Flat ribbon cable to hotend and extruder - is it a good idea?

I just got this ribbon cable:



Mine is 26 ways 28 AWG has stranded cores and seems to be quite flexible. I am thinking about using it to wire the hotend, extruder, fans and BLtouch. From SKR 1.3 schematics as far as I can see I could use common ground as all ground connectors seem to be connected together on the board. Thermistor ground might be different as it is labelled 'GND analogue' not just 'GND', so I am planning on having separate wire for thermistor ground. My plan is to use wires in the ribbon in the following way:

1 Motor faze 1
2 Motor faze 1 return
3 Motor faze 1
4 Motor faze 1 return
5 Motor faze 2
6 Motor faze 2 return
7 Motor faze 2
8 Motor faze 2 return
9 Ground
10 Ground
11 Gound
12 BLtouch VCC
13 BLtouch signal
14 BLtouch z endstop
15 Hotend fan
16 Part cooling fan
17 Hotend heater VCC
18 Hotend heater ground
19 Hotend heater VCC
20 Hotend heater ground
21 Hotend heater VCC
22 Hotend heater ground
23 Hotend heater VCC
24 Hotend heater ground
25 thermistor
26 Thermistor ground

Stuff in bold will share ground. All 3 wires will be connected to the ground PSU connector on the SKR 1.3 board. Heater will use 4 wires as it probably needs most current. 2 wires per faze hopefully should be enough to power a pancake motor which requires up to 0.7 A.

What do you think about my idea? Would different order of wires be better? Any suggestions?

That type of ribbon cable isn't made to flex a lot- it's more of a static installation type stuff, so you'll need to make a big loop with it to minimize flex or you'll have wires breaking inside the insulation. The problem you're going to run into that you have to terminate each end of the ribbon. That means laying out PCB''s with screw terminals or similar connectors. That's a lot of extra work compared to just crimping connectors onto the ends of wires. 28 ga wire seems pretty tiny, even using multiple wires to carry power to the hot end heater. I suspect the resistance of the wires and connectors may be a problem. You can get ribbon cable with heavier gauge wires, and even twisted pairs.

I used a flex ribbon in a printer I built about 6 or 7 years ago. It's still working, but I remember the PITA of laying out the termination boards, and I was just hand wiring them. I was in too much of a hurry to finish to wait for PCB turn-around. Maybe you're better with circuit board layout software than I am and have a bit more patience and that part will be easy.



The flex ribbon came out of a tape library robot that was dismantled at the Makerspace. Once in a while you can find them listed on ebay for reasonable prices.

Edited 1 time(s). Last edit at 05/07/2020 07:49AM by the_digital_dentist.

---

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

To terminate the ribbon, I was planning to use this little bord which is really cheap with the connectors shown below.




I am going to replace the socket with the latched one which should prevent the plugs from coming out. The current and the resistance of the wires is my main concern, but I thought that 4 wires in parallel should do the job. I guess I have to do more research on what is actually the current for the hotend heater (unfortunately I am using 12V PSU so it might be quite high, but the original heater wire doesn’t seem to be that thick)
I am aware that I will have to use large loop to prevent too much bending.

Does order of wires in the ribbon make a difference?

It was used quite successfully on the Mendel 90 for heated bed and extruder.. but they used d-type plugs so was a lot of soldering.

See [reprap.org] for details on this now 8 year old machine...

All that board is doing is converting the female connector on the ribbon to a bunch of pins. You can do that by just plugging a header into the female connector on the cable- no board is needed. But how will you connect your hot-end heater (or any of the other parts of the extruder carriage, or controller board) to the pins?
The order looks OK. Generally, it is best to keep high current, low resistance, switched lines away from low voltage, hi-Z lines like the thermistor input because of the potential for noise on the hi-Z lines. It is also best to use twisted pairs on those high current lines like the motor windings and hot-end heater. You can buy twisted pair ribbon cables. Here's an example: [www.mouser.com]
You still have to figure out how to get from the ends of the cable to the extruder carriage and controller board...

Edited 1 time(s). Last edit at 05/07/2020 10:30AM by the_digital_dentist.

---

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

You are right. I just realised that I don’t actually need this little circuit board. I can just attach the socket to the X carriage and then solder all the wires (BLtouch, heater, thermistor etc…) to the pins of the socket.
I know that twisted wires would be best, but I didn’t know that twisted ribbons actually existed. I have already bought the ribbon cable and I think I will go with it. In the original wiring of the printer (Anet A6) wires going to the x carriage were just bundled together (not even twisted) and it worked fine so I hope that twisting of the wires is not going to cause any problems.

NB not all ribbon cable is the same gauge wire

Check yours is up to the task of powering your hotend without itself getting hot. (its awfully embarrassing when your wires decide to strip their insulation!)

I have just measured current drawn by the hot end heater. It is 2.1 A max. With 5 wires in the ribbon (each is 28 AWG) for current in and another 5 for current return should be enough.

But I still have the main question. Can both GND wires of the BLtouch, GND of the part cooling fan and GND of the hotend fan be connected to the same point on SKR 1.3. If yes what would be the best place to connect all those GND's on the board?

GND of hot end and fan both go to 12/24v when off (0 when on) so is not compatible with digital pin gnd's

Also since the hotend and fans the GND is switched, you cant join these either... as they would interfere with each other.

Edited 2 time(s). Last edit at 05/08/2020 08:02PM by Dust.

Quote
Dust
GND of hot end and fan both go to 12/24v when off (0 when on) so is not compatible with digital pin gnd's

Also since the hotend and fans the GND is switched, you cant join these either... as they would interfere with each other.

WHAT ?????

