Slide bearings on Aluminium Extrusion (no smooth bars)

As I mentioned in another thread [forums.reprap.org]. One thing I've been considering for a while is to get rid of the noisy ball bearings and use plastic skates running in the slots of the aluminium extrusion. After Erik showed interest I thought I'd try it out over the weekend on the Y axis intially - rather than hijack another thread I thought I'd start a new one.


Briefly I've used a piece of extrusion for all of the support needs for the Y axis, along with bosch-rexroth acetal slide runners (intended for sliding doors, but ideal for this purpose I hope - I've actually cut the slides into three pieces ).


I've linked the two extrusions with a steel plate at each end (it's what I had lying around), currently the RS box, the ATX feed and the Z axis are all connected as before.
The Y motor and its idler are supported in the upper groove of the Y frame using bent T-plates from Wilko (I'll going to try rack drive and this is maybe a temporary solution). At the moment the Y axis stop switch is just taped on, I'll be mounting more stably soon.


The arrangement prints fine at the same speeds as before - the most noticable change is in sound of printing (the three roller bearings on smooth rod made a pretty harsh noise at 100mm/s, now only the X does this).

The skates give very good lateral contstraint, will allow simple adjustment if the bed isn't level.

I'll be replacing the bed with aluminium composite board (a la kimbrown)when it arrives - currently I've made the bed from a modification of Erik's quick and dirty solution (I sandwiched some cellular polycarbonate between the aluminium heatspreader and a spare bed pcb, unfortunately this pcb is curled, unlike the RRP one, so I tried sticking the three layers togetether with two part silicone and wieghting it while it set - this has made the whole thing very stiff, but unfortunately the PCB curve hasn't disappeared - instead the aluminium has conformed to its curve :/) I've glued some aluminium angle onto the bottom of the aluminium sheet to pride belt anchors for now, but this should take the rack instead.


I've attached a couple of photos of the mod, showing the new versus old footprint, and close ups of the skates, motor mount, and belt clamps.

Next in line is making a dual Z column, and using extrusion/acetal to replace the two remaining smooth rods...

Cheers

Ray

Beautiful! - we need frontrunner's like you!, it has been back in my mind since you mentioned those skates, I love the idea of the bed corners riding the extrusion, I love the direct line of support to the bed

Do you feel any slack between the extrusion and the skates and do they run friction less, or do you have to have a preload to keep away any slack?

About the outer skate - I prefer 3 point adjustment of the bed and by bed I mean the glass and with your setup the inner two corners are as stable as can be, I would want the outer adjustment point to support the glass without any chance of bending it, so I think my take would be to use a very short skate for that, would also save on the friction the inner skate should be more than able to line up the bed in itself, great experiment, I'll be back, must sleep now :-)

Erik

Hi Erik, I'll try the three point plan when new materials arrive this was just a quick fix-up to see how it runs (as it happens, it wasn't so quick, sourcing screws of the right length, etc) - I plan on having a triangular support structure for the heater and glass to make levelling easier - this would be larger than the heated area itself (you'll notice that the extrusion I used is a little longer than the Ormerod extrusion - it's 500mm, ideally I'll have something between an A4 and an A3 printable area eventually, but I'll need a bigger heater!). You'll also notice that I had to turn the PCB though 90 degrees to avoid the extrusion on this try out (this works quite well for cabling as it happens) - I'm not sure how large a gap to aim for between the two Y rails - this is because I was using the predrilled holes in the heat spreader and PCB to mount the skates.

There is indeed preload - the slides are pretty much floating in X and Z - they have the same shape in section as the groove does, but with a sizeable gap in each direction, to take up the slack in X I drilled the mounting holes in the end plates at 1/4inch (6.27mm) which gives enough wiggle room to leave the skates floating or bind them fully. I did consider spring preloads etc, but decided on this method for speed of building - I squared the "inner" extrusion to the end plates, then left the front one loose and straightened it by sliding the Y carriage back and forth until it moved freely with no play and no binding, then screwed it tight and locked it down.

