First DIY Router design

[1Jumper10]

Looks good so far.
On my machine I filled my gantry tube with high density polyurethane foam. I tried to do a before and after comparison of how it affected vibration and the numbers I got did show it affected how the beam responded to an impact. I still haven't cut any aluminum with it so I cant say definitively that it works well for that. But, I did use Clear Path servos on it and in their auto tuning routine, they assess the stability of the mechanical system they are attached to. The parameter the servo's came up with indicated my frame was very stable. So I think I can say for sure that filling my gantry tube with foam had a positive effect on stability. The foam cost me about $40 and an hour of time. Seems like it was worth it.

[jahnj0584]

just watch out, those toggle switches do not work well with 120V. I've had two basically fry themselves to be always-on, and the third one now will usually work as normal, but sometimes stick to always on. I have to manually unplug the machine at times when done for the day.

[A_Camera]

just watch out, those toggle switches do not work well with 120V. I've had two basically fry themselves to be always-on, and the third one now will usually work as normal, but sometimes stick to always on. I have to manually unplug the machine at times when done for the day.

They should work according the rating, but of course, if you run more voltage and/or current than they are rated for you may fry them. Better quality is the best solution.

[fretman_2]

Gantry is looking awesome!

[nikdan]

Looks good so far.
On my machine I filled my gantry tube with high density polyurethane foam. I tried to do a before and after comparison of how it affected vibration and the numbers I got did show it affected how the beam responded to an impact. I still haven't cut any aluminum with it so I cant say definitively that it works well for that. But, I did use Clear Path servos on it and in their auto tuning routine, they assess the stability of the mechanical system they are attached to. The parameter the servo's came up with indicated my frame was very stable. So I think I can say for sure that filling my gantry tube with foam had a positive effect on stability. The foam cost me about $40 and an hour of time. Seems like it was worth it.

Thanks for the tip. I'm currently doing some tests of "Polymeric concrete", or basically resin with some stone/sand or other additives. This test is of Shore 80 A polyurethane which I got for free and some sand blaster sand (50% SiO2, 30% AlO3, 15%Fe2O3, median size 0.5mm).

The tests are:
- 25 w% resin. Looks to be a bit too little resin.
- 35 w% resin. All sand is soaked.
- 50 w% resin. This looks to me too much resin, all sand is clearly soaked and then some.

I will let it cure until tomorrow and take a look at it. 35 w% looks most promising out of the trials. I want the highest sand content that is still completely soaked in resin.
My target is to fill the holes of the gantry-extrusions, which is about 2 dm^3 each. I haven't calculated the weight which will be added if I do so. Estimating 3-4 kgs extra.
The resin is not vacuumed after mixing, this because I will not have the possibility to do so when filling the extrusions, so I want to see the real result.

just watch out, those toggle switches do not work well with 120V. I've had two basically fry themselves to be always-on, and the third one now will usually work as normal, but sometimes stick to always on. I have to manually unplug the machine at times when done for the day.

Thank you for the warning. The idea is to use these for 5V, 12V and 24V applications. They are all rated for 250VAC, 3A. But I will keep this in mind if I change my mind!


Thanks, fretman!

[jahnj0584]

They should work then. Im using them at 110V 10A or so which is below whats printed on them, but alas what can you expect from China. I will be sourcing higher quality ones very soon.

[nikdan]

Some updates:

This week and weekend have been slow in terms of CNC progress, haven't really done much.
The tests of polymer concrete went well. The 20w% test showed some signs of loose sand and the 35w% showed to much resin. So the final result was a 30w% mix of Polyurethane and sand.
I plugged both sides of the wholes of the extrusions with 3D-printed plates and managed to add roughly 2.5 kgs of polymer concrete to each extrusion. Resulting in 5.1kgs extra weight in total for the gantry.
The sound of the extrusions was immediately changed with the added material, a more dull sound, which I take as a nice verification of vibration absorption.

