Let the fun begin !!! ???

Attachment 13199Attachment 13198
the frame is nearly finished. i have to weld in the table locators on each side and give it a lick of paint. it looks big ( it is ) but i said to myself that i may need upgradeability considering the growth of the DIY cncs over the past number of years and the bigger routers that are coming on stream.
Attachment 13200
i also welded the electric cabinet frame together, 800 x 600 x 200 deep ( just under 3 lengths of 20 x 20 angle). i have 1mm galvanize to sheet it and some hinges knocking about somewhere.
the frame and cabinet should be finished by next weekend and then i'll get a moat organised for the dreaded epoxy pour!!

Quote:

---

Originally Posted by PAULRO

Z1: 0-0.02C or Z2: 0.02C-0.05C . i'm sure either of these will do the job but would like to know which preload is the preferred choice for wood routers.

---

You get a slight increase in stiffness with higher preload, at the expense of lower life and higher driving force. For a wood router the change in stiffness is negligible compared to other parts of the system, so I'd go with Z1.

At long last the very kind people in shannon customs have released my profiles and ballscrews,:whistle:i can now start in the next phase of the build.

Well at long last I can get back to the build of this baby. i had to take a break and head abroad for a time but i'm back now with a clear run for the next few months . i poured the epoxy and left it cure for 72 hours and i'm pleased with the result. i had a heater running and had the cubicle locked up to control the heat. it has some air bubbles but i can live with that.

I assembled the gantry and attached the bottom linear guide with its housing plate. i then used this to mount a hand held router and skim the the top linear slide bed. i took 0.4 mm deep across the entire length to ensure parallelism. it turned out o.k. but the weight of the router left an angle on the base so used the z axis assemble instead to machine it again and it turned out pretty good.

i have the z axis built and assembled and i have the pulleys drilled and reamed ready to go but i rotated a ball nut to far and the bearings started to fall out so now i'm left with a basin of bearings and an empty ball nut . i have looked on line for some tutorial to enlighten me on the re-assembly but cannot find any thing.
the ballnut has 6 plastic change overs in it , 3 pairs at 180 degrees to each other . the ball nut is a 16 x 10 pitch with a double start thread. if any body has any experience of these i would be very obliged at this stage . i'm itching to get back at this build and move on to the electrics,

thanks in advance , Paulro

i will post more pics tomorrow once i have the ball nut assembled.

You should be able to find examples on this forum - plenty of us have made that mistake!

Two ways I've found that work well:

1) Put grease on the balls to stick them to the nut, then gradually insert the retaining tube as you go to support them.
2) Put the biggest magnet you can find on the ballnut, so the balls stick to it in side. Carefully place them in the right place with something non-magnetic.

Looking good so far, it happens see links below:

How to get the balls back in a ballnut?

RM1605 Ballnut Tubes ( keep the bearings in ball bearings in a ball nut, how many?

Thread: ballnut - balls back in via the side cover and using less balls than original[/B]

Thread: Re-filling a ball nut

There is also a few vids on ebay showing how to do it:

http://youtu.be/dGp56l0bsdA

http://youtu.be/4rjWH0UPA9U

.Me

I like the idea of skimming the top with a router, provided the bottom rail is perfectly flat.

Same ball trouble here, this was the best video I found, the method worked great.

https://www.youtube.com/watch?v=5qbVrRYzK5U

The grease did the trick ok. 3 circuits, 16 balls ( dia. of 0.1245") in the first two and 17 in the last . it took a couple of trys but i got there. so cheers for the advice and videos links, very informative. the extrusion i used had a slight angle on the broad side and an incline on the narrow side .it played havoc with my assembly so i had to re- machine some supports and change the design of the gantry. i also spent a fair bit of time on the bottom Y rail to get it flat.

