Kraken thread

[macanazon]

I have upgraded my Rostock max v2 to use the Azteeg x3 pro with the viki lcd and Kraken but I don't now how to obtain the correct firmware for these upgrades. Is there any way any one can send me the files that I need for this?

[pyrophreek]

I have done a similar upgrade, the easiest way is too use the repetier firmware configuration tool online.

[Renha]

answered you in another topic

[626Pilot]

I believe what I am seeing is a result of a slight twist in one or more towers, or difference in arm lengths on one or more arms.

You probably have tower lean. It's common on these printers. The towers are not fully constrained upstairs or downstairs, so they can have a tendency to lean a little. If you remove everything from the effector, does it sit perfectly flat on the glass? Probably not... I learned this when I tried using the E3D Chimera hot end. It was just fine at bed center, but halfway to the edge one nozzle would be higher than the other.

Re-doing your towers with something like this might help. I was able to get my lean down around 1 degree.

Another thing is the top plate. It can deform slightly from having the weight of the filament up there.

I just got my Metal MAX frame from Trick Laser, so hopefully I can be done with that problem once and for all.

[pyrophreek]

It actually was the slight slop in the rods. The added weight, and the lower position of the nozzles on the kraken, resulted in even tiny rotations of the end effector platform causing substantial relative changes in height of the nozzles. Replacing the stock arms with the trick laser carbon fiber arms solved the problem. Any rotation of the end effector platform is due to slop in the linkages, or the carriages not traveling on a straight path. The parallel arms constrain the system to translational movements only.

[euanfoster]

Hey there,

My work is currently in the process of upgrading a rostock 3D printer to fit a kraken with an azteeg x3 pro. However, our power supply unit was cobbled together from an old PC and is fine for single filament prints. As a result, we were wondering what sort of currents you guys were getting when running multi material prints? and if possible could someone recommend a power supply unit to buy?

Kind regards,

Euan

[pyrophreek]

One of the nice things about the azteeg is it supports separate power supplies for different parts.
I have actually upgraded ours with two separate power supplies, the stock atx and a 24v one for the bed. The bed draws a ton of power on its own, so having it on its own power supply has taken the load off the main power supply. Also, the bed actually heats up in a reasonable amount of time.

[euanfoster]

Ah that's a pretty smart idea separating the power supplies like that! I think I will just ahead and copy you on that one.

Thanks for the advice!

[euanfoster]

Hello everyone,

I now have my kraken working with an azteeg x3 pro board and two power supplies. It is a major improvement on the rostock machine we had before hand. I was wondering however if anyone has heard of a direct drive modification? I would like to print one nozzle with Poly Urethane and that specific material doesn't like being bowden fed. I have CADed up my own idea based off the airtripper extruder, but was wondering if a tried and tested mod already existed.

Looking forward to hearing what you guys think.

Euan

[pyrophreek]

The main problem with mounting a regular nema stepper motor to the end effector is the weight. I see two solutions:
1) use a small geared stepper (cheap unipolar ones can be modified to be used as bipolar).
2) use some sort of flexible linkage to connect a stepper to an extruder on the head.

[Ascensi]

I was thinking of a direct drive system as well and what I concluded might have to happen - if I were to do this is to mount either surgical rubber hose or a spring inside the cable management hose with just enough tension to lift the weight upward to the Rostock ceiling (perhaps just to lift the cabling rather than the endeffector). I have a magnetic end effector setup, I've had it tested on 2x speed multiplier in Matter Control (using Rambo 1.3V3) and it didn't come off but I had only a single hotend and fan shroud at that time.

I'm working on having Kraken set up with 4 E3D Volcano's on the magball endeffector and I would like to quickly swap another Kraken endeffector system with small nozzles for detailed prints etc.

I've been modeling a new endeffector for the Kraken to use with magballs and 40mm fans should ideally just snap into it rather than needing to print them -less weight & material.

