OmioCNC report

[Mmpie]

I had wondered the same, it was part of my thinking of filling the hollow parts of the bed with epoxy or cement. Quick, easy and cheap. You could fill the t slot channels on the underside too I guess. I think that with steel cross supports should offer a lot more rigidity.


Sent from my iPhone using Tapatalk

[dharmic]

Unfortunately most of the stresses of deformation on the bed are going to be tensile rather than compressive - in my mind epoxy or cement would just crack. Some kind of camlock between the slots might keep them together and lined up and spread the load across the planks but there isn't a lot of room to work with for such a mechanism inside those slots.

[RCaffin]

At what point do you accept the machine as it is - and start using the acquired knowledge to design a better one?

Cheers
Roger

[dharmic]

About the point at which I realise no amount of 4mm max thickness backing is going to make an appreciable stiffness to a bed which is only supported at the ends (not even side support) and a slab of 20mm 5083 is going to cost me $350 before I have it machined square, surfaced and drilled in a tooling thread/dowel grid?

Yeah, that'd be the point at which I murmur something along the lines of "fvck it", screw the deck back on as it is and find ways to work around it for the occasional job where it becomes a significant issue.

And investigate the price of getting three phase to the house

[RCaffin]

I suppose it would be too much to hope to find a second-hand cast-iron table complete with T-slots? (I find the T-slots very useful.) It might be worth seeking.

Ah yes, 3-phase. I got it installed before Elcom went all commercial with their charges. They mumbled, but provided. It cost me 4 rolls of 60 A wire, some 45 mm conduit, and an hour with a small ditchwitch. Yes, I went underground: power one side of the trench, telephone on the other side.

Cheers
Roger

[Mmpie]

I don't know enough about stress and different types of forces to comment on that. When I get round to it I will try it anyway and run whatever tests you can think of. I have a litre of epoxy expiring on the shelf doing nothing which will surely be enough. It's nothing ventured nothing gained as far as I'm concerned, it's not going to cost me anything so worth giving it a go. If it works - great.

Another option could be to flip the bed over, drill out some holes, slide some steel bar down the channels and bolt it in. With the 3/4mm strips running perpendicular underneath it should offer a lot more support?

Sent from my iPhone using Tapatalk

[dharmic]

That might be better, provided there were plenty of bolts. Or maybe epoxy the steel rod in - the epoxy with a steel spine is much less likely to crack as the steel would take the tensile load, preventing sufficient strain (deflection) to crack the epoxy.

The biggest one to sort out IMO is locking the adjacent planks together at the edges to spread the load.

Meh, over it now. I've put the deck back on as is and run a short job, it was fine.

[RCaffin]

I am trying to be realistic here, and not nasty.
Might it be appropriate to simply say "You can't get there from here?"
And then use it and appreciate it for what it is.

Cheers
Roger

[dharmic]

Not nasty at all, Roger.

Mmmpie, you'll see from recent posts I've gone through the whole "can I stiffen up the bed" thought process and abandoned it. I'm not going to say "can't", but I am saying "simply not worth it".

Because once you stiffen the bed and start depending on it being more accurate you'll discover that the frame isn't square, the gantry sides flex, the spindle mount plate flexes, the gantry twists on the rails because the carriages are too close together, it's impossible to reliably tram the spindle and a litany of other sources of error.

Much easier to say "It's good to about 0.25mm, maybe 0.1mm provided your part isn't too big, allow 0.5mm or more if it spans the bed" and get on with it. Most of what we do, those kinds of tolerances are workable anyway.

Need better? Put the hundreds, then thousands, that you'll burn chasing the OmioCNC down the rabbit hole in search of accuracy instead toward a Tormach or Haas or Bridgeport or converting a Hafco knee mill or something like that instead.

Oh yeah, got my quote - $1600 to a sparky plus another $600 to Western Power for the connection. If I end up getting a VMC or mill turn this wouldn't be a huge expense in the scheme of things, but I ain't going there until I'm sure I can't do it on a single phase

[RCaffin]

Hi dharmic

Right - understood.

Now, 3-phase. I run my manual mill off 3-phase, although single-phase would be fine. I run my machining centre off single-phase with (I believe) a 6 A fuse. This whole business of 3 kW spindles is imho a gimmick. I am machining steel and titanium with a 500 W Baldor DC Industrial motor. It is quite sufficient.

