI will disturb stepper signals more than it will disturb router operation?
But shielding does not only prevent EMI to act on a cable but also to come from a cable. So if you don’t want the stepper signals to be disturbed, I believe both should be shielded. Both carry dirty power signals. And you don’t know what additional wires you will drag in your drag chain in the future.

If you replace the cable of the router anyway, why would you use an unshielded cable at all? For cost? But you do this only once and if you have EMI problems it is sometimes very, very hard to find the cause. I’ve learned to avoid them where I can.

1 Like

Read the reviews?

2.0 out of 5 stars Not appropriate CNC cable
Reviewed in Canada on May 2, 2021
Verified Purchase

This cable is AWG18 4 core with a shield and copper ground line. The cable is sold as “CNC Cable” but is not appropriate for CNC at all as this cable is not designed to support constant flexing and bending. This cable is intended to run along a track and not move once installed. Another example of deceptive product descriptions from Amazon resellers. Typical.

Thanks! I appreciate the detailed feedback.

I was thinking the same thing after posting that link.

I found this from McMaster Carr for data and this for power.

( I have used McMaster Carr before for hardware.)

Perhaps it is because it is late here and I can’t find my glasses… but I not sure where to start on the two links you provided.

I looked for a CNC provider and found the CNCZONE forum where they are suggesting places I know such as Digikey and Mouser. But they don’t mention CNC specific cable.

Maybe you or someone elese can post a link to a specific vendor that services the US. No sense in reinventing the wheel if someone has found a known good cable.

Thank you again for your comprehensive response.

Edit: Found my glasses. I finally saw what I was missing on the LAPP site, “Number of cores and diameter.”


First, thanks for the heads-up regarding EMI. You probably saved me lots of time, money and aspirin.

Here is what I have come up with:


18/3 Chainflex
The Makita router only uses two wires but I was thinking that I could connect the ground wire as normal on a three wire plug (Chassis Ground) and the other end of the wire back to the braiding.

20/5 Chainflex
This is for the stepper motors: Again, I thought I could tie the extra wire to chassis gound on one end and to the braiding at the other end.

As best as I understand you only ground one end of the EMI shield.

McMaster Carr also has these options:

18/3 Continuous-Flex

20/4 Continuous-Flex

I pan on using the Muddyfeet drag chain system (Thanks for the STLs!!)

I hope these wires will be flexible enough.

Any thoughts?

Is there any real-world evidence that drag chains cause issues with the controller or the steppers on the OF? A lot of commercial machines use drag chains, and nearly every 3d printer and my past CNC machines have all had drag chains without any issues. I don’t recall my X-Carve using shielded wires for the steppers. I’d like to see some imperial data on cross talk and signal integrity before I assert there is an issue. Certainly for low level, straight binary control wires I could see integrity issues with high inductance switching loads of the steppers, but there wouldn’t be any of those in the drag chain.

Finally, I’d avoid reading too much into the response of one support person on the use of specific words. They have been using “not supported” on a ton of things recently, from spindles to monitors to wireless game pads. Certainly anything that deviates from the baseline would not be directly supported by the manufacturer; that does not mean it won’t work.


1 Like

I usually try to just get the number of cores I need, then solder a wire (pigtail) on to a length of braid that I separated and twisted together. I have linked two videos that do this in two different ways.
Soldering wire to cable shielding
Splicing shielded cable with solder sleeves
For the stepper cable, you could get the 4 core 18 AWG, or the 5 core 20 AWG you referenced. Many stepper motor suppliers I have seen use 20 AWG. If you get the 5 core, you could use the extra core as a ground between the stepper frame and controller ground. I believe Route1WoodDesign has added such pigtails on the 1F accessory cables he sells.

Note this is different than mitigating EMI using the single point grounded shield you describe above.

Hey Ziggy,

Makita Trim Router with 120 Volts sucks 6.25 Amperes, here it would take at least 1.5 mm² / ≙≈AWG15 (2+PE). For For 240 V / EU version of course you only have half the Amperes so a smaller cable would be enough (it’s not the power (Watts) but the current (Amperes) that gives a wire’s cross-section area).

Steppers in the Onefinity, if they suck about 3 Amperes (do they?), here 0.75 mm² ≙≈AWG 18 would be enough. But the controller provides up to 6 A for each motor, so in case of retrofitting a stronger stepper motor I also would take at least 1.5 mm² / ≙≈AWG15. Here 4 wires are enough, shield to chassis, of course a 5th wire to ground can do no harm.