[pyrophreek]

That reminds me of a delta printer I saw a while ago (http://deltaprinter.co.za/glass/index.html" onclick="window.open(this.href);return false;) that uses a pulley and counterweight system to hang the extruder motor above the end effector. You could try to make something like this, and adjust the counterweights to compensate for the added weight. You will lose acceleration, and it is still technically a bowden drive system, but the drastically reduced tube lengths should make it act more like a direct drive.

As I mentioned before, the cheap 28BYJ-48 geared unipolar steppers can be run as bipolars, and should work at 12V just fine (http://www.jangeox.be/2013/10/change-un ... polar.html" onclick="window.open(this.href);return false;). These are much lighter than the normal NEMA format, and are slower due to the gearing, but should work for polyurethane, as you cant go very fast anyways. Worth a try mostly because you can pick these up for a couple bucks on ebay. I don't know how long they will last, as the gears may wear out, but considering the low cost it may be worth trying.

[Ascensi]

Hey cool! I am working on an effector but plan to mill it in aluminum.

Cheers,
Michael

Did you manage to create/mill your Aluminum effector? I have a design/Stl that reclaims the full Z space using magnetic arms with space underneath for the Halo LED light and a fan shroud built right into the effector plate that aims the airflow directly between Kraken's 4 heater blocks. I have it setup on my Rostock and secured it with mini bungie cables "CamJam" between the delta arm's ball joints. I printed mine but I had to carve it in places with a deluxe crafting/wood burning exacto knife just to make sure parts like screws on the heater blocks and wiring could fit through and not burn the effector -basically I'd have to update the model if other people wanted to use it. I'd love to have an Aluminum version but it might have to be molded and cast. I don't have a milling machine so I'm looking for someone whom could collaborate.

[mhackney]

No, I'm using the stock V1 effector and my printed Kraken Konnector.

[Ascensi]

Here are a couple examples of what I'm working on to get the Kraken working in a smaller, compact effector with common features such as fan shroud and LED built in. Bungie cables have been added, these ones in particular can be latched on easily and adjusted on the fly. Please excuse the poor image quality, just took them quickly. I still have to rework the files as I explained earlier. This effector allows the original/stock z height. I haven't yet tested this effector, I am hoping I could get additional setup info.

Can anyone share what the offset values are for the Kraken? thanks.

[enggmaug]

Here are a couple examples of what I'm working on to get the Kraken working in a smaller, compact effector with common features such as fan shroud and LED built in. Bungie cables have been added, these ones in particular can be latched on easily and adjusted on the fly. Please excuse the poor image quality, just took them quickly. I still have to rework the files as I explained earlier. This effector allows the original/stock z height. I haven't yet tested this effector, I am hoping I could get additional setup info.

Can anyone share what the offset values are for the Kraken? thanks.

Interresting design.

It would not fit on my kraken, because it seems that my heat blocks are much bigger than yours.
Did you do anything to them, or did E3D change their heat blocks for the kraken ?

But how do you make sure your heat blocks dont melt your effector ? What material do you make your effector in ?

What offset values do you need ?

[626Pilot]

Is anyone here actively using a Kraken, or has everyone abandoned it to the Cyclops and other hot ends?

Any 3- or 4-color prints that don't look like a bunch of trash (which is literally all I could find online)?

[mhackney]

I'll consolidate the reply from the other thread here too...

I think the situation is that no one extruder is going to be good at everything and that there will be specialization. We are starting to see the early days of that now. For instance, I think a really good switching extruder similar to the Cyclops for PLA or ABS color work is probably not going to be able to do its color job well while also being able to handle high temp materials (Nylon, etc) or greatly dissimilar support materials. Similarly, a really good all metal extruder for doing work with exotics like CF PLA, woodfiber, Nylon, polycarbonate, etc probably is not going to handle PLA very well. These materials have different physical properties of both the solid and melt. A simple example, a short nozzle bore really helps when printing PLA but a longer bore is much better for materials like Nylon and polycarbonate which tend to drool with a short bore. Also, as you pointed out, the melt path needs to be "hardened" for materials like CF PLA but simple PTFE tubing is fantastic for ABS or PLA only work. The choice is going to be very personal - what do you need to be able to do and what tradeoffs are you willing to make?

