I have some 20-conductor, 24AWG cable I want to use as an umbilical. Problem is I don't know if a single 24AWG wire is enough for my hotend.

The hotend may be up to 60W, although less in operation typically. And, the umbilical cable will be about 6-10 feet (2-3 meters) long.

I find a lot of conflicting information on the current-carrying capacity of 24AWG wire...some figures as low as 0.6A/15W, and some as large as 3.5A/84W. It seems to be application specific, and "3D printer umbilical cable" is not a common application. What is your opinion?

Wire current limits are based on the insulation material, and whether the wire is in a bundle (3d printer extruder cable) or in free air. Essentially, the resistance of the wire will cause it to heat up at higher currents. If the increased temperature causes the insulation to soften or melt, it's going to be a big problem. Same gauge silicone and PTFE insulated wires can handle more current than PVC insulated wire because of the much higher melt temperatures of their insulation.

Is it a 12V or 24V heater? Higher voltage heater could use thinner wire because current through the wire is lower.

All that said, I'd use at least 20 gauge wire for a 50 or 60W heater, regardless of voltage.

---

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

I did some testing with the random DB37 cable I bought, presumably 28ga conductors, 6 feet long, and ran current through 6 adjacent conductors, which is how many I plan to use for my printer per the above.

I got a lot of resistance at first, but noticed the heat was coming from the DB37-to-terminal-block adapters I was using....presumably the traces on the PCB are too thin. I replaced them with wires directly soldered to DB37 sockets, and was able to run 11 amps through, with a voltage of 1.35V, or 0.13 Ohms, dissipating about 15Watts in the cable (and test leads).

The cable barely gets detectably warm even after an hour, so I think this is fine from a safety POV, however, I wonder what the implications of dropping the voltage from 24V to 22.65V. I'm sure the fans won't care, and the heaters won't either, but I don't know how the thermistor circuits are wired.

If the thermistor circuits use 24V as a reference, this could throw off the temperature sensing. Based on my limited knowledge, a thermistor should be wired up in a wheatstone bridge or with a 3-wire/4-wire arrangement that separates the voltage detection from the drive current. However in 3D printers I don't see any of these extra conductors that it would take to do a proper 3-wire or 4-wire circuit, so I really don't know how the thermistors are wired. If they are just wired as an element in a voltage divider, then dropping VCC at the printer will cause the reported temperature to drop, causing the system to overshoot actual temperature. One option would be to run yet another 24V wire just as a signal wire for the thermistors, but I'd rather not waste a pin on that.

A Wheatstone bridge is not useful for a normal (NTC) thermistor. In most cases, the thermistor is connected to a reference voltage on the controller through a series (bias) resistor with the other end of the thermistor being connected to ground. The value of the bias resistor is often the 25°C value of the thermistor.

Somewhat better linearity can be obtained by having the bias resistor's value as the geometric mean of the upper and lower temperatures of interest. This is calculated from the square root of (Rmax * Rmin) where Rmax is the thermistor resistance at the higher temperature of interest and Rmin is the thermistor resistance at the lower temperature of interest.

The reference voltage is typically the main voltage for the controller board, usually 5V or 3.3V - some controllers can have a reference supply for the bias resistor and the ADC.
DB37 connectors are generally not a good choice for current-carrying, even where the original specification gave an adequate current. This is because there has been a lot of progress on general connector design since the DB series was first designed - something like Molex Microfit is better.

Edit: A quick edit to add a photo showing very thin flexible wires plaited to be semi-self-supporting. The wires are 7/0.12mm and the 4 pink wires carry a total of 1.8A while the 2 white wires are for a thermistor. The brown solid wire through the middle is a Nitinol Superflex wire to support the bundle and entirely replace a cable chain.

Using multiple thinner wires has advantages for supplying the hotend as thinner wires with thinner insulation can be air cooled for a better current carrying capacity while being lighter. It should go without saying that anywhere you are pushing the limits of current carrying capacity the wires should be in open air along their entire length. Using multiple wires offers the possibility of making a cable bundle that is more flexible yet self-supporting by plaiting them together.

Mike

Edited 1 time(s). Last edit at 06/28/2024 02:12PM by leadinglights.

That is a very interesting cable bundle and gives me a lot to think about.

Back to the thermistor circuits, I still wish the blasted mainboard companies would just publish electrical drawings, but it seems this is impossible to get. But I see that the thermistors are plugged into the main board with a 2-conductor plug. However, I only want to run 1 wire through my umbilical for each thermistor. One of the 2 wires must be the voltage used to sense the temperature, and the other one must either be ground, or a supply voltage to the thermistor.

If it's sense+ground, which I think is what you said, that means I only need to run the sense wire, and I can attach the other thermistor lead to shared ground at the printer. The thermistor won't be impacted by voltage drop from other things on the circuit (assuming ground potential doesn't change), because the thermistor will be "driven" by the mainboard through the "sense" wire...apparently about 2.5V.

If it's sense+voltage, I still need to run the sense wire, and I need to know what the supply voltage is. If it's 24V or 5V I can probably also pick that up at the printer, but need to be careful it doesn't experience voltage drop from the heaters or it will impact reported temperature.

Quote
PCLoadPLA
Back to the thermistor circuits, I still wish the blasted mainboard companies would just publish electrical drawings, but it seems this is impossible to get.

Duet boards are open source and the schematics are available at [github.com]. Most Chinese boards are closed source.

On most boards one side of each thermistor input is connected to analog ground and you can use a common analog ground wire for multiple thermistors. The analog ground pins of the thermistor connectors are labelled AGND or VSSA on Duets and some other boards.

Edited 1 time(s). Last edit at 06/29/2024 03:36AM by dc42.

---

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].

" Most Chinese boards are closed source." this is over stating things

MKS and Bigtree Tech are both Chinese companies and provide circuit diagrams for most of their stuff.


From a previous post PCLoadPLA said you have a Artillery Sidewinder X1 which has a BOARD_MKS_GEN_L which has circuit diagrams [github.com]