Latest pics.... i clocked the bed top get it as true as possible . i will still M/C it when i get it powered up . i want to minimize the dust as much as possible. i have 2 nema 23 s , 4 NM for the x axis . i'm not sure if 1 motor would be enough and go with a belt drive for both ballscrews. i have read up on the different opinions here on this forum and on 1 versus 2 motors . the gantry with brackets , cables and spindle come in at 65 KGs .
i'm going to move on to the electrics next and spec the various drivers, power supplies and other miscellaneous parts that will be required to get this baby up and running
i would like to start on the electrics next so i'm hoping that someone here will give my nema 23s the thumbs up or down because i don't want to buy drivers for a motor configuration that is inadequate. i have pumped the numbers into Irvings spreadsheet ( what a piece of work , by the way) and they look good . i have read Eddys build a few times and hope to get my electrics as close as possible to his. Eddys ganrty is not as wide as mine and that why i'm curious about the 1 motor versus 2 on the x axis of my build and more to the point should i go with a nema 34 motor? any thoughts on the subject would surely nudge me on.
my build changed as it went on due to extrusion being extrusion. it all looked great in the CAD model but when i went to assemble it in the designed configuration ( L shape bolting 2 pieces together at 90 degrees) it kicked up and played havoc with the brackets supporting it so i had design around the headache. this in turn lead to a change in the z axis configuration. i have left scope for a revisit to mechanical build but for the time being i want to get it running and discover if it needs a revisit. Paulro

more pics .the motor config. is repeated on the far side .

Quote:

---

Originally Posted by PAULRO

more pics .the motor config. is repeated on the far side .

---

Why do you have what looks like a 3:1 ratio on the screws.? Even with a 10mm pitch screw 2:1 is bad idea if your planning on using as wood router.
2:1 ratio will slow you down to around 5 or 6mtr/min max rapids and for woods you'll want to be cutting at that. 3:1 would be even worse.!!
Stay with a 1:1 ratio is my advise if your cutting woods.

Regards the motors then it's preference thing mostly 2 x 3 or 4Nm 23's will easily handle 65Kg. I prefer single motor with belt for accurecy and piece of mind reasons but I'm equaly happy to use slaved motors provided the correct drives and precautions are used. Slaved motors work ok but I would only use them provided you have Digital drives that can detect stalled motors or have fault output so you can E-stop the system.
Digital drives have made slaved motors much safer in my opion with less stalling as they handle current so much better and can easily detect for stalls. They also dont loose steps as much as the old drives did.

it's a 2:1 ratio. 18 tooth on the motor and 36 on the screw. i read a post in one of the blogs here that 2;1 ratio would be good for detail work but i can change to 1:1 if i could still achieve the required detail. thanks for the advice on the digital drives though.

Quote:

---

Originally Posted by PAULRO

it's a 2:1 ratio. 18 tooth on the motor and 36 on the screw. i read a post in one of the blogs here that 2;1 ratio would be good for detail work but i can change to 1:1 if i could still achieve the required detail. thanks for the advice on the digital drives though.

---

What kind of detailed work.? 1:1 ratio on 10mm pitch will get you this kind of detail.

https://www.youtube.com/watch?v=wjDvimSdH8M

that will do it for me so . i'll go with the 1:1 . i could replace the 18 tooth pulley for a 36 tooth pulley to create a 1:1 ratio or replace the 36 tooth pulley for 18 tooth pulley . i suppose it wouldn't matter either way , or would it? i'll replace the 36 tooth pulleys and be done with it.
excellent work on the calender. if i can reproduce that detail i'll be more than happy. i might get some more details on the M/Cing of that when the time comes , if you don't mind.

By the way Jazz , i just had a look at the u tube video of leadshine closed loop stepper(https://www.youtube.com/watch?v=ZMZdCcLQc4M) you posted and i have a couple of queries.( great video, by the way) would it be worth my while to attach encoders to my nema 23s?. i had planned on using AM882 drivers but i don't think they take feedback signals. i know they have stall detect ( you mentioned that previously).i would like to get as much info. as possible before i decide to press the BUY button. the closed loop system is a more attractive option in terms of constant feedback but that's as much as i know at this point.

i had some play in the fixed bearing blocks (0.3 mm) . stripped them down and set them up again and still had end play. had to make some shims to eliminate it .

i have gathered power supplies over the years and i'm, hoping to use some of these.i will have to either make a power supply for the drivers or buy one . i haven't decided yet . http://www.mycncuk.com/attachment.ph...id=14624&stc=1

more bits gathered down the years. i hope to use most of these also.

