Bond Tech Extruders . Any good, whats the improvement?

Thoughts about Raise3D, 3D printing and making in general.
Spikeysonic
Posts: 98
Joined: Wed Mar 01, 2017 12:07 am

Bond Tech Extruders . Any good, whats the improvement?

Postby Spikeysonic » Sat Nov 04, 2017 12:36 pm

Bond tech extruders. Ive had nothing but grief with anything other than PLA. Ive seen suggestions of getting a Bond Tech extruder to replace the Raise 3d one. Are they any good, what are the advantages, how easy is it to uninstall the previous and install one of these Bond Tech Extruders when not done anything like that before. What problems does it solve, does it cause any problems or risk.?

it mentions It has come to our knowledge that if the ribbon cable to the break-out board gets unplugged when the printer is turned on, thermal runway might happen, make sure that the ribbon cable is securly held in place with hotmelt or a zip tie.

Whats a thermal runaway?


ie all detals please

Jetguy
Posts: 2207
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bond Tech Extruders . Any good, whats the improvement?

Postby Jetguy » Sat Nov 04, 2017 2:47 pm

Thermal runaway is when for some reason the controller is not able to sense the actual real physical temperature of the heater and in this state, thinks the temperature is lower than the setpoint and thus the heater should be on. Because the controller and thus firmware is not getting proper information (actual heater temperature) it cannot make proper decisions on how much power and for how long the heater should be on.

This manifests itself because of some assumptions based on mechanical facts.
The heater is a separate device than the temp sensor. The heater block is the mechanical and thus thermal connection between these 2 devices. Luckily for us, the mechanical arrangement is such that even a loose setscrew thus not locking the heater cartridge or thermocouple temp sensor- they are both somewhat unlikely to completely side out of the heater block inside the printer in normal operation. This is not always true on other brands of printers and this design and mechanical details on the Raise 3D N series does contribute to safety.

That said, it's not impossible to have a fault. The first known fault is that the early versions of thermocouples if damaged, the actual probe and thus sensing junction of the thermocouple could pull out of the tube and thus the tube stays in the heater block and the temp sensor is no longer in the tube, thus cannot sense the actual temp of the heater block. Again, V2 thermocouple construction is improved over early kickstarter supplied thermocouples and so any new printer has a new mechanical design that largely prevents this specific condition.

OK, so that covers most of the mechanical conditions. Again, assumption time!!! The system assumes the heater cartridge and thermocouple temperature sensor are "coupled" together by the mechanical heater block and that in no way can those 2 come apart and that the temp sensor is at least able to report and sense the actual temperature of the heater block assembly.

NEXT, is that tiny microvolt signal of the thermocouple is read by the thermocouple amplifier ICs on the extruder head breakout board and converted to a higher voltage logic signal that can travel across the long ribbon cable back to the mainboard. Thermocouple amp chips are a precision calibrated industrial chip. As such, they have a lot of safety features built in. One example is basic calibration. The chip actually has a temp sensor build in measuring the very temp of the chip itself. This is because a thermocouple sensor is actually sensing the DIFFERENCE in temperature between the hot end at the heater block and the cold end at the screw terminals of the extruder head breakout board. back to assumption time, the idea is the distance between the chips and those thermocouple screw terminals is short and there are no heat producing components on the board that would throw off this assumption by heating the screw terminals or the IC itself to be different than the screw terminals. So what happens is, this combination means that any 2 raise 3d printers and any thermocouples- are so accurate they are within 3C of each other!!! But back to safety, the thermocouple amp ALSO can sense the INSTANT a thermocouple circuit or wire is open circuit. In this case the thermocouple chip sends an extremely high temp report up to the marlin motion control board. Again, the extra safety layers are that the thermocouple amp calibrates the thermocouple by reading the ambient temp and it's own internal laser trimmed factory calibration and also can sense the wiring fault of an open circuit thermocouple.