Note that I’ve seen that often suppliers especially web shops for consumers choose wire strength at the lower limit. But if you calculate the wire strength and you are between two wire gauges, you got to always take the next-higher wire strength.

Hey Tom,

it is not the drag chain itself, but the arrangement of cables in parallel and close to the other. Providing the data you claim would necessite a little research, I’ll see if I find the time :slight_smile: . But you know that a wire arrangement like mentioned above promotes crosstalk. And both types of current are dirty, AC and stepper power signals.

• Avoid parallel arrangement of low-level signal wiring and power-carrying
or noise-prone conductors.
– VFD Manual

And as for low-level signals, who knows if you will retrofit proximity sensors to your machine. The controller supports them as limit switches. Their cable would find place in the drag chain too.

Also you can think of that in the U.S. many people still want to listen to AM radio and then it would be nice if all cables in the workshop were shielded. (even if the sparks of the Makita’s carbon-brush commutator can not be shielded :slight_smile: !)

I agree that going with a bigger cross section - IE Gauge - would be better.

But at what cost to flexibility? I spent a while on the telephone with IGUS tonight. They even suggested buying drag cable that would bend less than the cable to prevent damage.

Frankly at some point it seems to be overkill.

The Makita comes with 18AWG and I believe the stock Onfinity controller wires are 22AWG.

I am now considering these as a good compromise between upgrading the current capability while retaining flexibility.

17/2 1.5" Bend Radius

20/4 1.4" Bend Radius

They are a bit pricey at around $2USD a foot.

Thanks Tom!

I am still learning as much as I can and I appreciate you and all the others that know more than I do!

I had my system up and running briefly making parts for the final build and I ran into all kinds of accuracy issues. They were only around 1mm but that seems like a lot to me.

So I guess I am willing to give anything a try.

I believe you will find that the peace of mind quality shielded cables will give you is worth the cost. It ultimately should help to eliminate areas of potential signal loss/deterioration, as well as provide solid cabling for potential futureproofing/upgrades.

1 Like

Hey Ziggy,

One could easily think so. But I see it this way: When I replace my (household) vacuum cleaner cable, I know that I tend not to move the machine that often and that it is not switched on all the time.

But in a CNC that runs for hours and hours, and the cable is bent (left) and bent (right) and bent (left) and bent (right) and bent (front) and bent (rear) and bent (front) and bent (rear). A drag chain cable is certainly made with very, very fine copper wires so that it can be bent even more easily. But using a cable in a CNC is a completely different burden. And besides, dragging the cable, you only do it once, hopefully.

So with such applications, I’m more likely to be in favor of overkill.

The bending radius of special drag chain cables can be found in the datasheet. Note on the subject of the bending radius, when designing the drag chain you still have variation options with the Onefiniy. You can definitely try to design a larger bending radius.

The above is all very technical; some I follow, some I dont. Would it be easier to strap the router cable to the hose and run both from above?

I now have a spindle, and I purchased shielded cable as I could appreciate that it might interfere with the signal cable from the controller to the VFD no matter where it was (and it was a purchase option). I have this shielded cable strapped to the hose and hung from above; the cable is a bit of a beast and doesn’t bend easily. I have both hung over a piece of large diameter dowel, that I had handy, which rolls back and forth on some timber as the CNC demands.

Hey Andy, hey Ziggy, hey all,

@Dr-Al had the cable attached from the ceiling at that moment which I find a fast and quick solution.

One youtube link that I posted in the same thread has been clicked many times, it shows a simple solution if you find drag chains too complicated.

Note that for the cable it is better to be bent on different parts over its entire length (like in the drag chain) than always on the same part (as shown in the video).

If steppers are so dirty, why dies 1F ship the x and z gantry wires zip tied together? I have run them thru a drag chain. Seems ok, but is anyone having trouble?

What is everyone doing with the router power wire? The plug will not go through a chain unless maybe I disassemble at the router end or cut, run, then repair the plug end.
Just wondering

My understanding is that the router power is what is inducing noise into the stepper wires not the steppers to each other.

My drag chain flips open on each link so you can lay cables in and close it up.

From what understand, the higher voltage and current running through the stepper cabling makes them more ‘noise immune’ than the sensor/signal cabling - I think that is why many CNCs are sold without shielded cabling for this application.

I am not sure what you mean by ‘disassemble at the router end’ but the drag chain would require you to snap open all the links in order to put the power cable through - tedious, but better than cutting the cord IMO.

Thx. Just wondering. I guess I’ll probably go via the hose boom. Seems to be pretty solid answer.