I compared the ease of movement between the original and this version, I was surprised that on the original, movement is easier in one direction than the other. On mine it's got the same resistance in both directions but is not as free as on the original (this could be altered I guess by changing the preload, or using springs for preload) - one very noticeable difference is that with a load of two bricks on the bed, the ball bearings excel, the acetal does get stiffer - this is why I think for Z I'll need two posts (to kill the moment caused by the weight of X, and have the X load on Z be purely vertical), I'm pretty sure that it'll have no impact on the running of the extruder along X though, so that's probably next, using a 20x20 section to keep the weight down It'll take two skates, maybe one in each side groove, with the rear one having a variable preload to keep the extruder horizontal, or perhaps one on top and one on the side (depending on how I mount the extruder motor - this might be a good time to experiment with direct drive and lose the Boweden!)

Cheers

Ray

I like that.... Nice work there.....
Bet it's heavy thou....

---

Please send me a PM if you have suggestions, or problems with Big Blue 360.
I won't see comments in threads, as I move around to much.
Working Link to Big Blue 360 Complete

Great idea. I assume its much cheaper than the ground rods and linear bearings? Perhaps you should add it to the RepStrap page.

regards
Andy

---

Ormerod #318
www.zoomworks.org - Free and Open Source Stuff

Nicely done.
I was contemplating the second Z support - could you be a gentleman and remind me what size the extrusion is (width not length)?
I am away from my printer for another 10 days and would like to come home to parts...
My thoughts are to keep the X axis the same size, double the Y and the Z.
I was going to buy a second heat bed and mount them side by side, will look now and see what options there are for larger beds.
Again, good work....you have inspired me to keep modding mine.

Hi Kim, thanks - it's not that much heavier actually - I've used two 500mm lengths of extrusion (instead of one 350mm length) - total weight of extrusion is 730g, and I used two 230mm lengths of 3mmx80mm steel for the ends - this weighs1.7kg/m so the steel weight is 780g, giving a total of 1.51kg - the original lower frame (extrusion, smooth bars bearings and plywood/mdf weighs 1.125kg, so I got a weight increase of 1.5/1.1 (36%) for a length increase of 500/350 (43%) - not bad going really (and this frame is much more rigid). Until I get the rack in (or if I stay with belt drive, move the motor outside the base), I can't really use the extra space, but I should be able to do an A4 sized print bed pretty easily in the current frame, when I get round to X and Z I intend using 500mm extrusion there too, but will have to look hard for a heated bed big enough to use that much space.

Cheers

Ray

Quote
kwikius
Great idea. I assume its much cheaper than the ground rods and linear bearings? Perhaps you should add it to the RepStrap page.

Cheers Andy - I cross posted on the aluminium extrusion forum, I'll take a look at the repstrap pages too - I think it probably is cheaper than smooth rod and bearings - I paid next to nothing for the steel (something like £10 for three eight foot lengths of it a couple of years ago - it's just guillotined 3mm mild steel sheet) the extrusion costs £5.50 per metre (ie in total for the base), and the slides cost 1.35 each but I used 1/3 per corner. all in, it's probably £10-15 of parts if bought new (including screws, T plates and a tap).

Ray

Quote
badman.teddy.edward
Nicely done.
I was contemplating the second Z support - could you be a gentleman and remind me what size the extrusion is (width not length)?

Thanks teddy,
I bought this stuff (bosch-rexroth 20x40, 6mm bore):
[www.aluminium-profile.co.uk]
which is identical to the RRP stuff - it costs £5.50 per metre and they cut to length

I think two plates side by side is a good plan - you'd probably need a beefier dedicated PSU and maybe external switching for one or both of them (and maybe a second external thermostat for the second one) - I've found A4 bed PCBs, but they have the same resistance (and therefore power from 12V) as the standard 20*20 one - which means it won't get to 100° form 12V, so if i get one I'll need a 24V PSU or similar I guess...

Cheers

Ray

Quote
rayhicks
I should be able to do an A4 sized print bed pretty easily in the current frame, when I get round to X and Z I intend using 500mm extrusion there too, but will have to look hard for a heated bed big enough to use that much space.

Ray, I keep trying to track down a supplier of Carbon Fibre heating tape. It's used in lots of 12V and 240V heating applications - like heated seats, gloves etc. There's some on Ebay but it's in the Russion Federation so might as well be on Mars. Anyhow, just thought I'd mention it.