Where this has gone wrong and the recently discovered assumption that did not fall true. The assumption was that if the ribbon cable between the extruder head breakout board and the main controller failed, that a fault would be detected. This assumption is sadly false. Again, from a wiring high level diagrame The heater and thermocouple both get wired to the extruder head breakout board. If the temperature sensor wiring connection fails open circuit (loose or broken wire to the screw terminals) this fault is detected instantly and shuts down any heat because a high temp is reported. The problem is in the details, The breakout board has 2 connections, one is the non locking ribbon cable, the other is a 3 pin locking connector carrying the actual heater power directly switched on and off by the mainboard. So the temp sensor is on a non locking connector but the heater power is on a locking connector. This means that the chance is, the heater connector will stay locked- because it has a lock, and the sensing connection in the non locking ribbon cable MIGHT come free during motion or vibration. Since again, we found a condition where if and when this ribbon cable comes disconnected, the MAINBOARD is unable to detect this fault. This is not a firmware thing, this is a hardware thing!!! At the hardware electrical level, once the system is in operation and has read a valid temperature, if that ribbon cable comes disconnected, then the mainboard is unaware and the reported temperature remains the last temp. Again, this is the major concern here. Instead of sensing a high temp and thus being higher than the setpoint instantly causing both a logical warning in the code and also shutting off the heater, instead, the very hardware sensing pin of the mainboard thinks the temp is perfectly normal and is just the last reading received.

Which comes to the next layer of system- firmware. Firmware is just logic making decisions based on information. If you have a hardware condition where incorrect or invalid temperature readings are being read and that goes into the decision process to control the heater, there is limited safety code or decisions you can make to capture this unsafe condition. Again, the issue is, the firmware doesn't know there is a fault because the hardware does not know there is a fault.

So, the key is, this is NOT specific to the Bondtech. ANY raise 3D printer, stock or whatever, if the ribbon cable fails in operation, the above fault condition is possible. I'll let you think about this on your own and determine how likely that is- but again, understand the condition has little to do with the Bondtech. iI just so happens when changing or repairing the extruder head- they may move or touch the extruder head breakout and ribbon cable. But it does not require that to cause the fault. Just using the printer- long enough time, moving around, the cable can come unplugged because it is not a locking connector.

Jetguy
Posts: 2207
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bond Tech Extruders . Any good, whats the improvement?

Postby Jetguy » Sat Nov 04, 2017 6:20 pm

The second part of this is why do the bond tech upgrade at all?

This just comes down to basic extruder knowledge and understanding.
What we are really talking about changing is not the entire extruder but only the feeder portion. The hot end and nozzle remains the same on this current series of upgrades.
So the real answer is why would you want to change the stock feeder?
#1 The stock feeder is heavy. I mean REALLY heavy. Unless you take it off, many of you might be completely blown away with how heavy it actually is. The Dual extruder is even heavier. It's more than 2X heavier than the stock single extruder feeder. One part of the weight is the motors, but the other part is the massive CNC milled aluminum bar or block that forms the bulk of the feeder. Since the entire gantry system of rods has to support this mass or weight, this is why this matters. Even when you remove the second motor on a dual extruder stock setup, the dual extruder feeder assembly is still more mass and thus the effect of just taking the motor off is only marginally effective.
#2 The stock extruder uses a fixed distance idler that is adjusted by taking the cover off and moving the motor in it's mounting holes. Nothing about that is a precise or easy adjustment. If you get this wrong or your filament diameter is different, it either will not grip or be too tight and motor force is lost in the teeth deforming the filament and you skips steps when nozzle pressure increases. The feeder has one job in life and that is to 100% of the time move the exact linear distance of incoming filament as commanded. Slipping and grinding or skipping steps is the exact opposite of precision. Again, the feeder has one job in life, to take the commands sent to the motor and translate that into a distance of filament pushed into the nozzle regardless of what filament you have loaded and should only skip steps in the case of an absolute nozzle jam of epic and proportions. It must never grind or slip and should not require adjustments- pretty much the opposite of the stock feeder.
#3 Technically, the extruder head breakout should not be connected electrically to the frame or extruder head. The combination of the large aluminum bar or block that is the stock feeder provides an electrical path to the extruder head and through the bearings and the rest of the frame. This is because even though the electronic breakout board is in a plastic 3D printing cover, and the screws may even have a nylon washer, the through hole through the PCB can contact the metal screw, and since the screw is into the aluminum feeder block, and since that is screwed to the central crossing block, and all the bearings are metal and in contact- we have a massive way for static, EMI, shorts from an errant heater cartridge- all ways to shoot down the ribbon cable right into your electronics. The Bondtech feeder being made of plastic and the mounting of the breakout board to the Bondtech feeder- completely isolates the extruder breakout from the frame.

