Ah, okay. So the significance of the dimensions specified is just a requirement for the spring to fit, but anything in that range should work?

And yes, I'm not sure what to do with the arm piece, or how it all connects together. A photo of an assembled one close up would be immensely helpful.

Also, I think this might have gotten buried in the sea of comments, but how rigid should the Y bar be? Currently if I twist the bar I can get it to turn a bit. All the wheels are tight on the bar and roll well. I haven't connected the belts yet though, so maybe the tension keeps that in check? Or maybe just a consequence of using PETG since it's so springy?

 

I went to a standard mount for the motors and it seems to have at least compensated for the fact that I can't seem to get the play out of the z-axis gantry. I am going to re-print those parts as I may have screwed them up when I was putting the wheels on them, but for now.......

First Print by Matt Mathias posted Mar 28, 2016 at 10:05 PM

 

There are three general techniques for wire management ranging from easy to hard, depending on how many wires you are trying to hide:

Easy- Run small sets of wires in the Vslot channel and cover with these.

(Cable chains are a nice way to manage wiring between two moving sections)

Medium - Run larger sets of wires in black flexible plastic tubing and attach to the frame.


Hard Mode - Look, no wires! Run small sets of wires through the hollow center of the 20x40 rails.


When it comes to wiring within a control box enclosure, even when you trim wires to length in a box with extra room, it can still be a tight squeeze.

 

Any tips for getting a perfectly squared build *after* the frame is assembled? I have spent the last 2 night trying to do this and seem to always end up with a tiny bit of skew. I'm thinking of just buying a bunch of long corner brackets, loosening all the screws, then putting corner brackets on every corner before tightening it back up again and removing the brackets.

 

I eventually broke down and printed the 90 degree corner pieces for the inside which helped a lot. How are you determining that it isn't square, just curious because I used at least 3 different methods.

A couple of other things to note:

1) Its not that hard for some of the cross pieces to pull things out of square, the front cross piece in particular caused me issues.
2) Squaring up the frame to my satisfaction might have been the part of the build that took the longest.
3) I started with the offset motor mount, I feel like having the cross member on the bottom of the frame like that actually added a lot of rigidity and made it easier to square up a few things. I would use that now even with a different motor mount option. I used the small 90 degree brackets from the z-platform to mount it and it was easy to get perfectly square to the frame, which in my mind helps the rest of the frame be square.
4) Eventually I decided that as long as the XY Gantry rolled square and was within about 1mm of square on each side (by looking at the distance of the wheels from the corner brackets) then I was satisfied.

It also ended up that in order to make things square I had to move one of the pillars just slightly up, I think I must have cut one of them like 1mm too short and having them all flush on the ground was causing some skew in the rest of the frame.

 

My carpenter square always shows at least one corner out of alignment once I've finished.. I probably have a piece slightly short or long somewhere, but I have no idea where. I also might just be measuring poorly -- it's hard to get a good spot for the carpenter square in certain parts of the frame.

My XY bar seems square from the front, but one wheel hits the back of the assembly before the other does (if I roll it forward they both reach the front of the assembly at the same time). I'm going to take it apart tonight and install better spacers, then see if it rolls correctly. The XY assembly is pretty close to square as far as I can tell.

I have the corner brackets installed in the lower assembly, but I couldn't fit all 6 screws.. Maybe this is the issue, as I noticed the part that I don't have connected aligns with the vertical bar.

 

I went with the carptener square as well. Ended up putting little 1-2mm gaps on some of the corners to squares things off. I also put the metal 90 degree joints from openbuilds on all corners which helped out squaring things as well as keeping the required gaps in place and still maintaining strength.

In the end, everything was aligned, level and runs very smoothly. Too smoothly actual since the z platform always drops to the bottom when I power down the drivers... Thinking next time around using a larger lead screw with a smaller pitch to avoid that.

My XY bar was / is off on the back side as well by about .5 mm. But with things squared and running smoothly and no issues, I decide to leave it and attributed it to a printing error when making the rear top brackets.

My point, it took me and my son about 5 days of 2-3 hours sessions to finally get it all aligned and square.

 

Just to chime in: +1 the zero play. When I had both my leadscrews in the back and none in the front, I could definitely tell there was a good amount of weight from the cantilevered bed trying to tip the front of the bed down. Especially based on my build-size, I can easily extrude an entire filament roll, so I just put one on there as a test to see the weight effects on the bed. But I still made sure I had full wheel engagement.
By moving one of the leadscrews to the front of the bed, it helps to support it more evenly, and no more 'front droopage', especially on large prints.

