Core-XY based on Rat Rig V-Core and Duet 2 Ethernet

A Former User

@deckingman said in Core-XY based on Rat Rig V-Core and Duet 2 Ethernet:

... machine then maybe converting it to CNC at a later date. The conversion shouldn't be difficult - at least for a 3 axis machine. It's more less just a matter of replacing the hand wheels with stepper motors I'd have thought.

Slightly OT now, but, have had the same thought - seems the low end machines come with lead screws instead of ball screws so may have to replace the screws or accept some backlash. At least that was my conclusion.

snoozer

@bearer

I know in Mach3 you can work with backlash compensation, seen a bunch of stuff on YT. I believe in the F360 CAM module there is also something regarding backlash that can be configured. With a dovetail guide on a machine it would be a tight enough fit that you dont have it "bouncing" around and its likely to only come into play on changing direction. On an actual milling machine (remembering back to 1st apprenticeship) you only operate in one direction for cutting, not back and forth because the tool rotation and the type of cut dictates which direction to cut. Its not like a router. The move back is usually not a cutting move but instead a move to get the tool of the surface and fast travel and then re engaging. If you move in the axis that is responsible for your surface measurement then yes, you must take backlash or play into consideration. If you cut a perimeter around a part then you would have to compensate for backlash as well but you should be able to measure that and it should not change unless your lead screw or the lock is wearing out.

For me personally I am looking into a CCC (Cheap and cheerful China) type CNC router. 200-300mm bed size will do for me. Mostly light aluminium and sheet plastic stuff and engraving front panels etc.....

snoozer

@bearer

And one thing I have totally ignored in my post above, a proper milling machine would have linear encoders on the axis as such, not on the drive. That would by default compensate for positioning errors cause you have a closed loop system. Still you would take out the backlash play I guess by moving over the point and move back to accurate position you have the load on the correct surface on you drive. Closed loop steppers would not solve that in the same way cause that reading is not aware of the actual position of the work piece but only of the drive.

Jan

deckingman

@snoozer Ahh, those comments about "actual milling machines" and "apprenticeships" take me back almost 50 years - long before the days of computers let alone CNC machines. Long before linear encoders and DROs (Digital Read Outs) in fact. But yes, you are absolutely right about backlash. It was always assumed that any machine (mill or lathe) would have backlash so any cut was always made by moving the tool to a position beyond the start point, then commencing the cut from the same direction.

snoozer

@deckingman

My 1st apprenticeship was in 1986, not yet 50 years. In the workshop for apprentices we had DROs on the milling machines (Deckel, a well known German brand) and you could program the machine to cut out the feed after x mm but not real CNC there yet,we where supposed to learn I have also a vague memory there was mechanical sliders that could be set to cut out the feed...... just vague memory, can't remember exactly how that worked.

Jan

deckingman

@snoozer We are going off topic but as it's your thread, I guess it's OK. I pre-date you a bit - my apprenticeship was from 1969 to 1975. In the days when men were men, boys were boys and calipers had vernier scales. Calculators hadn't been thought of so we used slide rules and log tables.

Dougal1957

@deckingman said in Core-XY based on Rat Rig V-Core and Duet 2 Ethernet:

@snoozer We are going off topic but as it's your thread, I guess it's OK. I pre-date you a bit - my apprenticeship was from 1969 to 1975. In the days when men were men, boys were boys and calipers had vernier scales. Calculators hadn't been thought of so we used slide rules and log tables.

Takes me back a bit as well Ian I started my apprenticeship in 73 finished it in 77 but did get to use a bridgeport CNC Mill at college in 76 ish. (did actually do some real CNC Work in 78 sort of time whilst working in an Industrial engineering department for a large engineering company.

Doug

grizewald

@snoozer said in Core-XY based on Rat Rig V-Core and Duet 2 Ethernet:

Thanks for the info. I can't wait to get the V-Core kit, its two weeks now I think since I ordered, I hope they will ship it next week. I will find out how well its designed then regarding tolerances and stiffness.

After watching the YT video with the Bondtech extruders I had a nosy on their website, they do pretty nice stuff, Bondtech just was never on my radar.

Jan

Jan, I built a RatRig V-Core at the end of last year. I hope they have improved their quality a lot since then. Sure, they will go out of their way (even using next-day DHL delivery to replace incorrectly manufactured parts), but I would have preferred not to have the problems in the first place.

One of the worst was that the threads which should be cut in the ends of some of the extrusions seemed to have been done by someone who was drunk:

alt text

Checking all your extrusions when you receive the kit will save you time and trouble. It is not possible to build the printer properly if the threads are not straight as the bolts bind in the thread before they can tighten against the plate that they mount.

Once put together, the frame is rock solid. You could probably park a bus on it without doing much damage. Getting it square when you put it together will entertain you for hours!

The supplied print bed is indeed 3mm and anodised. Unfortunately, they didn't countersink the mounting holes at the corners and this robs you of around ten millimetres of print volume at the front of the bed (and the back, if you don't get stopped by your part cooling fan hitting the back Z-rails first.) The bed itself is 33cm square. My actual print size is currently 300 by 275 by 290 (X, Y, Z). I currently have just a 32cm square PEI sheet (just 0.4mm thick) as my print surface. The plate is heated from underneath by a 30cm square silicone heater with 420W at 24V.

My intention is to replace the PEI sheet with a 31cm square Anycubic Ultrabase. I need to countersink the mounting holes on the bed first.

If you want more information about how I built mine and all the various parts I used, just shout and I will get my camera out this weekend and give you all the detail you want.

snoozer

@grizewald Thanks for the info, Much appreciated ! I was promised a tracking number last week but nothing yet. I got in touch again and apparently some supplies got stuck in customs....... hoping for a tracking number now this or next week. I am not really impressed with this to be honest, they could have done a better job communicating this.