So basically, that was just a high level list if what amounts to cons against the stock feeder:
Let's talk about the pros of the Bondtech:
The original direction drive AKA mini style reuses the existing extruder motor and mounts the second extruder motor (if used) down on the frame. So as of this moment, this is the lightest moving mass version. It also again retains the stock motor, just now bolted to the Bondtech feeder. The downside of this option is again the remote feeder is pushing though a Bowden tube with ever slightly less control over filament pressure because the Bowden tube can stretch while the filament inside slightly compresses during high pressure situations. It's also not as good for flexible filaments in the second Bowden extruder. That said, it is optimized for single left side extrusion and that's because of the moving mass reduction.
The next variant is the newer BMG style. The BMG is the rear reduced extruder. The whole idea is use a smaller and thus lighter stepper motor, trade off the fact the motor can spin faster than we typically use it for and thus get the same if not better resulting torque and feed push on the filament by using a lighter motor. This works amazing well. The other difference though is this is sold as the double extruder. So this double direct drive (meaning the second extruder is not Bowden remote tube pushing style) is only slightly heavier than the single extruder "mini" style. Again, the idea is no more compromise. You get massive weight savings over stock and don't compromise dual extrusion capabilities and puts it on par with the original single Bondtech Mini upgrade that started all of this.
But let's get to the real magic of Bondtech, the grip. The magic is the custom machined hardened hobbed gears that mesh together and grip both sides of the filament together thus pushing the filament straight down into the rest of the extruder. Not to mention, the entire system has a spring pressure system to adjust for filament diameter and soft and hard filament. This means without the user making careful adjustments, the Bondtech spring loaded system will do it's best to properly grip any filament you load into the system. No more taking off the extruder cover, loosening 4 screws, shifting the motor, hoping you got it right before you tighten the screws.
Next, because the Bondtech either uses a smaller gear in the case of the mini style or the gear reduction in the BMG style, you have a higher step per mm of filament movement. That means that finer control and slight increase in force just from the fact the motor rotates further to move the same linear distance of filament.

The question you should be asking after seeing this list, using the system first hand for a while stock, is "Why did I wait this long to really figure out an upgrade"?

Now yes, that's hard and cheesy sales pitch. Absolutely folks make great prints every day on the stock extruder. You could use t for as long as you own the printer and it is a simple system. You can adjust it for various filaments and it does have the nice clear stock cover. Again it's not like the stock feeder is so broken that people cannot print. It's also not that the Bondtech is the only game in town or the only upgrade on can do to the printer. That said, it is proven and used by quite a few of us and backed as one of the few systems made just for this printer.

User avatar
jmp
Posts: 199
Joined: Mon Mar 21, 2016 12:20 pm

Re: Bond Tech Extruders . Any good, whats the improvement?

Postby jmp » Sun Nov 05, 2017 12:00 am

The installation is pretty simple, probably can be done in an hour even if being really careful (there is a video showing step by step installation). I first upgraded to the bondtech that reused the original motor and it worked flawlessly for a year and about 2000 print hours. I just installed the BMG so I don't have long term usage but it seems to work just as well while shaving some weight.
RL Name: Jason Preuss Thingiverse: JMP Youtube:http://youtube.com/c/patterntoprint Website: http://www.patterntoprint.com

User avatar
walshlg
Posts: 1346
Joined: Wed Mar 23, 2016 12:01 pm

Re: Bond Tech Extruders . Any good, whats the improvement?

Postby walshlg » Sun Nov 05, 2017 10:45 pm

20171105_172335.jpg

The heads are 1/2 the weight. I tried setting speed to 200mm/sec and infill to 300mm/sec. (Actually prints about the same speed as 100mm/sec due to acceleration limits, I think).

