Wire gauge and a near serious accident

Hi,

I am posting this as a safety warning and a suggestion regarding best practices for wiring the heatbed.

Yesterday while testing a MK2 heatbed for the first time I had the "magic smoke" coming out of one of my printer builds.

It turned out not to be PSU, not the RAMPS, not the heatbed either: the wiring from the RAMPS to the heatbed had overheated and melted.

The accident took place in less than one minute, showing things can go from normal to really bad in just a few seconds. I turned on the heatbed, 40 seconds later I noticed some smoke coming out of the printer, and by the time I reached for the stop switch I had already breathed in a good amount of toxic fumes. I am guessing a few seconds more and things could have been much, much worse.

Here is the piece of charred, melted cables:



Here is the same cable, stripped, showing the amount of copper, clearly insufficient to carry 10~11A, even when doubled.



This is the cable I am replacing this with, noticed the much larger amount of copper:



Lesson learned? I hope so. I'll be taking the heatbed wiring more seriously from now on.

A few guidelines for myself and others:

1. Always test the heatbed wiring for resistance. Remember, at 10A, 0.1Ohms means 1V voltage drop means 10W dissipated by the wiring!
2. Always test a new heatbed.
3. When testing a new heatbed, stay close to the printer and keep your hand close to the power switch.
4. Use a thick wire gauge for the heatbed, and verify the amount of copper by stripping a good bit of wire.
5. Make sure whatever connectors you are using are properly tightened and secured. A bad contact means overheating and things melting, burning and eventually causing a serious accident.

I always overbuild stuff. My 6A bed is being run from 8 gauge super flexible ptfe core car audio wire and it is soldered directly on the controller. Obviously such a massive wire gauge is overkill, but is is super bendy and it stays stone cold. Good luck with your wiring and be careful!

Quote
iamdarkyoshi
I always overbuild stuff. My 6A bed is being run from 8 gauge super flexible ptfe core car audio wire and it is soldered directly on the controller. Obviously such a massive wire gauge is overkill, but is is super bendy and it stays stone cold. Good luck with your wiring and be careful!

That's very good advice, thanks! Point taken and I'll be extra careful from now on!

I've used some left-over cable from my R/C models in my printers - it's very flexible, has a very high current carrying capacity, and has a silicone rubber sleeving that can resist up to around 200C. My delta is wired up with 16AWG silicone cable that has a capacity of 22A - more than enough for my needs, considering that the heat-bed will take around 11A or so (wired separately from the RAMPS/Arduino and its stepper motors + hot-end). Not too expensive either.

Here's a UK supplier - Component Shop - but any model shop that sells make-it-yourself electric radio-controlled models will stock it.

Edited 2 time(s). Last edit at 04/13/2015 05:25PM by David J.

What gauge wire was the one that burned up?

And what size are you using now?

I am about to buy or take apart electronics for to find the wiring I need to hook up a 310mm onyx heatbed, and all the power. Making sure I don't have this happen.

Check the suggestions above by iamdarkyoshi and David J. In my case, I believe I am using 14 gauge speaker wire right now, that seems to be good enough for the approx. 10A MK2B heatbed I am using.

This is a usefull website to check which gauge wire you should be using.
AWG Calculator

Ohm does not like his law being violated.

Heated beds are big current eaters. Mine is 34 amps when running. This requires serious copper - running 10 gauge stranded feeder from the power supply.

Quote
itchytweed
Ohm does not like his law being violated.

Heated beds are big current eaters. Mine is 34 amps when running. This requires serious copper - running 10 gauge stranded feeder from the power supply.

For that sort of power, IMO a mains powered heated bed is generally a better option. That is what I have on my large delta. Of course, using mains voltage to heat the bed brings an additional set of safety concerns, especially on a machine with a moving bed.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
dc42

Quote
itchytweed
Ohm does not like his law being violated.

Heated beds are big current eaters. Mine is 34 amps when running. This requires serious copper - running 10 gauge stranded feeder from the power supply.

For that sort of power, IMO a mains powered heated bed is generally a better option. That is what I have on my large delta. Of course, using mains voltage to heat the bed brings an additional set of safety concerns, especially on a machine with a moving bed.

Mains voltage on a heated bed? That brings it own huge bag of issues with it. Going from low voltage to high voltage ( > 48 volts ) and the attendant issues of arc flash protection, different wiring practices, etc. Even though the current is higher, I will stick with the 14.2 volts that runs the whole show.

Quote
itchytweed

Quote
dc42

Quote
itchytweed
Ohm does not like his law being violated.

Heated beds are big current eaters. Mine is 34 amps when running. This requires serious copper - running 10 gauge stranded feeder from the power supply.

For that sort of power, IMO a mains powered heated bed is generally a better option. That is what I have on my large delta. Of course, using mains voltage to heat the bed brings an additional set of safety concerns, especially on a machine with a moving bed.

Mains voltage on a heated bed? That brings it own huge bag of issues with it. Going from low voltage to high voltage ( > 48 volts ) and the attendant issues of arc flash protection, different wiring practices, etc. Even though the current is higher, I will stick with the 14.2 volts that runs the whole show.