If anyone is to do this after reading this, some tips:

- Get someone to help you, a extra pair of hand is awesome!
- Use Polyurethane instead of epoxy/polyester. It has low crimpage, low exothermal properties, and vibration dampening.
- The finer grade of sand the lower percentage of resin you need. I realized that as I ran out of the first sand I tested. The second sand I used was much finer and required far less resin to soak entirely. It might have been that the finer sand had a higher density. Not sure.
The higher sand percentage the better (at least financially).
- I used Axson UR 3450 Shore A 80 Polyurethane. I wouldn't go lower than Shore A 70, but you could go higher. Maybe as high as Shore D 40 / Shore D 50.
- Try using some kind of forcing bag if you can, would save a lot of time. Took me roughly 4 hrs to do this with help.




I received my new spindle on Friday, it's a 3kW air cooled spindle from Aliexpress, from one of those who ships within the Eurozone, good for taxes. A total of $496 for Spindle, VFD (HY-220V-4kW) and a full set of ER20 collets.
I did a quick check on the runout but couldn't see any. Which is good.
I'll try to test the spindle this week to make sure that everything works A-OK.

They should work then. I'm using them at 110V 10A or so which is below whats printed on them, but alas what can you expect from China. I will be sourcing higher quality ones very soon.

Mine are rated 250VAC, 3A. So if you are having the same switches as me and pushes 110VAC, 10 amps through them, they will fry, yes.

[nikdan]

Weekly update:

Last night was my final day at the workshop. I managed to complete all essential parts for the final CNC to work and it feels awesome!
- I added an 8.5mm stainless steel plate on the back of the Z-plate to add stiffness. The maximum thickness I could use before it touched the rails.
- Completed the spindle plate
- Added a mounting for the cooling on the spindle extrusion.


The stainless plate might look unnecessary but the steel was cheap and I had the time. So why not?
Steel is roughly 3 times as stiff and any stiffness to this long plate is probably good. I chose stainless so I don't have to worry about the corrosion.



The spindle plate has a cutout for the bearing mounts and ball screw mounts. This pushed the plate 6 mm close to the Z plate to reduce the moment arm.
What I realized just before doing the holes was that the drawing of the spindle was not the same as the real thing.
CC of the hole-rows was 0.7 mm off, height ~6mm off and width 1.8 mm off. I was only lucky I waited with this plate. This spindle was bought from "CNC star" on Aliexpress (Most likely the same seller as 'POWACE cnc', 'Changzhou Rattm Motor' and 'Rattm motor store' on Aliexpress).

Now let's just hope that the rest of the Chinese parts will actually fit...!





I have been thinking of where I should put the cooling tubes, and while inspecting the spindle I realized that the square aluminium is only a heat-sink which meant that I easily could mount the cooling control block there without damaging anything.
The cooling system was bought from China. $14 with shipping. The problem with the very cheap cooling system is that within the flexible hose there was a stiff plastic tube for the air flow, and within this, there was another plastic tube. The result was that the flexible hose could not withstand the spring force of the plastic tubes within and made it impossible to control where the hose was pointing. I guess you get what you pay for...
A small operation and removal of the middle tube and shortening of the flexible hose made work far better, as long as the flexible hose can hold the pressured air enclosed. We'll see!





This week I hope to do a mockup and assemble all parts I currently have. It will be interesting to see the full size of this thing!

[nikdan]

Quick update.

Not much has been done since last time. I did the mockup. Felt good to see that all parts did fit where it was supposed to and nothing clashes... so far.
A few pictures below of the mockup:









I been thinking about the rails and are still not convinced what size. My design was 15mm rails but I am reconsidering it for the Y-axis. A quick fix would be to make an adapter-plate until I can mill new ones in the CNC itself.
Screws will probably be changed to 10mm pitch.

Currently, the move is taking all my time so there won't be to much action until that is done. I will most likely buy the motors, drivers and such first and finish that before going for the rails.
I will most likely be working on that during November. A first run before the end of the year will be my target!

I'll update as soon as something changes.