I've had the best dual color printing experience with the Cyclops over all others by a big margin. Drooling and stringing have not been an issue. I worked hard to test and adjust to get the lowest extrusion temps - which translates to less drool and stringing - and that revealed a weakness in the extruders I use. That resulted in investigating a more forceful extrude. That led to the BondTech QR. The combination of the QR and Cyclops gives me really excellent results. I think you've seen some of the 2 color prints I've done.

I no longer use the Kraken for multi color work. I think if I went back now, with what I know about "calibration" and getting a well calibrated machine, some of the dragging issues would be alleviated. I might try this over the next few weeks.

[626Pilot]

I have become convinced that to get the effector REALLY tram with the bed, with arms that last for years without losing any tolerance (rather than weeks to months), you have to use either magnetic or tension ball joints (steel balls held in place with Spectra line/spring). I have some steel balls ordered from Haydn (deltabot forum) and will see about getting them adapted to the Trick Laser platform and carriages. If that doesn't work out, I think the answer is to CNC machine an acetal effector with the standard SeeMeCNC hole pattern. If acetal is good enough for the wheels, I think it will be good enough for the joints as well.

I'm also working on a counterweight solution for flying extruder platforms based on Spectra line, V wheels, and 500-gram weights. I'm concerned that any differential weight put on the platform (from extruder bowdens/electrical cables/water tubing pressing at different angles) will change the tension on the GT2 belts. My idea is to suspend the platform from the carriages, and then counterweight the carriages so that they're mostly "weightless." The platforms will eventually have a strain relief for the huge bundle of wires going down to the hot end, so it won't press as hard either. Eventually, I'd like to look into replacing the GT2 belt with either highly-tensioned Spectra line, or some type of chain that obviously won't flex at all. I don't know how feasible chain is, but I've seen others mention the idea.

[bot]

Have you thought about nema 11 motors at all? I have a couple geared 5:1 on the way from china to see if they will work out. One seems to be less than 200g. I've seen others use larger, close to 400g nema 11 steppers with aluminum mechanisms. The micron3dp also uses nema 11, although geared to 14:1 I think. I'm putting it right on the effector. In my mind, the added mass to the effector might actually be beneficial in decreasing any harmonic frequencies of the effector, possibly reducing ringing (or at least lowering the frequency, and therefore making the ringing less prnounced).

[Xenocrates]

I almost think that a swashplate and sprung hotends is a good answer. Only need a single motor, no matter how many hotends you want to control, can be made relatively simply as well as relatively lightweight and compact. There have to be issues with this, but I might try it latter. Maybe a water cooled set of E3D's, designed to have a shelf for a spring that goes around them, and put three of them on a heavily modified aluminum carriage, with a small stepper, and a swashplate. That seems like a good test platform, although rather obviously you end up having trouble with scaling upwards in terms of heads compared to a kraken like solution, as each needs it's own water cooling jacket and springs, and they need to be arraigned circularly. That being said, perhaps a sprung hotend and a solenoid is better, as then your arraignment is more flexible, but then you end up needing more electrical gear and wiring to the effector.

Firmware or slicer changes for the swashplate would be interesting, as the slicer would have to track where the swashplate is versus where it needs to be for each tool, and then rotate it to position. 626, since you've done a lot of firmware stuff (for the smoothie, admittedly), how hard do you think it would be? Is it possible to do it without homeing the plate? Or would it be easier to move it to a known reference position after each tool switch, and then program the correct rotation for other hotends? I thing being able to not home it would be far superior, but that may prove impractical to implement in a reasonable amount of time for the available firmware.

Also, in regards to a counterbalance system 626, have you thought at all about those retractable key lanyards? It seems something like that would be easier and more compact to implement (although you would be somewhat limited in granularity), and reduces your part count, as well as the number of moving things at any given point, hopefully reducing waste energy in the system.

