This topic contains 10 replies, has 5 voices, and was last updated by 
Charles Jackson 3 weeks, 3 days ago.
-
Posts
-
After being kindly loaned another Robox after an incident with mine. I found it had similar nozzle fluctuation issues.
Fix was
1) rewire the 24v cable with thicker gauge soft silicon wire
2) Really make sure the head wheel is secure.. i actually redesigned the headwheel to use a hex nut so it wont fail like the Robox OEM one. Can share the file if anyone interested
Anyway Top tip to diaganoze issues either with loose 24v connector or loose head wheel…
Type M104 Sxxx Tyyy in Gcode window and watch nozzles heat up to whatever xxx and yyy are. If the plot in Automaker is very steady then its likely head movement thats causing nozzle temp fluctuations and therefore either loose 24v connector, loose head wheel or both!
Aerospace Engineer with a creative itch being scratched by Robox!
@chaz33gtr Just for info, this is a late 2014 Robox, with a factory DM upgrade, so it has all the revised parts fitted, including the later generation cable support to the rear of the printhead, plus it had only just started doing repeated pauses in printing whilst things heated up in the last month or so.
It has an unknown history, as the previous owner passed away, so I have no idea how hard the head has been tightened up previously, as I just calibrated it, after giving it the once over.
Re the pauses for heating cycles I was seeing, I didn’t really give it much thought, and it all makes sense, re loose connections, and the stiffness & resistance of wiring. I’m not sure the latter is really the case, as CEL should have got this bit right (current rating), but degrading push fit connections with movement or vibrations used to be the norm in automotive circles, and fortunately it has moved on since.
@17bt @chazr33gtr The head power cables are rated for 4.5 amps per the wiring and connectors. What I have found is that early power cables were not crimped correctly, which lead to an early failure of the head power cable. Once the early ones were eliminated, the factory parts are rock-solid. I suspect you received a DM kit with an early power cable. The fittings themselves are Molex and are designed for moving or vibrating environments.
Please note that the M04 command you specify is for the DM head only and not for the SM or S2 heads.
I operate two Betas and four Production Robox.
I am the US/Canada Technical Support engineer for the Robox.
See my 3D Hub site at https://www.3dhubs.com/phoenix/hubs/ben@bhudson I’m sure the current rating is fine, but BTW @chaz33gtr Robox is a factory built DM machine, not an early one by any stretch of the imagination, not even 10 months old.
Unfortunately, your response is somewhat tiresome, because I’ve since looked at the problem too and the @chaz33gtr observations and findings stand on inspection.
That Molex connector definitely isn’t upto repeated vibrations, there is no positive location, so it rocks about the pins, and even if it was…if you look closely you will see the stiffness in the red and black wires transfer the cable support motion into a left right wobble at the connector in the X direction printhead motion, and additionally, in my case with a factory upgraded later model cable support, I can see evidence of the plastic retainer cutting into the red heater wire insulation, which is indication of too stiff an insulation, or too short a cable.
So this whole printhead action along the X axis will in time degrade the outer sleeves, no matter what KoolAid you’re drinking. Sorry to be a Dr Death, but even the Z height motion affects the 2 pin plug in the other direction, so that poor old Molex is getting a right pasting.
Mike
PS I’m sure someone at CEL proper will understand this, and it’s not a biggy to improve & mod for the future, just passing it on @pete. 🙂
@17bt I am the US technical support lead for CEL. I have personally refurbished, supported, and tested over 200 printers. I personally operate 6 Robox printers, all with the cable chain mod, all with over 3000 print hours on the printers. I have seen the behavior reported many times, mostly with the old bundled power cable/Bowden tube design, and have always been able to solve it.
You have also contradicted yourself. You stated that “this is a late 2014 Robox, with a factory DM upgrade” and then you state that “chaz33gtr Robox is a factory built DM machine, not an early one by any stretch of the imagination, not even 10 months old.” If the printer was from 2014, it can’t be a factory DM printer. The factory DM printer is the RBX02 and has silver aluminum wraps on the ends instead of the black. Any printer that has a DM upgrade and black wraps is an updated RBX01 and has a chance of getting one of the older cables.
Yes, there is some movement in the connection, but the Molex Micro-Fit 3 connector is made to allow some movement without degrading current capability. This is why it was chosen. The cable chain mod removes all but a tiny amount of motion from the power cable at the connector. There are however a few DM kits with power cables that were not assembled correctly. I have found about 10, all from DM kits that were made close to the start of the production run. In every case, a replacement power cable has resolved the issue and the behavior has been eliminated.
Based on what you have reported, I suspect that the connector has been damaged. The Micro-Fit 3 uses a tight tolerance barrel to locate the housings to each other and an internal spring contact to maintain positive contact even under vibration. The connector has a spring latch to keep the connector fully seated into the receiver. This connector is an industry standard for this type of connection.