[jahnj0584]

Wicked man. didn't realize the square spindles were basically just aluminum. Do you have a tramming system or idea in place for the spindle?

Id go with 20mm HGH rails. Automation overstock has great rails for a great price available to Hiwin specs!

I'd also do 2010 screws. 05 lead really isnt worth the extra accuracy unless you dont really want to go over 200ipm anways.

[NIC 77]

Nice work so far!

You are flying along.

Are you going to add some ribs inside your gantry box? I would. Also was the epoxy expensive? I'd also consider filling up the gantry with it, not just the T slot (after adding ribs). That or high density Poly Urethane, which would be much lighter.

Of course, you can build it, see how it works, and then decide if want some extra filling, but I think the ribs are a good idea as a minimum.

For the ballscrews, IMO, the choice for X and Y is between 10mm lead and 20mm lead. 5mm is not a good choice. Did you mention what kinds of speeds you want to achieve? Do you have a graph of the torque vs speed curve for the motors?

For the Z, I'm thinking 10mm. How much will all of the Z parts that move up and down weigh?

Great work. I think you will have a nice machine when it's done.

[nikdan]

Well, not much to update. Been busy with the move, new city, new job, new everything. The move with all new funiture and all was more expensive then I thought so I will have to postpone the remaining mechanical parts for the beginning of 2018.
However, I just bought the all electrical parts, motors, BoB, motion controllers, the whole shabang. So I have some tinkering to do until christmas at least

It will be based on the Pro stepper kit, 4 axis, 4Nm nema23 and the UC300eth-5LPT. 3 proximity sensors, 6 limit switches.

First up will be mounting it all inside the electronics cabinet. I'll update when I have something to show!

[nikdan]

Quick update.
Been working on my Electronics cabinet now for a week or two. I thought it would take a quick days work... not so much...

I got all nice parts from cnc4you as well as the UC300-eth directly from cnc drive. Will try to fire it all up when the box is ready.
The spindle took a nice spin the other day, it purred like a cat


When your cutting fluid is out, just borrow your girlfriends Le Creuset - olive oil sprayer and you are good to go! (Just dont tell her you did)



This is some of the progress. I have two 80mm fans with filters blowing inwards at the bottom of the sides. 6 - GX20 for steppers (X,Y,Y1,Z,A) and 1 extra. 3 - GX12 for emergency stop, limitswitches and home switches. 1 - GX30 for the spindle.
On the front is 3x 16mm steel buttons, 3x switches, an emergency stop and a 15A (Max) analog ammeter. My plan is also to lift the Control panel for the VFD out on the front. (Silver tape marks when I'm not allowed to drill).


I hope to be done with the electronic layout this week, some paint to make it look pretty and doing cable work next week.
Allways a nice feeling to see some progress

[jahnj0584]

nice. Wheres the enclosure from?

[nikdan]

nice. Wheres the enclosure from?

I bought it from eBay, roughly $60 with shipping. I wish it were 5 cm wider tho' would have helped a lot!

[nikdan]

Hi everyone, back again for some updates. I've been quite busy lately both with life and the CNC.
I did not anticipate that the control box would take as much time as it has. I've been working a lot with it and it's still not 100% done.
The biggest issue is, of course, my inexperience in electronics and with that, I do a lot of mistakes and seem to always miss some stuff to do all connections properly.

However, since last time I have finalized all holes on the box, painted it industrial grey with Hammerite paint and managed to fill the box (after a lot of pondering...) with all the electronic equipment I believe I need.
One major thing that took time was to enable the VFD to be installed into the box, the box is just a bit small which led me to sacrifice aesthetics and just go for functionality. I'm quite confident that it will bite me in the ass, but that's life! Just a tad bigger would have helped me a lot!

But, everything works!!

Power supplies, check!
VFD, check!
Amp-meter, check!
LEDs, buttons, switches, homing switches, limit switches, Check!!