Mains voltage does indeed bring additional safety concerns such as protective grounding, insulation, and preferably RCD protection; but it avoids the need for huge power supplies, switches to control high currents, and very heavy gauge wires. Delta printers are particularly suited to mains-powered heated beds, because the bed is fixed, and all the mains wiring (including the PSU for the rest of the printer) can be put under the bed and made inaccessible to prying fingers.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

May I know what is your wire gauge, sleeve/insulation type, length and resistance?

Quote
sarf2k4
May I know what is your wire gauge, sleeve/insulation type, length and resistance?

If you are referring to my mains powered heated bed, then the bed heater was supplied with a 2 core double insulated silicone lead. For the mains wiring under the bed, I used the cores from 3 core 5A mains cable. See [miscsolutions.wordpress.com] for photos.

A moving mains powered heated bed would require extra flexible cable with a suitable voltage rating, such as the type of cable used for multimeter test leads, with strain relief at both ends.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I bought a 16awg wire with silicone sleeving rated around 100'c, 5 meters for a pair of them from aliexpress. I didn't cut them to my required minimum length yet so the wire resistance are rated at 0.4ohm at 5 meters.

This wire will be used to connect from the psu to the board. I still haven't measure the current stock wire that I have now on my gtech prusa printer

Quote
sarf2k4
I bought a 16awg wire with silicone sleeving rated around 100'c, 5 meters for a pair of them from aliexpress. I didn't cut them to my required minimum length yet so the wire resistance are rated at 0.4ohm at 5 meters.

This wire will be used to connect from the psu to the board. I still haven't measure the current stock wire that I have now on my gtech prusa printer

Here is a chart of resistance per meter for various copper wire gauge:

[www.daycounter.com]

So either there is a very big error in your measurement or your copper wire gauge is incorrect.

Personally I have settled on inexpensive 14 gauge (1.5mm^2 here in Europe) multi strand speaker cable. It does the job for an MK2B heatbed at or around 12V/11A and costs less than 1€ per meter. Also check the other recommendations above. And as dc42 wrote, for a moving bed some sort of strain relief is recommended at both ends

Not another awg resistance reference again

Thanks for sharing this..

Nice gentle reminder to improve wiring before it's too late.

My bed is 24v at 20amp and this is really weird but I have 14ga stranded wire and near the power supply I noticed some discoloration and the sucker burnt me. I tested it and the wire is about 70c but very strange is that just 100mm away the wire is only warm and near the power fet it is room temp. I just ordered some 12ga silicon wire so that should work considering the wire is less than a meter long.

---

_______
I await Skynet and my last vision will be of a RepRap self replicating the robots that is destroying the human race.

When dealing with high currents the ability of the conductor to dissipate the heat is important. Even very thick wire can overheat if it runs through an area where it can't be cooled. Un-necessarily thick insulation (including double insulated cables) bunching the cables with others, running cables through or under thermal insulating materials and passing the cables through an already hot environment all prevent good cooling.

Even the lists of current ratings such as the one quoted previously ( [www.powerstream.com] ) are not to be trusted im some circumstances: An extension cable rated at 13 amps but fully rolled up and covered by a hay bale has been known to start a fire even though it was supplying a bank of lights drawing 4 amps at most.

Conversely, if the resistive heat can be dissipated, then really quite a thin cable can be used. On one of my heated beds, I use 2 lengths of 22awg thin-wall silicone covered wire for each connection (4 in all) These are plaited (braided) into a flat open assembly and the whole assembly is in open air, connected at each end with 20A connectors. Even at 12A, the temperature rise is well under 10 degrees.

It is possible to calculate the temperature rise of a cable by finding the resistance with a low current and at the working current and using the temperature coefficient of resistance of the copper.

Mike

Jeez. I know the prime spec for most hobby printers is that they have to be made of all the cheapest parts on the planet, but come on, wire is cheap. Why does every part in a printer have to operate right at the hairy edge of its specs?

If there's any doubt, use thicker wire. Sure, it might cost you a few cents more, but it may keep your house from burning to the ground. Ever heard the phrase "penny wise and pound foolish"?

If funds are so tight that you can't afford to wire the bed properly, or lack the imagination to scrounge wire out of a junked appliance, you should probably find a lower cost hobby that doesn't require so much creative thinking. How about paint-by-numbers?

Edited 1 time(s). Last edit at 09/27/2019 06:54AM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

There are good engineering reasons not to use wire that is too thick. Insufficiently flexible wires and the addition of moving mass to the bed are points to bear in mind as well.

The heating that Dark Alchemist mentions is possibly due to high resistance joint in the connector/crimped terminal/soldered joint at the power supply end.

BTW, the 22 AWG (0.327mm²) thin-wall silicone wire that I mentioned above was about twice the price of 14AWG (2.08mm²)

Mike

Quote
leadinglights
There are good engineering reasons not to use wire that is too thick. Insufficiently flexible wires and the addition of moving mass to the bed are points to bear in mind as well.