Re quality of extrusions etc..... I have ordered under the assumption that I have to rework the ends anyway to make then square. I bought a router with routing bits that should allow to cut aluminium. I have read about kits like this (not Rat Rig now but other V-Slot stuff) where the extrusions have not been cut to length properly etc...... So before I assemble anything I will make sure all is correct length or at least same length where it needs to be and square.

Re the print bed. I an actually just waiting to see what comes with the kit. As you said, non countersunk holes is what I expected. The pictures show the heads of the bolts sticking out the top. @deckingman pointed me to a UK supplier for aluminium stuff. They don't deliver to Ireland, I have asked that already but DHL offers a "virtual" UK address, so they can ship there and DHL then sends it on to me. So I don't expect to use the original plate in the long run. My idea is to use a glass surface, the original Prusa Mk3 PEI I have on my printer is starting to crack and it generally becomes messy after only a few months. With the inductive probe I cant use glass on the Prusa without modifications. On the V-Core I hope to use a piezo probe which would work on any surface.

Well.......... I am hoping the kit will ship next week and be here about 7 working days later..... I am glad I have started this as a project with no time constraints, otherwise It would have been very frustrating.

Greetings from sunny West-Cork
Jan

deckingman

@snoozer Sorry to hear that you are still waiting. IIRC I had to wait a long time too for the initial batch of Vslot extrusion that I ordered from RatRig. When it eventually arrived, it was well packaged with no sign of any damage and all the ends were cut nice a square and accurately to the lengths that I had specified. So I had no problems on that score. I tapped my own holes so can't comment on that aspect but given that the extrusion already has the correct tapping size hole in the end, it's difficult to understand how they could get the tapping so badly wrong for @grizewald - they must indeed have ben drunk at the time

snoozer

Finally, got a shipping number...... Supposed to arrive 1st week in May........

Jan

deckingman

@snoozer said in Core-XY based on Rat Rig V-Core and Duet 2 Ethernet:

Finally, got a shipping number...... Supposed to arrive 1st week in May........

Jan

That's bad. You must be very unhappy about how long this is taking. How long will it be between you placing the order and receiving the goods?

snoozer

As I said before, I treat this as a project so the time is not really an issue but ofc its not satisfying to have these delays. However....... the delivery has arrived at my local depot this morning. If I am very luck the delivery will be made today otherwise tomorrow.

I have ordered the kit on the 27/02/19, that makes it over 2 months. Gave me time to play with the steppers and the Duet 2. RatRig could have done a better job communicating the lead time but I also accept there are sometimes issues out of their control as well such as customs.

I will post pictures of the unboxing and show details about the quality of the extrusion cuts etc......

Hope to be able to update here this evening or tomorrow !

Have FuN!
Jan

PS: I was actually tempted to order an Anycubic Predator for the meanwhile to have something to play with. They are unbelievably cheap and have not so bad reviews and they can print big. I might still do that for a fast-print large nozzle machine. I hate changing nozzles between uses. I have the "Nozzle X" in my Prusa for everything but played around with 0.6 and 0.8mm nozzles as well with good results and FAST !!!!!!

snoozer

The kit has arrived. No visual damage on the outside. This evening I will open the box and check the content. I have other obligations 1st so probably wont start the build until the weekend.

Have FuN!
Jan

snoozer

Hello,

today was the day I started the build. I was positively impressed with the precision of the extrusion cuts. I have no precise squares and I can measure the length only with a ruler but the extrusions of meant to be of same length really exact. The threats in the extrusions also worked flawless where straight. I had to remove some metal shavings but hey, that's OK. Here a pic of all that came in the box (except the tools etc on my workbench ofc).

And the extrusions:

I started at 10:00 with assembly and have the metal work done now 18:30. The instructions could have been better but its not a Prusa, I have seen to late that there has to be a washer between the bearing in the rollers. Had to remove them cause I had seen that to late. I am not yet 100% sure the cube frame is actually square yet. I will have to do some adjustment. I have used a ratchet strap to tie all together before tightening all bolts but the X gantry does not hit the ends equally, one side is approx 1mm off at the front and the other end 1mm at the back. Not shure what way would be a good procedure for that, suggestions are welcome.

Here the state I am at so far.

Now printing a tool mount in Colorfabc's Carbon Filled XT-CF20. I will continue the build tomorrow. Time for pints and friends now.

to be continued.....,,

mrehorstdmd

You can get things squared up by measuring diagonals across each parallelogram in the frame. If they are square the diagonals will be the same.

snoozer

@mrehorstdmd

I know how to measure it, just I have no clue yet how to adjust it. I was thinking of criss crossing the frame work with adjustable wires....... but that would only bend the extrusions, not fix the actual adjustments.

THX
Jan

mrehorstdmd

I'd square it up as I assemble it instead of trying to assemble it first then make it square.
Since the XY relationship is critical, I'd start at the top of the frame and get that part square before bolting the rest of the stuff to it. Match the bottom square to the top, then add the verticals. If the frame is also the Z axis, get those verticals square with the top first then add in the other two. Check squareness of everything with each frame member that you add.

Keep in mind that a 1mm error over a span of 500 mm is only a 0.11 degree error.

deckingman

@snoozer Your X gantry issue is most likely caused by unequal belt tension.

snoozer

@deckingman

This is the plain mechanic only sitting idle, no belts yet hooked up. The actual frame is the issue I think. Once I have finished printing parts for the V-Core on my Prusa I use the steel table that I have for the Prusa to check the V-Core frame is level and square. My "workbench" is a rather light Ikea desk, probably not really flat and level.

Jan