---

RS Components Reprap Ormerod No. 481

Interesting Radian - I wonder if it's "special" or just carbon fibre tape they've measured the resistance of and sell as heating tape? (there's a lot of plain carbon tape on ebay, but it tends to be fatter - there's other stuff on the general web on composite/fibreglass sites that's cheaper and may be worth a punt - I'll get some and try it!). Looks like the resistance is around 8-20 ohms per metre (for tape ranging 15mm-25mm), so I'll have to try to remember Kirchoff's (?) laws and work out how to spread a network over an A4 area that'll work. Looks like people sew copper wire into the tape to connect it.

A couple of months ago I bought some screening copper tape (30micron thick, 5mm wide) from China, and it arrived last week, haven't had a chance to play with it yet, but it looks to have a resistance of around 1ohm for 10 metres - not sure I can squeeze enough onto an A4 plate to matter (but I'll give it a go) - my back up would be to tape nichrome to the back of a glass plate with kapton then use silicone to stick that to something like dibond - nichrome is appealing since it comes in a wide range of ohms/metre and it's easily handled/joined (the copper tape is pretty fragile and it would be hard to rejoin if it tore halfway through taping out a bed if taping to glass - maybe some thin tuffnol would be best so I could solder over breaks),

Cheers

Ray

Beautiful new edition!

But I've tried plastic bearings..

Ormerod bamboo bearings are the true heart of the machine .....
A bit more expensive but much more silent.... and fast..

As a economical solution yours is beautiful!

Bearings in the original kit made ​​me laugh too many times...

Dario

---

Ormerod 187
Firmware Electronics: Duet 0.6
Firmware Version:1.18.1 (2017-04-07)
Web Interface Version:1.15a
Slic3r 1.2.9a and Simplify3D 4.0.0
[www.dropbox.com]

Thanks Dario - you made the bearings out of bamboo? I thought it was just the structure - respect!

Cheers

Ray

Hi Ray
Thanks for the extrusion sizes - ordered my upgrades.
In a way printing has been so nice I am hesitant to mod out of fear, I guess worst case a days work can return it to stock.
I have been looking at [www.ebay.com]
It is a digital thermostat. Many others but most I find have a high temp of 110 degrees. That one in the link is 300. (Those numbers are both Celsius)
I can power that unit from the PSU, I need to find out what the amps of the ormerod heat bed are before I think about external PSU options, or whether I could get away running it all from a 900w modular PSU I have from an old PC.
Not sure if the sensor from the above will work, (too big) so will need to see if another can be used and calibrated.
Downside is there will be an extra control unit to power up before printing. I can't see myself rewriting the pronterface software so it could manage 2 heat beds...and the net is so slow on this boat even searching for whether the duet has enough ins and outs for such a thing is patience testing. I had a look last night and large beds were hard to find.
In my dream world, I would add a third horizontal extrusion (imagine yours plus another same distance out again) and have four of the ormerod beds allowing for a 400mm x 400mm print bed. I would guess the stepper would need upgrading but that is no major concern. Except for one project I am not sure I need a bed quite that big.
Time is on my side (even if net band width is not) so tonight I will search for a multi input thermostat and see if the 400 x 400 is too crazy.
Keep it up!

Quote
rayhicks
Thanks Dario - you made the bearings out of bamboo? I thought it was just the structure - respect!

Cheers

Ray

my sincere apologies
I thought about bearings...
it's my first time in a forum...


I leave immediately
ciao
Dario

Hi again teddy - you're another seaman then ?(KimBrown does much of his thinking and printing on a tug)

I know what you mean about the modding - I've been buying bits and bobs with a view to keeping the original Ormerod intact and building my own beside it, but the time felt ripe to try out at least one axis on the weekend (pretty much everything I THINK I need is here now, just a case of finding time to try (and then revise over and over again!).

My main goal with the development is to get something that a colleague can walk up to, switch on, and print - without having to jockey it too much, and that is stable enough to not need frequent re-calibritation. I bought a few 500mm lengths of 20X20 and 20X40 and it seems a shame to cut them, so i'll probably end up with a machine that's a 500X500X500 cube - so potentially a 250 in Y (I like the skates in the corners) by 400 in X print area keeping the same axis geometry (though there's nothing I intend printing at that size and it'll be end-stopped to a smaller bed until I find/make a suitably powered large one ).