Last test was to set up the HY VFD for my spindle which gave me some trouble. But It's fixed and works great.
All cables for the steppers are done and the next step is to try to connect to the UC300ETH. I'm currently waiting for the software license and when that's fixed I will start to move my steppers to see that everything is a-ok!



I also made a small adjustment to the spindle heatsink and installed a thermometer I had extra. A good way to see that I'm in decent temperature range for the spindle. Will mostly be used in the beginning I suppose.


Next up will be software related stuff which there is not much to update on. I will probably work with Fusion 360 CAM software as a start, combined with the UCCNC.
I hope to order my rails and screws in January so I just might start assembly as soon as I know the electronics are working properly. Mechanical is so much more fun!

[jahnj0584]

looks good man. Control boxes always take a long time. Its funny but a too small box basically requires "unclean" wiring practices for everything to actually fit. Is the case itself grounded? I noticed you used power line filters from the get go.

I also looked at temp probes but too much effort with another PSU or converter and all that. I just added a single fan and can hand test the exhaust air. Its never very warm if it all.

[nikdan]

Thanks, Jahnj,

Yes, the box itself is grounded, so no worries The worst part about wiring for me was that I had to use tweezers to get everything in place!
The kit I bought recommended two EMI filters, especially 1 to shield off the VFD, but I thought I'd use two for the sake of it. It was included in their recommended wiring scheme and as my knowledge in electronics, I thought I'd just go with what they say, they were $8 a piece.

I used a digital temperature sensor that is running on batteries, cost me $1.2 with shipping from Aliexpress. Made it all a whole lot easier

I've now managed to install my UCCNC with the license and after a lot of scratching my head the steppers now go the correct way and does what I tell them to, they acted a bit strange in the beginning. As you can see in this video they were very aggressive one direction.


I've also installed the VFD in UCCNC which makes it possible to control it directly through the software, by a push of a button or by an M3 command in the program. Quite neat. Still not working 100% tho'. If I set the spindle speed to 10000, the spindle varies in step between 9700 and 10300. So nothing in between. It just varies between these exact two values. I find it quite strange but I think this is due to a badly shielded breakout board or fluctuating voltages. It seems to vary depending on the load of the steppers.
I also don't get any feedback in the software of current spindle speed, I can only see that on the VFD's display. Will do some troubleshooting this weekend.

What I didn't expect installing the motors and spindle was that it was so quick. It's not complete of course, but most of the tuning I need to do when the machine is assembled, so I'll now order the rails and screws and start the assembly!
So if their shipping times are correct it should all arrive mid-January!

[ger21]

Quite neat. Still not working 100% tho'. If I set the spindle speed to 10000, the spindle varies in step between 9700 and 10300. So nothing in between. It just varies between these exact two values. I find it quite strange but I think this is due to a badly shielded breakout board or fluctuating voltages

I had a similar issue when using analog speed control with my Huanyang VFD. RPM would fluctuate by ±1000 rpm or so. Adding capacitors to the voltage input made it better, but it still varied quite a bit. The issue is apparently due to noise somewhere, as my analog voltage was very steady.
I ended up switching to a Huanyang plugin for UCCNC, which gives accurate speeds, and is very simple to use.

[nikdan]

Thanks for your help on the other threads Ger,

I saw that thing on your youtube channel, do you need to have it connected to your laptop directly? I would prefer if I could be in another room and running my CNC though wifi or similar.

My analog voltage is far from accurate, varies up to .8V as I run the program. My issue is probably the BoB, I'll try to solve it, but most likely I'm going to try with some capacitors!

Apart from that, the only electrical and software issue I have is that the limit switches do not activate when running a homing sequence. I have 3 proximity switches for homing and 6 mechanical limit switches at all end of each axis. I basically want to use my limit switches as E-stops, but with the possibility of jogging away. Not really sure how to add this since I can't find anything in the settings, perhaps a macro? Do you have anything like this for your Screenset, Ger21?
Other than that, since I have all my homing switches in series I've also installed a macro for UCCNC that jogs of the homing switch before the next axis starts to move.