The heating that Dark Alchemist mentions is possibly due to high resistance joint in the connector/crimped terminal/soldered joint at the power supply end.

BTW, the 22 AWG (0.327mm²) thin-wall silicone wire that I mentioned above was about twice the price of 14AWG (2.08mm²)

Mike

That is the strange part because there is no kinks or bends in it but it does have spade lugs to go into the terminal so with the silicone wire I am not going to use them again.

---

_______
I await Skynet and my last vision will be of a RepRap self replicating the robots that is destroying the human race.

Quote
Dark Alchemist
That is the strange part because there is no kinks or bends in it but it does have spade lugs to go into the terminal so with the silicone wire I am not going to use them again.

A poor crimp will lead to a poor contact that will worsen due to the heat cycles. Most failures occur at the contacts, not the wire itself unless grossly undersized.

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)

Quote
MKSA

Quote
Dark Alchemist
That is the strange part because there is no kinks or bends in it but it does have spade lugs to go into the terminal so with the silicone wire I am not going to use them again.

A poor crimp will lead to a poor contact that will worsen due to the heat cycles. Most failures occur at the contacts, not the wire itself unless grossly undersized.

Yep, but they haven't discolored a bit just the 100mm of wire leading up to them. I bet they are getting hot and the heat is leading down the wire though it is strange they have no discoloration anywhere on them. I sized my wire based on the total length of the run and the amperage demanded of it and my figures never said it should get to 70c+. I figured it would be warm to the touch now if I doubled my length I would have needed a higher gauge. As it is the 12 gauge silicon wire should take of it and no more spade lugs but they sure were handy.

---

_______
I await Skynet and my last vision will be of a RepRap self replicating the robots that is destroying the human race.

I assume that the spade lugs are crimped on. Crimping is the best way of connecting a wire but it comes with a big BUT. It is only good if it is done right. The metal of the wire and the lug must be deformed sufficiently that there is a large area of contact and, very importantly, the deformation must be plastic: Metals, when bent slightly, will spring back but when deformed past a critical point will deform plastically and not spring back. Crimping with a worn out or otherwise inadequate crimping pliers risks not getting to the plastic deformation and the metal springing back and giving a poor contact.

The genuine manufacturers crimping pliers can be breathtakingly expensive but some cheap ones will do the job but don't take it on faith. Try making a few test crimps and try to pull the wire out of the crimp using a lever if necessary - if it can be pulled out then the pliers are not up to it.

Soldering is genuinely not as good as a sound crimped connection but much much better than a bad crimped connection.

Mike

Quote
leadinglights
I assume that the spade lugs are crimped on. Crimping is the best way of connecting a wire but it comes with a big BUT. It is only good if it is done right. The metal of the wire and the lug must be deformed sufficiently that there is a large area of contact and, very importantly, the deformation must be plastic: Metals, when bent slightly, will spring back but when deformed past a critical point will deform plastically and not spring back. Crimping with a worn out or otherwise inadequate crimping pliers risks not getting to the plastic deformation and the metal springing back and giving a poor contact.

The genuine manufacturers crimping pliers can be breathtakingly expensive but some cheap ones will do the job but don't take it on faith. Try making a few test crimps and try to pull the wire out of the crimp using a lever if necessary - if it can be pulled out then the pliers are not up to it.

Soldering is genuinely not as good as a sound crimped connection but much much better than a bad crimped connection.

Mike

Nope, they were crimped with a proper ratcheting tool with no bends or kinks.in them.

---

_______
I await Skynet and my last vision will be of a RepRap self replicating the robots that is destroying the human race.

Silicone wire is aluminum (forgot about that) so has the same current carrying capacity as the 14ga pure copper but it isn't even warm to the touch now.

---

_______
I await Skynet and my last vision will be of a RepRap self replicating the robots that is destroying the human race.

You should get 12 guage silicone coated wires, in that way it won't melt wire coatings. My Tronxy 802D hotbed came with two red thick coated12 guage wires and its coating material made of silicone. You might want to check out

This is what I'm talking about on link.. It's just 18 guage braided wire for high temprature but I wouldn't recommend for Hotbed, it's for Hotend tho.

[www.showmecables.com]

Same wire what my Hotbed came with:

[www.amazon.com]

I think that you're about 2 years too late with this information...

One very important thing about cable for high current is that they should be in free air. There is a lot of difference between the current-carrying capacity of a cable that is in the open and one where the heat can't escape - That is the reason that all of the cable on an extension cable drum should be loose outside of the drum if it is to be used anywhere its maximum current. Thick electrical insulation is also a thermal insulator and the best cable for heated beds is a cable with many fine strands and a thin, flexible and abrasion-resistant jacket. Multiple cores also help with the heat dissipation as there is more surface area on the jacket relative to the cross sectional area of the conductors. It goes almost without saying that there should be a good fuse/circuit breaker and/or current limiting power supply for the heater.

Mike