On the bed-design front, the Prusa-derived PCB bed heaters (like the Mk2a on the Ormerod) have a design resistance around 1 ohm (call it 1.2 for maths sake), so they pass around 10 amps from 12V, giving 120W of heat on a 64sq inch (or around 440sq cm) surface which works nicely - it reaches 100°C in a reasonable time and can hold it from the 12V line of an ATX (which many people have shown is nearer 11V on the PSU's supplied)- so the nominal power to area ratio I'm aiming for on a larger bed is the same 120/440 = 0.273W/cm² The larger bed PCBs I've seen unfortunately seem to have the same resistance as the Mk2 (and the Mk1), so they have a larger surface area, but the same power (for a given voltage), so would take longer to heat up, and may not actually reach the same target (since they have a larger area to lose the heat over. Doubling (or quadrupling) the Mk2's like you've suggested and powering them in parallel using an external controller is probably the easiest/wisest route, providing you've got a PSU that will regulate 12V at those loads (240W seems readily available on eBay in the 20's of pounds, as dmould and dc42 have testified, 480W may be trickier, but bobc suggested server PSU's and it sounds like that's the kind you have). One drawback is sourcing 4 nice flat PCB's - I bought one Mk2 from eBay and it works fine as a heater, but it's got a pretty noticeable dome, I'd hate to have four of them and try to press them flat with the glass cover (I'm thinking of using the thermistor hole as an anchor for a puller actually and using a flat thermistor elsewhere, or just not using it and trying to fabricate a heater as described above in the thread using radian's carbon fibre suggestion or maybe copper of heater wire).

Looking at the thermostat link you provided, it looks like it uses a thermistor "Sensor: NTC25℃=100K B=3977" - and I guess you could pop it out of the metal shell or fit your own the beta is too far out for my favourite thermistor package [www.rapidonline.com], but pretty close to a few others - one spec that would be worth asking for before buying is what the DC rating on the relay is (5A@240V AC doesn't necessarily translate to a greater amperage at 12V DC, depending on what the relay contacts are) maybe look for a thermocouple based temperature controller (but the ones on eBay don't seem to have much info that I can see atm),

Happy surfing/cruising

Ray

Quote
Ormerod187
I leave immediately
ciao
Dario

oi! come back!

Quote
badman.teddy.edward
Downside is there will be an extra control unit to power up before printing. I can't see myself rewriting the pronterface software so it could manage 2 heat beds...and the net is so slow on this boat even searching for whether the duet has enough ins and outs for such a thing is patience testing. I had a look last night and large beds were hard to find.

Instead of looking for a larger bed or additional controls for a second bed, assuming you have other 12v rails available, you can always remove the ribbon cable and use the other outputs on the board for the bed heater but direct them through a relay, using that to control the on/off of the additional 12v supply to additional beds. The thermistor I would try placing in between the beds (if using 2 beds) or at the meeting corner (if using 4 beds), it should be relatively easy to add a gap with interfering with the track of the heater. I would also place the thermistor wiring to the duet away from the ribbon cable (next to the heat bed output).

I am sure there are others who can explain it better or even if it would work, but food for thought.....

Mike

Edit. If you look at the following link Think3dprint3d and look to the lower right you will see the bed heater and bed thermistor points, the Duets we have, don't have the mounts.

Edited 2 time(s). Last edit at 03/07/2014 04:46AM by Mykey.

Good points Mykey , thanks - with an aluminium heat spreader over the PCB's there shouldn't be much difference between a thermistor reading from between the boards and one from the centre of a board I guess, and running a remote relay/s would be easy and cheap for providing power to the boards - I'll certainly give it a go with the two PCBs I have and see how it goes.

I've already added header pins for the thermistor, and two 1.5mm² solid copper insulated wires to the places you've indicated on my replacement RRP duet and I've used terminal blocks attached to the solid copper to feed power to the bed via 20amp multistrand flexible cable (rather than order some PCB mounting terminal blocks which i find fiddly to use). Much safer to attach the thermistor over there rather than pick the right pins from the ribbon header

Cheers


Ray

As it's your topic Ray, I'll post this here - although it's not related to the slide bearings, but to your digression towards larger bed sizes...