Cheers

[Mmpie]

Oh don't get me wrong I completely agree, you can't polish a turd. I'm just trying to put some ideas forward. Although I don't see any harm in making cheap, easy minor upgrades such as these ideas for the bed. £10 of steel and 30 minutes cutting, drilling and bolting it in is nothing if it helps even just a little. I'm not about to go spending a fortune on servo motors, fancy electronics and whatever other things people do chasing as close to precision as they can get. If I really need that sort of tolerance there's a machine shop over the road that gives us a decent hourly rate.


Sent from my iPhone using Tapatalk

[dharmic]

All on the same page.

I keep surprising myself with what a decent job this little machine does and how much of the tolerances we demand seem to be just chasing numbers.

I'm not saddened by the fact my deck went back on untouched and every time I look at a better machine I look at what realistically will be $20k (by the time I get power sorted, tool holders blah blah blah) which I'll never get back on a hobby machine.

Bang for buck, this little Omio is pretty ace.

[extent]

Some depends on where you are too, aluminum prices in aus sound really high. I got a sheet of 3/4" 6061 big enough for two full sized fixture plates (not full bed replacement) for about 200 bucks. Machined it myself and surfaced it in place, because that's plenty good enough for my needs.

[dharmic]

Half the battle is buying stuff in less than full stock sheet sizes. There are places that do it but they take their pound of flesh, understandably. Then there's the WA Tax where the price of things double when you have to get stuff shipped across the continent if you don't buy from one of the very few local suppliers. The 350 Australian Pesos for the sheet does work out to about USD250 I think. Still - it's enough to have me asking whether I really need it and the answer was "no, you don't"

[madeinoz67]

Half the battle is buying stuff in less than full stock sheet sizes. There are places that do it but they take their pound of flesh, understandably. Then there's the WA Tax where the price of things double when you have to get stuff shipped across the continent if you don't buy from one of the very few local suppliers. The 350 Australian Pesos for the sheet does work out to about USD250 I think. Still - it's enough to have me asking whether I really need it and the answer was "no, you don't"

Maybe a WA group buy? I'd be in


Sent from my iPad using Tapatalk

[dharmic]

Two more takers and I'd at least go to the trouble of getting a price on a sheet of 20mm waterjet cut to four pieces of the right size for the bed.

[Mmpie]

Get the grinder on it


Sent from my iPhone using Tapatalk

[dharmic]

Got a mate who works at a waterjet / laser cutting place. The cut would cost nothing compared to the cost of the metal and they get stock cheaper than I can (and pass those savings on).

[javanree]

If you are going to replace the bed with a slab of aluminium, try to get tooling plate quality, which is cast and thus has no stress in any direction. This means that drilling won't warp it too bad, unlike normal stuff. Would be a shame to spend $100+ on a piece to see it badly warp after putting in some holes to make it useful.

I've posted some pictures in the past of my X4-800, I went lazy and got a prefab T-slot from cast aluminium from vakuumtisch.de delivered to my door for roughly $250 and so now I have a nice T-nut table. For me it was worth the money I spent, now I just need to align my machine better (the Y-axis is within 0.01mm of the bed over 250mm, but the X-axis has a nasty 0.2-0.3mm lean to one side. Haven't checked Z yet...

[dharmic]

One of my clients was already putting in an order for a bunch of stock so I saved myself the hefty shipping fee by lumping in with them, I have a 600 x 800mm "offcut" of 20mm 6061 on its way for way cheaper than the 7075 option and much, much less than cast plate. Still a touch over AUD200 though. Add another $50 for the two carbide drill bits and ream, and this got expensive for a hobby machine. Still, should make for some tighter tolerances and convenience in setting up each job.

My intention is to fill it with a 30mm spaced grid of alternating M8 thread and 6mm dowel holes, plenty big enough for the kind of work I throw at it. I'll use the extra to make up some sub-plates to hold the 4th axis and tailstock, vice etc. Considering buying a piece of 500x750mm float glass to use as a reference surface to square up the machine first - or at least work out where to shim the bed for flat. Then clamp down the new bed, drill, tap, and finally surface it as it sits to deal with any warping issues.

Haven't quite figured out yet how I'm going to drill and surface the areas outside the working range of the spindle, I may end up begging a favour of the client's workshop super to run it on his Bridgeport or Haas machine over a weekend instead of trying to do it here. If he can get it fully drilled and surfaced parallel then all I need to do is set it up and get it level.

This might also work out in the timing of things, give me a chance to get my machine back up and running after I killed it yesterday. I suspect I over-taxed the power supply I added to the black box to run the UC400ETH controller and laser module I added to it because it's just died and refuses to connect to the laptop any more after running the laser at full power for a longer job.