Did a torture test, 10cmx10cmx1cm 1 shell and layer box. Minimal ringing, fill had a problem but WOW capable of at least 2x the printing speed.

Spikeysonic
Posts: 98
Joined: Wed Mar 01, 2017 12:07 am

Re: Bond Tech Extruders . Any good, whats the improvement?

Postby Spikeysonic » Mon Nov 06, 2017 12:49 am

Polymaker PC Plus & Support Attempt 1.pdf
(4.34 MiB) Downloaded 26 times
Polymaker PC Plus & Support Attempt 1.pdf
(4.34 MiB) Downloaded 29 times
Polymaker PC Plus & Support Attempt 1.pdf
(4.34 MiB) Downloaded 29 times



How many types of Bondtec upgrades are there? Several appear to have been mentioned.

Think \I got most of that above though a summary would help.

I have the Raise 3d N2 Dual. I got it at the TCT exhibition 26 Sept 2016. (have there been changes in the models sold since then which may also explain problems Ive had.)

If you look through my posts you will see I have had pretty much no success with anything other than PLA since december.

So far what you have been saying looks promising.

What Ive tried has included.

PVA supporting PLA, The PVA just jamed and blocked the nozzle.

Nylon and Nylon using PVA as a support i the second extruder.This just jammed the Hot end which broke during attempts to clear it. Plus shrinkage pulled it off the base.

Ninja flex on own and then trying to use Polymaker Poly support in the second extruder. This just resulted in a strandy base where none of the strands were boneded and all like bobbly material on top. Or resultsed in an open latice resembling steel wool or those scale collectors you used to have in kettles.

Tried polymaker poly PC and Poly support... just ended up with shredded prints, which also stik to the hot end coatign in in burned molten plastic blobs.

Attempts to try and use ninjaflax with a 0.8 nozzle did not go further and ended up wity skipping gears or ninja flex wrapping around gears

The ninja flax also clogged up the hot end and attpts to clear it ended up breaking another thrao on a hot end.

hipps just resulted in too much mateial comming out and the nozzles sliding through the material few mm deep .

Often seem to have problems with dual material prints and poly support was not much better nor were attmpts with glaze and some other flexibles.

So double checking how would the Bondtec work with all of these in particular duel extrustion with one material as a suport such as PVA, Hipps or Poly support, and particular interested in materials like ninja flex, stopping blaocked, slipping ad does this stop material curing around and ending up
all over the nozzle etc?

15 Oct PLA Attempts to make a PLA Moai Statue .pdf
(337.55 KiB) Downloaded 14 times


I found bonteck pots that appeared to cover development of the bond tech 35 pages long whihc were mentionign all sorts of custom made parts using SLS etc.


Is there an all done kit?
Attachments
Raise 3d Materals Problems.pdf
(651.77 KiB) Downloaded 20 times

zemlin
Posts: 281
Joined: Sat Oct 21, 2017 2:02 pm

Re: Bond Tech Extruders . Any good, whats the improvement?

Postby zemlin » Wed Nov 08, 2017 12:55 am

Just fwiw ... I recently swapped out my dual for the Bondtech. My upgrade was not just a bolt-on swap. The screw heads on top of the gantry slide interfered with the Bondtech body, so I needed to drill some clearance for the heads into the bottom of the extruder body. Also, the teflon feeder tubes were too long. I had to trim about 1.5mm from the tops of them.

These weren't a big deal, but I started it rather late in the evening expecting a simple bolt on. I ended up getting to bed pretty late for an old guy.

and then there's the bloody RTV on the servo motor connectors ... sorry RAISE, this is NOT a good idea. Please stop.

IMG_0480.jpg


IMG_0479.jpg
Last edited by zemlin on Wed Nov 08, 2017 5:21 pm, edited 1 time in total.

EldRick
Posts: 243
Joined: Thu Jun 22, 2017 5:55 am

Re: Bond Tech Extruders . Any good, whats the improvement?

Postby EldRick » Wed Nov 08, 2017 5:10 pm

Mine was a perfect fit - a 15-minute swap.

Jetguy
Posts: 2207
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bond Tech Extruders . Any good, whats the improvement?