Quote:

---

Originally Posted by PAULRO

i had some play in the fixed bearing blocks (0.3 mm) . stripped them down and set them up again and still had end play. had to make some shims to eliminate it .

---

Well unfortunatly this happens but when shimmed they are fine. Looking at the shim I'm not quite sure your shimming the inner race or outer.?
If it's the inner then the bearings are in the wrong way around. You want them back to back with a shim between outer races.


Regards the Leadshine Closed loop steppers then they are nice but expensive. If your using Mach3 then just attaching encoders to your steppers won't help at all because Mach is not closed loop and all it can do is report the encoder position it cannot correct for error like the leadshine servo stepper drives do.

The closed loop stepper drives actually close the loop within the drive not in the control software so when you see it correcting position Mach doesn't know anything about it. Now this is good and bad at same time.?
It's good that the drive monitors position using the encoder and corrects for any error but at the same time if any error as happened it's still an error which causes a break in continuity between drive and control software. Because the loop is contained within the drive and never reports any small positional errors to mach then In affect Mach becomes in front of the drive for a short period before the position catches backup in that axis. Other axis will carry on in front So Depending on the error amount it can affect the job so the error still shows up in the finished job.!! . . . . . This is also true of servos or any other drive that doesn't have true closed loop system that works in tandem with the control software to sync axis together.

If the positional or what's called following error gets too great then the drive will fault and e-stop the system (which in affect is what Stall detect does but much cruder) This following error can be set in the drive parameters to a small amount so if goes outside parameters then drive faults and system E-stops.
This is how most servos work but like I say the small errors have still happened and at some point will show up in the work.!!

This is why you'll often see people who have experience with both Servos and steppers on slower machines like mills etc say that Servos are no more accurate than steppers if steppers are setup and tuned correctly.! . . . Error is Error and will show in the work with either system.!!
In some ways servo or Semi closed loop is worse than standard stepper system because you'll never know where the Error comes from as each axis have caught back up so when job finished all looks well and in position but the work piece still shows error. With steppers any lost steps can clearly be seen in the axis as it won't return back to where it started.!!

Servo's rule when heavy loads and high feeds are required Or when going point to point very fast like in positioning systems. For Small/Med routers or Small/Med mills steppers can match servos no problem without closed loop.

ya , that's the way the shims are in o.k bearings back to back and shim in between on the outer races. cheers for the run down on the closed loop . i think i'll stick with the original plan. steppers it is with the AM882 drives.

Quote:

---

Originally Posted by PAULRO

y i think i'll stick with the original plan. steppers it is with the AM882 drives.

---

Good plan but make sure you make use of the Fault output so it E-stops the system when stall detect kicks in.!

well, i've been reading most of the weekend on the electrics and control side of things and i'm still in 2 minds about my next step. i started on sourcing a power supply for the motors / drivers ( AM882) i was quoted 346 euros + vat for a 60 V DC SUPPLY so i think i will be building my own one !!!
i still have a few queries about the twin motor set up on the x axis ball screws. i'm a bit confused about the homing of each ballscrew separately. am i correct in saying that when i home the x axis, both motors engage and head for home simultaneously but because 1 motor might have lost steps during the course of machining they wont reach their respective home switch together? will the motor that has lost steps get to its home switch eventually albeit some time after the first motor and if so why do we need to home them independently.
i was reading( http://www.mycncuk.com/threads/6418-...-CNC-Mk2/page8 ) about independently homing each axis because of missed steps and the strain it buts on the gantry. so my next question is about the distance they travel away from their respective switches before they stop and return. or have i got this concept wrong also!!:confusion: if i did decide to go the 1 motor route and upgrade the motor to a nema 34 and belt drive both ballscrews would the am8882 drive and 60 VDC POWER SUPPLY need to be upgraded also? it a question of cost at the moment and that's why i'm trying to have some upgradability if i do box myself into a corner with dual motors that need regular independent homing. all these choices are really playing havoc with the hamster in the hamster wheel upstairs.:confusion:
I THINK I FOUND A ANSWER TO THE ISSUE OF HOMINGhttp://www.mycncuk.com/threads/6565-...dy-Eddy/page30 / 32 , I wouldn't mind but i read that before!!