In case you didn't know it already, the Duet has a provision for an external switching transistor (MOSFET). These are typically cheaper than electromechanical relays and have far superior characteristics for handling high currents. The other bit of magic is that (unlike a relay coil) the Gate takes no DC current - although it looks like a capacitor so needs current to charge it quickly. The point is that you should be able to parallel up a few external MOSFETs if you simply copy the way T2 is shown wired in the Duet circuit here:



The input capacitance is something in the order of 5nF so with 4 in parallel the switch-on time of each will be the resistance of the pull-up resistor R18 times the total capacitance which come to something around 20us. Discharge will be faster because it goes through TR1. The aim is to make the transistor switch on and off as fast as possible as it dissipates power when in between states, but 20uS with long-period on/off cycling shouldn't generate much heat. If it did, R18 could be dropped to 270 Ohm.

The only tweak I'd do would be to add a 68 Ohm resistor in series with the gate of each MOSFET to combat ringing form the inductance of the connecting wire back to EXT_FET

---

RS Components Reprap Ormerod No. 481

Brilliant. I like where these ideas are going.

Mykey, that is good thinking. My brain didn't go to just placing the thermistor in the centre of the entire table.
It would still keep it simple as Ray wants (and I want) where you turn one thing on and start printing.

Radian, nice one.
I am going to try and find the duet specs and run with that idea. A few mosfets and it is pudding time.

Ray, I work in offshore oil and gas. Most of the time on fpsos (boats that produce, and store oil) and sometimes on rigs. Atm I am on a boat, and it is crazy big! Oddly this was the first boat I ever worked on about 7 years ago, I always revisit places but normally not with such a large break. Everything is the same yet different. Kinda odd. Meh.
I work in the integrity side of the industry and this seems to see me offshore for about half the year.
I am going to steal your idea and build this as a project printer, keep my first ormerod (relatively) stock. It has the obvious advantage of being able to easily print if something comes up during the upgrade process.

I will go ahead and do the full 400x400x400 print size and see how it goes.
I just spoke to the OIM out here and it seems I may be kept back until the 26th! Heh. This always happens when I get excited about playing with toys.

Am I the only one left printing on glass? It seems that way. I am thinking alu under a 400x400 glass sheet. The alu is probably redundant with glass too though. I just started using a 200mm wide roll of kaptan. Applies really well and my obsessive compulsive take 2 hours to lay the tape down so that each row is perfectly in line is now not required. Unlike when you damage the narrow tape and you can peel a row or two off with this it's replace all which is a bit silly.

Ok, now to order some hopefully not warped heat beds.
Great thoughts team. I appreciate the ideas - I have found some momentum.

Oh, and apologies for thread hijacking.

EDIT: I just scrolled up and noticed you still print with glass, Ray

Edited 1 time(s). Last edit at 03/07/2014 07:08AM by badman.teddy.edward.

I understood not a word of Radians post, way over my head! But if it means that even a relay wouldn't be required and it can be done on board, then thats a bonus, I just wonder about the power requirements and the wiring to each bed?!?

I need things simple (hence the simple thInking!)

I think this is definitely heading in the right direction, at this rate there will be multi headed Ormerods capable of 400 x 400 x 400 prints relatively soon.

If someone can point me in the direction of where to get the extrusions, sliders etc, I may well start to build one of these monsters!

Thanks,

Mike

Thanks Radian - I'd idly thought about the Ext_FET connection, but didn't know how best to use it, thanks for pointing out the design issues (I'd have overlooked just about every consideration you raised!).

As a further digression, it looks like the copper tape I bought has a resistance of around 0.25 Ohms/M (my bench supply says that it takes 1.5V to push 5.0amps through 1.2 metres of it anyway), which means that if I want the same amount of power per surface area from an A4 bed (which I calculate to be 0.8Ohm to get 15A from 12V giving 180W), I'll need to run a total of 3.2 metres of it if I do a single strip, or 2 strips of 6.4 metres in parallel. A4 is 30cm long, which means I'd need "10 bars" for a single continuous strip or 40 bars for the parallel configuration (A4 is 21cm wide, so theoretically I could squeeze 40x5mm bars on there with a minimal gap between, giving almost total coverage of the plate...reality checks accepted ) Is 15A pushing the Duet's onboard FET or tracks?