I do not see any evidence of insulation damage on any of my 6 printers, so I suspect that the power cable in question is also an older unit and was cut a little short or has moved a little. Please check that the cable is tight as it runs from the end of the cable chain to the connector on the side of the print volume. If it is not, you can pull some of this extra cable down to the head end. If the cable is too short, yes, there will be a higher load on the cable, but the cable chain should eliminate all but an acceptable level of movement in the cable connector.
Tiresome or not, the only Kool-Aid I am drinking is a lot of experience with the printers, both personal and professional.
@chazr33gtr Please be careful when using the soft silicone sleeved wires. We have found through long-term testing that they do not hold up to repeated bending and fail much earlier than the standard wires. Bed cables made with larger gauge silicone insulated wires fail after as little as 200 print hours and we have stopped using this type of wire for bed cables because it doesn’t hold up.
I operate two Betas and four Production Robox.
I am the US/Canada Technical Support engineer for the Robox.
See my 3D Hub site at https://www.3dhubs.com/phoenix/hubs/benThe confusion over contradiction (which is not the case) is that my Factory DM is about 10months old and has gone pop and is on its way back to CEL.
Mike has kindly lent me one of his older upgrade kit kit printers, Both mine and the upgrade printer experience the same results. Not sure of usage but mine has had almost constant daily use for 10months.
Temperature is rock steady if the head isn’t moving. but add in head movement e.g. a fine rectilinear grid support and it starts happening (and not when swapping nozzles).
I will keep an eye on my silicon wiring but its working lovely so far on both machines, but I struggle to see how it will fail given its desinged to be specifically highly flexible with a 0.75mm2 conductor area, 380strands rated to +150C / -80C and has a current rating of 15Amps
Without wanting to escalate tensions any further, I was just reporting my findings with my 25years experience in Military Aerospace fast jet avionics.. (i.e. hot, high vibration +9g / -3g environments) as I thought CEL might find it useful feedback. Lets not forget this was a kickstarter printer and surely free R&D done by the community at large especially people with significant technical skills should at least be considered if not highly valued to continue to evolve your already good product.
Aerospace Engineer with a creative itch being scratched by Robox!
The connector is not perfect. It was ok before the cables and tubes got constrained in the cable chain and did get an improvement after some failures as Ben mentioned. The design and PCB manufacturing was fairly mature when we realised it was breaking in this area and confused by other issues.
Softer cables may indeed reduce the problem. A different connector could remove the problem.
Good tip with the test Charles, thanks.
For official support please visit www.cel-robox.com/support/ and create a ticketFWIW Thanks for the honesty, and @pete I am surprised you had failures with silicone cable, and having the connector locate and snap into its male counterpart would indeed stop the pins flexing, and transfer the motion into the cable and the crimped connector junction, which might remove the problem.
If you look at the movement of the connector when you flex the cable support, a softer or more pliant cable would not transfer the motion so easily. The connector does move quite a bit despite being pushed hard into its seat.
I use “Thinwall” cable in my car & motorcycle based projects, as it works well with Tyco AMP connectors, and is de facto in automotive circles, because it is a higher current carrying cable in a smaller diameter compared to the normal PVC coated insulation variants, and as @chaz33gtr says the silicone based wiring is used in aerospace as a step up again.
Just another data point:
I’m surprised that silicone wire failed quicker than other options, in the drone world it’s accepted as the best choice. I’ve used it for some other 3D printer builds, and never had a problem with it.
Compulsive tinkerer
I was surprised too, but there has to be reason other than the choice of wire.
I am also using silicone wire in my classic car “points replacement” project, using a 3D printed casing, and for now just using tiny Veroboard circuit boards loaded with discrete components, and my silicone wire is soldered in situ.
However, this does fail where the solder to filament strands join, but only after repeated snapping of the “PCB” in and out of various 3D printed shells, as I tune the sensor position to suit various profiles.
This is to be expected as my multi strand wire gauge isn’t quite large enough, plus I haven’t potted the PCB in the case yet to constrain this junction, so eliminating this weakness, but once I do, I don’t expect any issues with the wire either, as it is flexible enough to allow the baseplate to move under the motion of the vacuum diaphragm.
Thanks for the feedback. If your interested, my printer with the wiring mod should actually be back with CEL at the moment for an unrelated issue.. so feel free to inspect it. I actually went a little further and soldered the wires directly to the head carriage pins, heatshrinked to prevent a short and then cable tied the cable to the cable carriage. So the stress point is now where the cables are pinched by the tie not at the head carriage.
Possibly although i cant confirm this .. maybe the actual fault could be twofold.. the connector working loose and the solder around the pins inside the head carriae fatiguing over time. Obviously heating with a soldering iron from the outside will reflow the solder on the PCB too due to heat transfer.
Your slic3r integration into automaker is awesome by the way,I use it all the time now that mike has shown me how to set it up.
Aerospace Engineer with a creative itch being scratched by Robox!
You must be logged in to reply to this topic.