Postby Jetguy » Wed Nov 08, 2017 5:22 pm

On the RTV thing, I cannot stress enough, if you took the ribbon cable off the breakout board, you must, must, must either ensure you lock it back in with some mechanical or glue form. This is in addition to the HIGHLY recommened mainboard mod. Failure to do this- I could have damaged my own printer just the other day had I not done the additional safety mod. This is just as bad on a stock printer- but knowing that you just unplugged and moved some of the wires- I cannot tell you enough how important and how bad the problem is if that cable comes unplugged.

Vice Chief
Posts: 52
Joined: Wed Aug 23, 2017 5:04 am
Contact:

Re: Bond Tech Extruders . Any good, whats the improvement?

Postby Vice Chief » Wed Nov 08, 2017 9:04 pm

Jetguy, can you point me to the info about what happens if it comes unplugged? I glued mine in place with some hot-melt, but I'd like to learn.

zemlin
Posts: 281
Joined: Sat Oct 21, 2017 2:02 pm

Re: Bond Tech Extruders . Any good, whats the improvement?

Postby zemlin » Wed Nov 08, 2017 9:30 pm

I did not disconnect the breakout board - only the motors. I am aware of the thermal control issue and have noted the resistor mod to prevent it.

Jetguy
Posts: 2207
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bond Tech Extruders . Any good, whats the improvement?

Postby Jetguy » Wed Nov 08, 2017 10:15 pm

can you point me to the info about what happens if it comes unplugged? I glued mine in place with some hot-melt, but I'd like to learn.


The issue is, that I incorrectly assumed that the Raise 3D hardware mainboard and extruder breakout (not even talking about firmware) was built such that all failure states (cables and connections coming loose) was detected. What led me into this false sense was that if the ribbon cable is unplugged or partially connected in such a way the thermocouple analog voltage signal pins going back to the mainboard are unplugged, the system boots up and instantly detects the fault. It's all about context and situational awareness. I knew the AD597 thermocouple amplifiers on the extruer head breakout also have safety built in to detect an open circuit fault of the actual thermocouple itself to the screw terminals. So end to end, you have a few "layers" of safety here. Or think of it as a chain. The thermocouple has to work for the AD597 to not show an error, and then I mistakenly thought that the AD597 talking to the mainboard connection would detect a fault under all logical failures. This is where this was later found to be untrue. The issue is, the mainboard and really the mega2560 processor itself uses the ADC (Analog to Digital Converter) inside the mega2560 to read the analog voltage coming from the AD597 and thus the thermocouple. In and of itself, nothing wrong with any piece of that. Many other brands of printers and machines used this exact combination of parts. But it's the devil in the details. The fact the AD597 is on a separate board connected by cabling- and that that cabling is lacking locking connectors. If the AD597 was directly on the mainboard, then no cabling or connectors to come loose. The tradeoff on that is, then the thermocouple wiring itseld has to extend all the way to the mainboard. You cannot just terminate a thermocouple into copper wiring and get an accurate reading.
Anyway, probably too far in depth here but the issue is, the wiring for the heater power has a locking connector, the temperature sensing ribbon cable does not have a locking connector. In failure, what happens is the mainboard is totally unaware that the current temperature it is reading is not valid because the ribbon cable came unplugged. Again, the fault is, the temp or voltage that the actual processor pin is seeing stays the same when the cable comes unplugged. Without a change, there is no way code (firmware) can detect this state of error. If your PID is properly tuned, your system is printing and the last reported temp is perfectly normal but say 2C under the set temp, the heater will be in the on state because you are trying to reach the setpoint. If the cable comes unplugged in this state, the last temp keeps being read as if it's being updated even though the cable is physically not plugged in. As such, the heater cannot be controlled because the logic is being fed bogus information. This is why this requires a hardware fix- a set of resistors that change the voltage on the pin should the cable come unplugged. That way, the very hardware system only works if the cable is working as a system.