Cheers Ray

Quote
Mykey
If someone can point me in the direction of where to get the extrusions, sliders etc, I may well start to build one of these monsters!

Hi Mike, I got my extrusion from here [www.aluminium-profile.co.uk], the sliders are also available there. I've only done a few hours printing on this, all's going well but no long-term guarantees! it might be necessary to use wheels to lower the friction (and chance of binding) particularly with an overhanging X arm like on the Ormerod (Im planning to have two Z posts and twin leadscrew drive so both ends rise together, so slides should be ok in that situation), but it remains to be seen.

There are two other extrusion types (OpenRail, which adds V guides to standard T-slot extrusion as used on the Ormerod, and V-Slot, which has chamfers on the upper surface of the T-slots to guide wheels), but I'd rather just see what can be done with the cheap and available T slot (I guess that both OpenRail and V-Slot will have lower friction and better constraint, but I don't think they're necessary for the low forces used in printers - I may be wrong)

Cheers

Ray

Quote
rayhicks

Quote
Mykey
If someone can point me in the direction of where to get the extrusions, sliders etc, I may well start to build one of these monsters!

Hi Mike, I got my extrusion from here [www.aluminium-profile.co.uk]

Bonus, I just had a look at their site, they're only 30(ish) miles from me!!!

Thanks.

This is a horrible mock up but oh well (<5minutes - it is bed time) - laugh at will Roughly what I plan to do. May turn it into a box too...shrug.

Mike I am going to put together my parts list tmrw evening.
I am happy to undergo this mod journey with you - unfortunately I won't be able to commence the build until (hopefully) the 17th, or (more likely) the 28th due to work away from home.
Will post when I work out what I am doing with the mosfets - probably make a design so the controller box fits it all- may as well make it so it will take the Duex4 at the same time...

will let you know..

Quote
badman.teddy.edward
This is a horrible mock up but oh well (<5minutes - it is bed time) - laugh at will Roughly what I plan to do. May turn it into a box too...shrug.

Mike I am going to put together my parts list tmrw evening.
I am happy to undergo this mod journey with you - unfortunately I won't be able to commence the build until (hopefully) the 17th, or (more likely) the 28th due to work away from home.
Will post when I work out what I am doing with the mosfets - probably make a design so the controller box fits it all- may as well make it so it will take the Duex4 at the same time...

will let you know..

Better than I can draw!

I know what you mean about time! Stoopid work!

I will probably order some extrusions, sliders and see if I can find any suitable sides over the next few days, or (preferrably) drop into see them. And find a good place to get motors etc.

I think it would be a good idea to start a thread documenting the build....

Mike.

Quote
Radian
The only tweak I'd do would be to add a 68 Ohm resistor in series with the gate of each MOSFET to combat ringing form the inductance of the connecting wire back to EXT_FET

With the on-board mosfets, about 200 ohms is a better value to use. Without it, the mosfets turn off far too quickly and this causes a lot of ringing in the load wiring.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Thanks DC..

I took the step of giving the larger print size / quad bed section of this thread it's own topic, it has migrated to [forums.reprap.org]

@ Teddy - didn't see your reply till now - you're not hijacking, that's the idea of putting ideas on the forum I print on glass, but not directly (I started using plastic cement when I got fed up of kapton and didn't have any abs and acetone to try - it works pretty well for PLA and ABS and seems to have caught on). I like glass since it's hard flat and elastic (you can bend it a bit but not much, and it springs back to shape, aluminium plate bends more and doesn't spring back so well...). Actually today I've run with the idea of fabricating a glass heater - using the tape I mentioned above, looks like it will work OK [edit from small-scale tests], I've used an A4 picture frame for the substrate and stuck strips of copper on that, then soldered the strips together - I had to go and see the kids before I got the full rehearsal ready, but might get it finished and tested over the weekend or Monday, if so I'll post pics)

Ray

Edited 1 time(s). Last edit at 03/07/2014 05:16PM by rayhicks.