The steppers will generally only lose steps if they're tuned too close to their maximum speed and acceleration from my understanding and reading of Jazz's posts. There's plenty of people on here using twin setups and not many posts to say they're losing steps and racking the gantry. I'm in the same situation as you regarding the twin vs single stepper and I'm leaning towards the single Nema 34 with a different driver. The AM882 will not drive the Nema 34 with any sort of speed, 240V drives are best for them but they are more expensive! I'm going to use my spare 3.1Nm Nema 23 stepper and driver for a 4th axis when I get the Nema 34....

that's good to know about the AM882 drive. originally i bought 4 nema 23 4Nm with the intention of using 1 each for the x,y,z and a spare for the fourth axis. but soon realized that i would probably need 2 for the x axis. ( 1 on each ball screw). maybe down the road i might just follow you down the road and replace the 2 23s for a 34, now that i know i will just keep going with the 23s and 60 VDC power supply. cheers for that.:thumsup:

Paulro The homing with the am882 drives is a doddle all you need is two inductive switches like this : http://www.ebay.co.uk/itm/DC-3-Wire-...-/390880485798 mounted one each side and adjusted so that they both trip at the same time when the gantry is square.
Then set homing slaved axis in Mach3 X with A (ask how to do this nearer the time) what happens then is when the home button is pressed the gantry will move to the end with the switches on and trip the switches. Then the gantry will back off the switch until it trips again. You can adjust the gantry square by adjusting one of the switches a little.
The machine will then make sure the gantry is squared every time you home the machine.

When using the AM882 drivers you wire up the stall detect (fault) terminals so that if one motor stalls it estops the router and saves any damage by racking the gantry. Hope this helps. ..Clive

Clives answered the slaving question correctly but I'll give a little more detail.!
When the motors are slaved both move together towards each home switch. The first switch to hit a switch stops just it's own motor not both and reverse's off the switch. The other motor carry's on it's merry way until it hits the home switch and does the same.
Now in practice this happens so fast it appears as one motion. This is because if the gantry is setup correctly the switches should trip exactly at same time.
If your gantry is not quite square you can tweak it square by moving the switches slightly but all the same it still appears like one motion as the difference will and should be tiny.

Regards the Motors and single or twin motors then it's personal thing and close call in some respects. If using quality Digital drives with stall detect like AM882 then I wouldn't worry about it and it's far easier to setup and less messy on wide machines.

One thing I would suggest to get the best from the motors is that you use more than 60Vdc with 80v drives and if using toridal psu you'll easily get away with running 70vdc. The extra 10v makes all the difference and worth sizing psu to maximise the motors.

Regards the electronics then don't get carried away like Eddy current did because for home use it's really not required. YES safety is important and a decent emergency stop circuit is required but you don't have to go to lengths of using expensive safety relays like Pilz. Normal Relays will work just as well if sized correctly.
More important is that e-stop or safety circuit can not be reset unless any fault is cleared and that it removes power to any devise that might do you harm.
Limits I don't really class as emergency stop condition and more positional error so can be dealt with in a differant manner. Instead of killing Power to devices they can just inform the control software limit as been tripped so can halt G-code and at same time remove siganls to the drives by using the drive enable signal.
This should be done in such away that it halts the machine and won't allow reset until limit is cleared or over ridden but at same time leaves the drives powered and holding the motors under torque.
E-stop system should done in such away that it responds to outside events like Fault signals from drives, IE Stall detect, VFD/spindle fault or emergency button press. In which case power should be removed and again system not allowed to restart unless fault cleared.