 

I am considering building a C-Bot, and am now looking at the electronics and motors available.
Why are the motors specified as needing to be 68 oz-in or better?
What proportion of the rated current should I run the moters at?
Thanks.

 

the motors are suggested at that much torque so that you have at least enough to move the effector around, and more as a baseline, since youu can find motors in the Nema17 form factor that have as little as 12 oz-in of torque, so it is more of a guide to make sure you have enough torque to move the whole XY Gantry as well as the hotend/effector.

as for the rated current, that depends on both your motors, and the particular stepper drivers you are using in your board

 

Looks like you mounted your solid state relay to plexiglass, one way to help avoid potential overheating in the relay is to mount it directly to one of the Vslot rails, they work well as heatsinks.

 

Ok so after a year long delay, I've finally pulled the trigger and started ordering parts for a C-Bot machine. AK Eric was helping me out last year by answering a lot of my questions about the machine (don't know if he remembers from the Google group) and then I kinda dropped off the face of the 3D-printer planet. So after packing for a move, a new house, bunch of renos, a crappy sled season and a C-Beam Machine build, I think I'm ready for the C-Bot build. Lol

I've been frustrated at times with the limitations of my Kossel. I think what finally got my arse in gear again was seeing the direct extruder file. Which brings me to a couple questions. I assume it uses a standard MK8 gear? Also, has anyone been using a smaller/lighter Nema17 motor now after going direct?

And lastly (for now) where the heck can I get a 1/2 decent dc-dc solid state relay that won't burn my house down? I ordered the E3D VariPower heated bed and I'd like to power it sufficiently. I see a ton of cheap ssr online but have read a lot of horror stories about them burning up.

Cheers guys

 

I don't quite get the obsession with ssr's... Ye olde clickclack relays are inexpensive, and can handle ludicrous current ($30 gets you 75A current for 150,000 cycles with 12v control).

Yeah, it's going to go click any time your heatbed kicks in, but compared to all the other noises, big deal....

 

I just pretend mine is a grandfather clock

 

You know, I never much thought about the life cycle of those things:
Let's say it clicks on\off 6 times a minute = 360 times an hour. For a 12 hour print, that's 4320 clicks. 150k\4320 = 35ish.
So, the relay is good for 35 12-hour prints? That actually seems a bit sketchy if you print just one 12 hour thing a week for a year.

 

Another option is going the custom mosfet circuit route. You don't get the isolation protection a ssr provides and you have you build the circuit correctly, but you get the benefit of using pwm based heating without destroying a relay. The cost is about the same as a relay as well.

 

I use an MK8 gear in my direct drive, however there is a massive amount of variation in MK8 gears available online. I went through 3 or 4 of them before I found one that could actually grip filament properly. Somewhere in this massive thread I posted pictures of some of the gears as well as graphs of of how much force each could generate before slipping. You need to make sure the grooves in the gear are cut deep so the space between them becomes a sharp point.

I switched to a lighter NEMA 17 motor too when going direct drive. This 40mm version rated 2A 1.1ohm and it is somewhat close to those specs in my rough measurement. It weighs about 15% less than a full 48mm kysan. I only run it at 0.5A so the stepper will slip before the filament shreds.

Changing topics, I have been trying to make my own mods to the XY motor mounts and am completely failing to use 123D. What I am trying to do is move the motors 2mm back from the printer (so the motor housings don't make contact with the printer) and similarly reduce the diagonal supporting bits to 3mm thickness by the motor (again so it doesn't make contact) and buff them to 5mm solid parts on the top. The problem is just my general stupidity I suppose, because I sketch the profile I want on the face of the motor mount and then extrude it, however it seems that when I make the sketch, 123D snaps to the nearest grid point and not the corner vertex. That means it looks ok from far away:


But if you zoom in you see the bits aren't actually connected. I've tried to be as precise as possible, but no matter how I try to line them up, I only end up with them as close as the grid for that zoom level is.


I've tried making them as separate pieces then snapping them face to face but that only works for one face and still requires repositioning. How in the heck does one extrude a sub-face that extends from one end of the face to the other? I can do it easily in Sketchup because of how their snapto works, but 123D only has various resolution grid snap?

 

The particular relay I was quoting specs from was essentially a random general purpose relay-- But that was 150k cycles at 75A. I doubt any 3D printer heat bed is pulling anywhere near 75A, so the life should be much, much, much longer. The mechanical endurance is rated for > 1 million cycles. Without much effort, I found another rated for 30 million cycles ("heavy duty" being the keyword).

 

I have a cheap Fotek that I am eventually going to use to control my LEDs since the current will not need to be switched on / off very frequently I am a lot less concerned about it being burned up.

