Dual Bowden Extruder for Robo

With the help of Tessarect, we've been working on dual extruder designs for the Robo 3d. I think it's time I make a thread about it to discuss it.



It's a fairly basic and contained design. The E3D nozzles mount to a plate with a V saddle insert. This secures the heat sink extremely well, requiring a large amount of force to twist slightly. The mounting plate then connects to the carriage via a screw acting as a post. This is them able to be adjusted height wise to calibrate the nozzle heights, with a spring providing resistance.



I've got most of the parts ready except for half the cold end. I have some concerns that I'd like feedback on if possible.

1) The use of 2 leveling inputs gives me concerns about the nozzle being slightly off angle. I'm thinking using a caliper to mesure the height off the plate would allow for verifying the plate is level, but still leaves margin for error. A single calibration point per side would be ideal but haven't been able to come up with one.

2) Would the stock E3D fan provide enough airflow to cool 2 heatsinks? My instinct tells me the fan is very oversized for the the heat transfer load, but not knowing the specs of the fan I can't verify for sure.

3) Is there any danger of blowing cool air on the nozzle? I'm trying to aim my part fan away from it but it's been difficult to completely miss it.



4) Any ideas for a simpler belt tensioner?




Any other critiques or criticisms are appreciated. Hoping to have it finished by next month.

 

1. If the nozzle's J-head is already flat it shouldn't be a problem. But if it isn't it could cause it to not be flat. Have you thought about a three axis leveling system? I don't know if there's room in the design for it.

2. I agree with Mark. Can you rotate the fans 90 degrees and get both in there? Another option is a new hood with piping to the heat sinks and use a different external fan to bring cool air to the sinks.

3. Only danger I could see with blowing air on the nozzle is if it cools it enough compared to the heater block at it seizes the filament inside the nozzle.

4. I can't really tell how the belt tensioner should work from the drawing (unless it's just like the current Robo one).

 

Mike, have you out any thought of using magnets to hold the hot-ends plates against the main carriage? It would make swapping even easier.

I've been thinking if power also. Maybe use color coded Anderson Power Poles. They attach and detach very easily so you don't need much force.


Sent from my iPhone using Tapatalk

 

Anderson Power Poles seem pretty heavy duty!
I'm using 4 pin mini-fit-jr plugs (ATX 12V plug) for my head(s) -- 2 thermistor (overkill), 2 heater -- good to 9A (108W -- more if you use the HRC contacts) -- was hoping for smaller, but....

I was thinking JST connectors, but those don't lock (zip ties FTW again) and I'd have to keep the thermistors and power organized...

1) If you wanted 3 axis adjustment, you could have 3 screws for each extruder plate (rotates in 2 planes and can go up/down in z -- but not straightforward as 2 screws) -- I like having a separate plate that actually holds the heads using magnets -- easy to fix jams -- and you don't have to re-zero it afterwards (not sure about the forces though -- and you don't have to re-zero if you take it out via the 2 bolts for the J-Head mount) . I'd flip the height bolts so the heads are topside and make hex captures on the bottom plate to make it easier to adjust. I think you're stuck with calibration via printing because the hotend lengths may differ.

2) I think 1 fan for 2 E3D's is good, but I don't have my E3D yet...

3) As far as excessive fan cooling of a nozzle goes, if that is a problem, you can do what I did (ugly but works category):
Put a layer of Ultra-Copper permatex on the nozzles -- this insulates them quite a bit and is removable if necessary (keep the last 1/8th inch bare so you don't. touch your model!).

4) I like the zip-tie method -- it seems to work and is simple enough.

And OMG:
I know it's horribly complicated to flip everything, but I think having 2 extruders on either side of the rear rod might be better (though it looks like you fit 2 extruders into the same space as one, so you might not gain as much as I think by doing so)

• Positions center of mass under the positioning belt -- no torque created for a move (though, compared to stock... not much mass there so...)
• I have lots of overkill (compared to my heated bed area) in the Y direction)
• Even if not, easier to upgrade Y size (replace rods/glass and call it even vs stretching Robo3D body)

 

But... if you want to remove the head using the J-Head securing bolts, you need a slot or something between the extruders (plus connector sized hole somewhere ) or you need to run the wires on the outside (nuts for the neat package).

 

And #2 -- I'd designate the nozzle nearest the fan as "cold" and the other one as "hot" so you don't risk melting at problematic places due to hot air heating the "colder" nozzle.

 

Thor, it would really help if you included pictures with those explanations.


Sent from my iPhone using Tapatalk

 

1) What would you have in mind? I finally got some prints assembled last night and I noticed that once securely tightened down leveling of the nozzle/heatsink is a non-issue. I'm more concerned now because once the jhead plate comes a little bit loose, ie to raise it, it induces a sway along the x axis. I'm thinking of going to a 4 point mounting plate but that just seems excessive. I'm hoping the deviation between the nozzle heights is small enough I can get by without that though.