---

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

@MKSA

On most boards The -ve pin on the mosfets plugs are not actually GND's

if you measure voltage from the power supply gnd to a mosfet + pin is will always read +12v or +24v what ever your system is on (some board do generate 12v from 24v for fans)
if you measure voltage from the power supply gnd to a mosfet - pin it will read 12/24v when off (ie difference between + and - pins is 0v)
and will read 0v when on. (ie difference between + and - pin is 12 or 24v)

And yes this confuses most people trying to debug why their mosfets don't work.

technically the -ve pin floats, but 12/24v flows through the connected device (fan/heater etc) and you read 12/24v

Edited 2 time(s). Last edit at 05/09/2020 02:29AM by Dust.

Quote
Dust
@MKSA

On most boards The -ve pin on the mosfets plugs are not actually GND's

if you measure voltage from the power supply gnd to a mosfet + pin is will always read +12v or +24v what ever your system is on (some board do generate 12v from 24v for fans)
if you measure voltage from the power supply gnd to a mosfet - pin it will read 12/24v when off (ie difference between + and - pins is 0v)
and will read 0v when on. (ie difference between + and - pin is 12 or 24v)

And yes this confuses most people trying to debug why their mosfets don't work.

technically the -ve pin floats, but 12/24v flows through the connected device (fan/heater etc) and you read 12/24v

I react to the "GND of hot end and fan both go to 12/24v when off (0 when on)" which in fact is NOT a GND.

Indeed hotend, part cooling fan are in most boards permanently wired to V+ and powered via a N mosfet connecting them to 0V.

Note I am familiar with signal ground, digital ground, power ground , shield, "earth" and the funny issues when you improperly connect them as they have to be connected, like ground loop, noise etc.... and therefore I react that way.
Plenty of 3D printers are improperly wired, the worst being the famous main ground/earth that poses a real threat with the cheap chinese power supplies.
Even some controller boards are improperly designed regarding their power tracks/connectors.

---

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

After considering all issues - current limitation, not being able to use just one wire for the ground and the fact that it is actually going to be quite difficult to physically attach the ribbon in a way so it is not bent too much and is not in the way I am giving up the idea of using ribbon to wire my hot end. When i rebuild my x axis I will just stick to the factory Anet A6 wiring method - just bunch of wires bundled in a plastic slewing.

Proper wiring of our 3d printers is not an easy task...

Many thanks for all help.

A method I have found very satisfactory is to braid (plait) the wires. This eliminates the need for sleeving, gives better cooling of the high current (heater) wires, is more flexible than a sleeved or even a naked bunch of wires, has reduced electrical interference as well as reduced fatigue at the wire ends and quite possibly cures scrofula - I like it anyway. The picture below shows braided 7/0.1 wires with a springy Nitinol wire to give the bundle a little stiffness and keep it away from snagging bits of the printer when moving. Self-supported bundle on the right



Mike

I like it!

Looks like my next upgrading is by braiding.

Twisting, not braiding is recommended to reduce noise.

I have tested braiding as shown in the photograph and it is effective in reducing noise. it is also a method sometimes used in low level multi conductor signals cables e.g. ecg or multiple photodetectors.

Mike

You can always braid twisted pairs....

I use this to do the twisting: [www.youmagine.com]

https://vimeo.com/213427411

Edited 2 time(s). Last edit at 05/11/2020 07:07AM by the_digital_dentist.

---

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

Quote
etfrench
Twisting, not braiding is recommended to reduce noise.

Quote
leadinglights
I have tested braiding as shown in the photograph and it is effective in reducing noise. it is also a method sometimes used in low level multi conductor signals cables e.g. ecg or multiple photodetectors.

My understanding is limited, and I'm not an electrical engineer, so I'd like to understand better here.

My understanding has been, roughly: Twisting cancels noise if you are twisting circuit pairs because the EMF cancels out. Braiding reduces noise in two ways; it causes the wires to cross at mostly roughly right angles so that it mostly doesn't induce current along the crossed conductors, and it increases the space between adjacent wires substantially, which decreases noise exponentially with distance.

Quote
the_digital_dentist
You can always braid twisted pairs....

Since you're an actual engineer and I just "play one on the internet," I'm curious: Is there any electrical noise benefit to braiding twisted pairs, or is it only for physical and/or aesthetic value?

Thanks!

Braiding cables keeps the wires together without having to use any extra jacketing or sleeving. It's looks nice, too. It can make rewiring things a PITA, but so can a sleeve, depending on the type.

Twisting high current wires keeps the magnetic fields mostly confined to the wires and reduces currents induced in adjacent wires. Motor drivers are high current/low Z connections. Twisting wires doesn't do much to prevent capacitive coupling, but capacitive coupling is primarily an issue with high Z connections. Shielded wire limits capacitive coupling and does almost nothing to prevent inductive coupling.

---

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

Braiding together twisted pairs would electrically be better than the wide braided cable in the photograph. Cross talk is not reduced as much between the cables in the flat braid but is about as good with when related to conductors not in the braid. The advantages of the flatter braid are in the greater flexibility - in the case of braided twisted pairs, each twisted pair is slightly less flexible than two loose conductors. The other advantage of the flat braid lies in the cooling: The 4 pink wires in the photographed braid are 7/0.1 wires to the heater element, 2 each for positive and ground. The free flow of air along with the reduced thickness of the insulation and greater surface area of the insulation allows this to be used for up to 2 amps with only a few degrees of temperature rise. Bundled wires in a sleeve would have a greater temperature rise.

And it also looks good.

Mike

Thanks both!

(I've used multiple smaller wires for increased cooling in my electric tractor conversion, where a loose bundle of 12awg wires is cooler than a previous pair of 4awg wires of larger cross-sectional area. There, I am sometimes pulling more than 200A, so it's a different scale, but same concept. ☺)