Again, here is the recommendation:
#1 secure the cable so it cannot come unplugged.
However, please note, without the resistor mainboard mod- a failure of the cable still can happen. Again, if your glue or cable tie or the cable itself fails still = problem. You reduced the risk but did not eliminate it.
#2 When you apply the resistor mod to the mainboard, now the very mainboard itself is capable of detecting a connection fault to the head breakout and "fails safe" in that a high temp is read when the cable is unplugged or broken. Again, we are applying a failsafe mod to the very mainboard itself since this is what controls the heater. When used as a pair (both a mechanical fixing of the cable and connectors, and the resistor mod) you have layers of security. You should not have a cable failure but if that still happens, now, beyond a shadow of doubt, the mainboard can detect this failure.

Jetguy
Posts: 2207
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bond Tech Extruders . Any good, whats the improvement?

Postby Jetguy » Wed Nov 08, 2017 10:27 pm

The reason why this is such a big deal again is the fact that I think some problems people have mid print with a failure and a jammed extruder- they might actually be undetected ribbon cable connection failures. And the reason again is bad testing and thinking about how the system works. If the cable is unplugged before the printer is turned on- failure is detected because a valid temp sense voltage was never read even one time. System starts up with a big fat MINTEMP error. The issue is, you can be printing or the printer just on, and the cable gets bumped loose- this is not resulting in an error message. To the user, the firmware, and everything else, hey, the temperature looks fine. You cannot control a heater if you don't know what the real temp is, that's the whole point of a heater control- feedback. So a user has a failed print where it jammed mid print. They turn off the machine to fix the error, it boots up and now has a mintemp error. In their head, this mintemp error just happened. That may not be true, you may have been trying to print and because the printer was on- and had read a valid temp. Or the other scenario I see repeatedly is a user gets a mintemp error. They look at some stuff, they turn the printer off and on 2-3 times and one time it fires up without error. They think it's some bug or software thing and try to print. Literally, since they did not diagnose a root cause and positively fix and understand the error- this false sense all is OK to print, I booted and didn't get an error this time is REALLY BAD. They got a warning message, ignored it and when the message went away with no real correction- all must be OK is far, far, far, from OK.

Jetguy
Posts: 2207
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bond Tech Extruders . Any good, whats the improvement?

Postby Jetguy » Wed Nov 08, 2017 10:32 pm

Also, the whole reason this is in the Bondtech thread but really applies to every single owner and user is the simple fact when you install the Bondtech upgrade, you are moving the head breakout and mounting it to the new extruder and disturbing the factory wiring. So again, sadly, this shouldn't "just" be a Bondtech only discussion, it's for any and all printers if not just to be aware of it as something that "could happen".

Jetguy
Posts: 2207
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bond Tech Extruders . Any good, whats the improvement?

Postby Jetguy » Wed Nov 08, 2017 10:35 pm

Just reposting the link to the resistor modification details viewtopic.php?f=4&t=3524&p=19044&hilit=resistor#p18989

Vice Chief
Posts: 52
Joined: Wed Aug 23, 2017 5:04 am
Contact:

Re: Bond Tech Extruders . Any good, whats the improvement?

Postby Vice Chief » Thu Nov 09, 2017 2:44 pm

Got it. Thanks!

User avatar
woofy
Posts: 95
Joined: Sun Jun 12, 2016 4:09 pm

Re: Bond Tech Extruders . Any good, whats the improvement?

Postby woofy » Sat Nov 11, 2017 7:19 pm

Vice Chief wrote:can you point me to the info about what happens if it comes unplugged? I glued mine in place with some hot-melt, but I'd like to learn.

Yes.
viewtopic.php?f=5&t=1470#p15549
Raise3D N2+, Ultimaker2 & UP plus printers.

spid
Posts: 15
Joined: Mon Apr 25, 2016 7:40 pm

Re: Bond Tech Extruders . Any good, whats the improvement?

Postby spid » Tue Dec 05, 2017 8:10 pm

The lastest is Direct dual drop in replacement from bondtech

http://shop.bondtech.se/ec/upgrade-kits/raise3d/raise3d-dualdirect-upgrade.html

It works with every filament, it never skips.

Bondtech is popular amongst serious makers for all their machines.


Return to “General”

Who is online

Users browsing this forum: Bing [Bot] and 4 guests