(If you are wondering why I need a relay for my LEDs right now its pretty easy, the 24v LEDs that I have look like the second coming of Christ when I turn them on, where as the 12v ones are very nice. I have the relay and a 12v supply but I would need to buy a 12v step down so meh)

On the other hand....for my Silicon heater that runs off mains (120v 750w) I got an Opto-22 relay and an Opto-22 heat sink. This was not cheap, however I'm also confident that they will not burn up even if I do run PWM with them since they are designed to handled it.

 

I k now that 123D is a lot like Fusion 360 and in Fusion 360 I solve this problem by typing the distance I need it to extrude rather than relying on the snap feature. I am not sure if this is possible in 123D or not because I don't use 123D very often. That being said I can tell you how to get a model from 123d into Fusion 360 if you want.

 

That makes sense! On another build I used a mechanical relay with e flyback diode and it has been in service for about 2.5 years and that printer is my workhorse. It prints on average 3-4 hours a day 7 days a week. Still clicking along happily!

I'm using an SSR on this new build mainly to take advantage of PWM. Mistakenly had PWM enabled when I first tested the mechanical relay. It did not like it very much. LOL.

 

I've got about 10 hours into assembly, and now i'm waiting on the post office to bring me parts. The frame is fully assembled with exception of the gantrys (waiting on bearings for v-wheels and motor couplers for the leadscrews). It was virtually square upon assembly but will need some minor adjustments. When mounting the right XY motor mount i found two bolts from the 2x2 plate below it were pushing against the motor. I removed the washers from the top bolts in that bracket which allowed the motor sit flush as intended. The XY and Z gantry crossbars (drilled on both ends) also had screw clearance issues. To secure the sliders to the extrusion ends i was made to use low profile screws which seem to provide the clearance necessary, i'll be certain when my bearings arrive.

A note about T-Nut counts and the Z-motor mount options. The BOM seems to exclude any additional T-nuts necessary for the alternate Z locations. I used an additional 12 T-nuts just on the bed front piece. Always order extra so you don't pay shipping twice like me.

Electronics on the way include a Smoothieboard 5XC a 30cmX30cm 120v silicone heater and an E3DV6 (which will become dual once the machine is mechanically dialed in.) I've gone for the Fotek DC-AC SSR but i'm having reservations about this mains powered bed idea.

I'll try and remember to snap some pictures when the gantrys are mounted.

 

In honesty from what I have read you *should* be ok as long as you don't use PID and stick with Bang-Bang. If you go to PID you will need a heatsink and some luck. I have heard people having great results and I have heard of relays melting in < 1min.

 

Some videos about SSRs and other options from a source I trust.

3D printing guides: Going big with Solid State Relays!

Guide: Properly picking and using MOSFETs!

3D printing guides: Everything about heated beds!

Guide: Heated bed thermistor mounting done properly!

 

i can tell these are Toms guides, just by the title. and im sad to say you beat me to it. i came here just now to post these very videos!

10/10 would watch 153 times

 

I had somehow missed that Tom did an SSR video! Thanks for the link.

Has anyone placed an inline fuse on the AC bed wires? That would give me a little more peace of mind.

 

I think I found the problem.

I was loosening the belts to test out that the motors were spinning freely. As I loosened on of the belts, I noticed the geared pulley on the motor dropped. When I looked at it, one of the set screws was almost completely out of the hole and the other one was AWOL. Luckily, it had fallen into the motor mount area so it was easily retrieved. I have now reseated the set screws using loctite to make sure they stay in place. I haven't tested the printer yet, but I'm pretty sure this is the issue.

 

Ok so i'm just kind of thinking out loud here and checking if anyone else has had a similar experience. When doing my homework about the C-Bot, I had to make a decision about what material to print the structural parts in. I figured that although PLA prints nice, I wouldn't do as I eventually want to chamber the bot to print ABS. My delta has had problems printing ABS in the past despite a PEI surface. So that left me with PETG. Printed out most of the parts and what a PITA!!! Have to print SOOOOO slow and darn near bonds to the PEI, stringy as hell, likes to glob on the nozzel, but I eventually got it done. I'm surprised actually that people like printing with this stuff so much. Doesn't look the prettiest but it's functional. So when it came time for the direct extruder, I knew that I'd have to print that in ABS regardless and bite the bullet. The ABS is printing like a dream now! I'm using Hatchbox black PLA @232C and 95C on the bed. I'm actually contemplating re-printing the rest of the parts in this. The only thing I've done a bit differently this time for the ABS is that I have the temp in the room cranked like a sauna. About 28C so it's like the Bahamas in here .... lol

Ok ..... that's my rant