Chain doesn't seem like a good idea. It's much heavier, much bulkier, and has more in the way of stiction in most assemblies. Fishing line doesn't seem like an ideal solution either, as it is designed to slip rather than catch, and thus would be a bit of a bitch to get to move any significant weight, and at high tensions I worry it might deform things like idlers or carriages. GT2 seems like the best choice for now. It's not so finicky as line, nor nearly so bulky as chain, is strong, accurate, and standardized.

[626Pilot]

I currently have three steppers weighing about 500 grams each (NEMA 17, geared 5.18:1, short can) on a printed platform above the extruder. They are tied to the carriages with paracord, and they definitely keep the belts tensioned! The tricky part is getting the height of the stepper platform just right so that it doesn't push/pull the effector down/up via the Bowden tubes.

The geared motors are nice. You can run them at much lower current than a standard NEMA 17. I have mine at 1 amp, and I think they could run with a little bit less. (At the standard 1.5, they were quite hot to the touch.) I would say the smaller steppers would need to be significantly lighter to incur much additional benefit. 200g sounds about right.

The trouble with supporting extruders, whether from the carriages or the effector, is that you can never just turn the machine off. If it's homed and you turn it off, it will plow the hot end nozzle right into the glass at top speed. It's fast enough to illuminate LEDs on the controller! I'm working on a counterweight system. Spectra line is tied to the carriages on one end, and 500g weights on the other. The lines are routed with V-groove pulleys from Amazon. Initial tests (without the pulleys) were good. The effector wouldn't fall regardless of what XY position I put it into. Werner Berry uses a similar system on his printers, but he takes the weight off the effector instead of the carriages.

Key reels do sound like an interesting idea, and it would be nice not to have a bunch of weights hanging off the side of the printer. Would need to find some with about 500g of retraction force. They could be mounted right to the tower tops, no V-groove bearings required.

The swashplate idea is interesting. I assume you mean something like the thing on a microscope that swivels and lets you select which lens magnification to use. It would require mounting a servo or tiny stepper on the hot end. In theory, it's doable if the actuator is repeatable enough. I don't know how repeatable hobby servos are.

[blackrabt]

I had to really stare for a while to determine what all was going on with your printer. What a beast.

Maybe I am off base here but I think even with just the right balance of weight and retraction force you may find that you are putting more work on the extruders still to move that weight around. Probably negligible though. Is there a reason you wouldn't want to mount the extruders to the frame and just run multiple bowden tubes to the hot ends?

As far as retraction force, those key hangers are usually just hairsprings (similar to those in a pocketwatch) inside a housing, if you could do the math to figure out the thickness and number of coils needed you could probably print the housings yourself and add mount points to them for ease of use, then order springs from a specialty store like http://www.acxesspring.com/spring-force ... lator.html" onclick="window.open(this.href);return false;.

[Krinje]

Pardoning my summery knowledge of the problems being approached here, I'd like to add my $.02. Along the swash plate idea, perhaps the servo could drive some sort of transmission that selected the appropriate extruder? Or perhaps a servo for each extruder (the servos could be quite small) that engage/disengage the idling wheel in the extruder, effectively the geared wheels spin on all extruders but only grips when a servo engages the idle. (or more complicatedly the other way around) I'm sure some mechanical advantage could be found to use very small servos for this task. Just a couple specific examples: 6.2 gram servo can output 1.2kg/cm, or a 30kg/cm in a 66g servo. Anywhere from 5 to 40+$

[626Pilot]

Just got done installing my Chimera. I'd like to switch to the Kraken soon, but I'm not going to bother unless and until I figure out this mess:

Has anyone EVER got a 2-color print out of a Chimera, or a 2+ color print out of a Kraken, that didn't look like unmitigated trash? I said it before, so apologies for repeating myself, but I have NEVER seen a photo of a 2+ color Kraken print that didn't look like it belonged in the sewer. I'm starting to wonder whether anyone has ever been able to print something decent-looking with one of these, other than only using one nozzle at a time.