2) It's possible to rotate them I think, I was trying to avoid it though because that area is already very cramped. Sanjay Mortimer responded to me on google+ confirming that one fan is adequate for cooling both heatsinks in my design.

The major advantage of rotating them is I think I'd lose less x movement. I'll look into it though

4)It's pretty close to the existing one but instead of using zip ties for tensioning it has a screw hole you use to press down on the belt. Someone suggested one from the Mendel90 I might incorporate if I end up printing a new carriage plate.

I considered using magnets in the design but went away. It doesn't really offer a way to calibrate nozzle height, plus they're more expensive than nuts/bolts. I like the idea of it but just couldn't ever wrap my mind around a solution.

I'd need to increase the hole size on the carriage plate to 25+mm to get the heatsink out from the top. I might do it because it was a bear of a problem getting the heatsink on the jhead holder plate.

I like the idea of a connector but honestly a DB-12 or DB-19 would probably be adequate. The nozzles only draw like 3.3A so you could get by with at most 2 connectors per nozzle (4 leads total) so that's 8 for nozzles, 2 for fans, 4 for thermistors.

1) I could use a drawing on this. If you look at my response to melody above you can see what I'm currently struggling with. on the design

3) That's a good idea . If I run into issues I'll add that. I got some confimation from sanjay it shouldn't be an issue.

4) Yeah I was trying to think of something more robust.

That's not a bad idea, but yeah as you described I'm essentially not losing any space. My design is about 10-12mm wider than the stock, but only to strengthen the carriage. It's definitely possible to fit 2 on the existing plate if you leave the sides more open and enforce the middle more. I think another issue is the firmware just has a "distance between nozzles" option, and not a direction. Might be wrong though


I"m not sure the fan blows that hot of air. I will be curious to see what happens if I don't decide to rotate the fans.

 

What or where is it decided concerning dual head extruder to use one or the other extruder?

And what about colour choosing? Does the software handle that?

I heard that obj objects contain the colour information and that is translated. But stl files don't seems to have that option.

But extruding filament, dually, one side or the other, what decides?


Sent from my iPhone using Tapatalk

 

Well first thing if you need to specify in the firmware you're running dual nozzles. Then in repetier or matter control (I assume MC has it, haven't used that) you define the printer properties to include a second nozzle. This should then give you the option to specify which nozzle you want which body to be for slicing. I'm not sure if you can define a single stl file with multiple extruders. I believe they need to be unique bodies. Though that side is something I haven't explored thoroughly.

 

The reason I was thinking magnets is that if they're done right, they make lining up the extruders very easy and reproducible. They also can keep you from stripping screws if you have to swap things out. This makes swapping nozzles, for example easy as instead of swapping just the nozzle you'd do the whole hot-end.

The reason I suggested the Anderson Powerpoles is that they allow very easy connection/disconnection but allow for positive connects. They also have very good vibration rejection.

I get really frustrated when you have very small connectors (such as molex) that need you to use a ton of force to disconnect. I have diminished hand strength so this makes it difficult to do without using tools -- yet using tools causes you to deform the connector which is not a good thing.

APP, because of their modular nature, allows you to make the plugs in whatever combination you need. They also come in multiple colors so you can color coordinate with the intended purpose.

But ultimately I think it'll be up to the person who implements the new carriage.

 

If you're trying to do a 2 color print (or support+model), your active area is reduced by 2x the distance between nozzles (A can't get all the way to B at one limit, nor can B get all the way to A at the opposite limit). Slicer has the "offset to 2nd nozzle" in X/Y coordinates. If you rotate it around, the "where A can't get to B" can fall outside the heated area.

'Tis a horrible amount of work though.

Here's a sketch for the 3 bolt idea, but after seeing "offset to 2nd nozzle", if Z is the same, it really doesn't matter where they are XY wise (ugly quick sketch). This does enable you to counter non-perpendiclarity of the nozzles, though I'm not exactly sure how you'd check for such...

 

Yup. 1 STL for each nozzle -- there is a way to save the combined work in Repetier. Clunky compared to a Dimension...
I'd also like Slic3r to be able to automatically adjust the layer height (if you're doing a vertical wall go up to 0.3mm, and go down to 0.1mm), but I don't grok Perl yet...

 

Well you're gonna lose distance on either the x or the y. I'll look into it some more to see about the possibility of having them along y instead of x. I guess it would be better to lose distance on Y (254) than on x (219 for me)

BTW did a acetone bath for the fan duct:

 

It's possible to gain more y distance by going with a different rail system like on the 1.2 robos

 

It would have to wait a few months I'm in trouble already for spending so much money on my printer lol

 

Some progress update pictures. Waiting on some m3-10mm flathead screws to arrive before I can finish the cold end but the last piece is printing now.

X carriage:








New heatsink fan mount:







Dual ezstruder cold end:



Design intent:


Test fit:


with parts:

 

Looking great, Mike! You and Jeff are to be congratulated but I will save my full on YAY for the first filaments spewing forth from both ends!

 

Looks amazing. So Mike are you releasing your own kit for this upgrade?


Sent from my iPhone using Tapatalk