Building a control box isn't difficult and doesn't need to be complex but it does need to be safe which is easily done using relays.

well i'm glad i held off with the purchase of the 60 VDC PSU. i must get a sketch of a psu organised and post it here. i remember reading a thread posted here about the grounding issues and potential for error if the power supply is grounded incorrectly . i think it was in relation to the main screw of the coil touching or not touching the base plate . i'll have to dig it out and give it a read again. the homing issue can be put to bed for the time being at least , i 'm clear in my mind following the info from you guys here. with regards to the safety relay i think i will leave it for now and get the baby to take its first step with a decent level of safety using the relays that i have . down the road i will tackle the safety relay when i get more familiar with the whole experience. as always guys ... the advice is first class, cheers :applause:

Quote:

---

Originally Posted by PAULRO

with regards to the safety relay i think i will leave it for now and get the baby to take its first step with a decent level of safety using the relays that i have . down the road i will tackle the safety relay when i get more familiar with the whole experience.

---

That's the point you don't need a safety relay at all. You just need standard relay/s wired such that they create a latch and drops the latch if an E-stop occurs. Latch can only be reset via momentary button after fault is cleared. Very easily done and costs very little.

Quote:

---

i think it was in relation to the main screw of the coil touching or not touching the base plate . i'll have to dig it out and give it a read again.

---

Paulro I think what you are referring to is the bolt that holds the toroidal transformer through the centre to the chassis. What you must not do is connect an earth wire to the end of the bolt that is not touching the chassis. The reason of this is that it will give you a shorted turn on the transformer by connecting the bolt from the two ends to the same point ie earth. ..Clive

so i have a coil with 2 insulating rubber pads ( 1 per side of the coil) a steel dish washer to sit on top of 1 rubber pad and a bolt to secure the whole lot through the disc on to a metal plate that will be earthed anyway. so the steel disc will be earthed through the frame and if i put another earth at the top i'm creating a shorted turn. Standard Range Toroidal Transformers- CM0750225- 750VA 230v to 2x25v that is good to know , i've been searching the forum for a thread that explained the construction but couldn't find it . by the way is my understanding of how the coil is assembled correct or am i assuming the assembly incorrectly?

http://www.mycncuk.com/threads/8421-...6783#post66783

There's a diagram of the power supply....there are a few posts on the forum where people have put links to the actual items that they have used...I'll be looking for them myself again soon!

Here's a picture of how mine looks, the bridge rectifier is under those wires between the terminals and the capacitor.
Attachment 14656

And here's a typical latch circuit

Attachment 14657

thanks guys i'm about to put the first elec dwg. together and when it's finished you might critique it and let me know:dread:

Quote:

---

Originally Posted by EddyCurrent

Here's a picture of how mine looks, the bridge rectifier is under those wires between the terminals and the capacitor.
http://www.mycncuk.com/attachment.ph...id=14656&stc=1

---

Nice pic. Eddy but I only see one cap!! How big is it:barbershop_quartet_ ..Clive

Quote:

---

Originally Posted by Clive S

Nice pic. Eddy but I only see one cap!! How big is it:barbershop_quartet_ ..Clive

---

22000uF 100v

well here is the PSU and e stop circuit . i'm not sure if my calcs. are correct so i could do with a second opinion. i have calculated for 5 motors but i'm only using 4 to get the me up and playing . the 5 motor will be for the A axis further on up the road ( https://www.youtube.com/watch?v=EIwEh6ZmFFs) i have calculated for the lowest cap. and rectifer but will go with something with more head room. any way , feed back would be great , cheers in advance

Quote:

---

Originally Posted by EddyCurrent

22000uF 100v

---

Wow that's big for one cap I am using 3 to get 14100uF may I ask where you got it. ..Clive

Paulro That looks fine to me the rectifiers are usually 30 - 50 A anyway. Not sure I would switch the earth out though. ..Clive