[tahustvedt]

I decided to try a gearbox so I made this real quick. The housing is made from 10mm aluminum, 6mm aluminum and 22mm MDF. I'm using two 10x19 mm ball bearings on the shaft. The pulleys have 20 and 45 teeth, because that's what I had.
Attachment 222746

[tahustvedt]

I just installed it for testing, and that really did the trick! It's able to accelerate ridiculously fast now. It will accelerate up to 2000 mm/s/s, and at that rate it tosses the housing off the magnets. 1500 mm/s/s works without releasing the magnets on the housing, but it's way more than I need. I will set it to 700 for now and have lots of headroom. Maybe increase to 1000+ as I test more.

Attachment 222782
Attachment 222784
Attachment 222786

The blocks are there to act as replacements for the melamine sheet I normally use. I need to route a hole for the stepper before I can mount it. The plastic pulley is temporary, because it's the only GT pulley I had that would fit with the belts I had available. I will make new aluminum pulleys for a wider belt I have later.

My gantry weighs ~55 kg and the housing ~15 kg.

[cpr]

Am still inspired by this, keep the updates coming!

(for what's it worth: I have some looong makerslide coming... Longer than it can be reasonably installed.)

[tahustvedt]

Here's what the finished installation looks like. There's an opening under the belt drive assembly to allow some air to flow through the white plate. Attachment 223174

The router is almost finished now. I just have to make a door for the backside of the gantry housing, and add some rubber flaps under the gantry and housing to keep chips from flying out when I cut aluminum.

[ger21]

Nice job. All your work is really clean, well designed.

[tahustvedt]

Thanks, Gerry.

[tahustvedt]

I'm going to try making new pulleys before I order from SDP-SI.

Made this today Looks odd. Teeth are wide compared to the 44T pulley, but both at from cad drawings of GT2 pulleys. I also made a 44t pulley, but managed to mess it up when I bored the shaft hole in my lathe. :P I'm making a new 44T pulley now.
Attachment 223504

Gotta order a new belt from SDP-SI if the pulleys turn out ok. Maybe I'll get the SDP-SI pulleys as well, just in case.

[tahustvedt]

Here's the finished 44t pulley. It has what to me looks like normal GT2 tooth profiles.

[tahustvedt]

Well, they work. I tensioned the belt pretty hard. I measure 0,07 mm total X movement of the gantry when I pull on the spindle with a force of 10 kgf im both directions, and it flexes back to the same middle position when I release. I think the added backlash of the belt drive is negligible.

[tahustvedt]

Made a new 22t pulley from a new tooth profile drawing. This one looks normal.

[tahustvedt]

The router is just about finished now (as if a DIY router can ever be finished). The only thing that remains is to install a handle on the door on the back of the housing, and a stronger gas spring. I added curtains under the housing and gantry to keep chips from flying out, and to keep coolant fumes inside. The hose on top of the housing is for extracting coolant fumes, and the one on the side is for dust extraction. It's all working great. I think I'll make something now.

Attachment 223954

Attachment 223956

[tahustvedt]

Oops. I stalled the spindle when I tried to pocket at 15000 rpm with a 22mm bit, and the router bent the 6mm shaft as it tried to chisel oak. :P I hope the spindle shaft isn't bent, but I think it's time to upgrade to a HF spindle anyway, because the machine is obviously limited by the Kress. I'll have to retram now.




Update:
It didn't go out of tram. The spindle shaft doesn't seem to be bent, though I measured 0,03mm runout with a cleaned shaft and collet, which is a little disappointing. :P Maybe it's not bad for a Kress.

[tahustvedt]

I made a hinged support for the long X ball screw from 10 mm acetal. I also made a bush to support the screw where it exits the gantry beam on the opposite end of the nut. I could have used another ball nut, which would give even better support, but this'll do. I can run 8000 mm/min now without whipping causing problems. It's whipping slightly in the exteme ends, but not badly. Without the supports it would only reach 6000 mm/min. Still, I'm running a max of 6700 mm/min to keep the kernel speed low. 6700 is pretty fast, and a great improvement over 5500, which is what I ran before. I'm a big fan of acetal. It's so easy to work with, and prototyping stuff takes minutes, even with tiny tools.

A video of the gantry moving at 6700 mm/min:


I have a new plan for the housing. I'm not entirely happy with the magnets I'm using. They work fine for the most part, but I fear that when doing quick and detailed work they can cause problems for the X-axis accuracy, as they will not hold the housing tightly to the gantry. I will add two 180 N electromagnets to secure the housing to the gantry, and trigger them with a switch, or maybe the spindle output. I already have 12V running up to the housing for the LED lights inside.

[tahustvedt]

I made a new pinion for my belt drive today. I wanted to try a lower ratio. 1:1.4666........ this time. Seems like it's a smidge stronger at speed than it was at 1:2, if there was any difference at all. Maybe the sweet spot is somewhere in between. I'll have to do some testing to see if it will handle small sudden acceleration jobs as well as 1:2 did. It's difficult to quantify stepper performance.

[wizard]

I made a hinged support for the long X ball screw from 10 mm acetal. I also made a bush to support the screw where it exits the gantry beam on the opposite end of the nut. I could have used another ball nut, which would give even better support, but this'll do. I can run 8000 mm/min now without whipping causing problems. It's whipping slightly in the exteme ends, but not badly. Without the supports it would only reach 6000 mm/min. Still, I'm running a max of 6700 mm/min to keep the kernel speed low. 6700 is pretty fast, and a great improvement over 5500, which is what I ran before. I'm a big fan of acetal. It's so easy to work with, and prototyping stuff takes minutes, even with tiny tools.

Really nice with the lead screw support! The first thing I thought of was those spring centered hinges used on commercial doors, like at restaurants. Those sorts of hinges probably cost more than your lead screw and nut assembly though.

I have a new plan for the housing. I'm not entirely happy with the magnets I'm using. They work fine for the most part, but I fear that when doing quick and detailed work they can cause problems for the X-axis accuracy, as they will not hold the housing tightly to the gantry. I will add two 180 N electromagnets to secure the housing to the gantry, and trigger them with a switch, or maybe the spindle output. I already have 12V running up to the housing for the LED lights inside.

It might help to sit the housing on alignment pins. I'm not sure what you are doing here as I've only visited the thread from time to time. In any event if the goal is a quickly removable cover, mechanical latches are hard to beat.

By the way this machine looks awesome.

[tahustvedt]

I had a mechanical latch on my old router. I simply lifted the rear of the housing a fe mm up and it would dropp into place when it reached its seat. It's not practical on the new machine. I need to either make a mechanism that operates both sides, or operate separate mecanical latches on both sides of the housing, and I can't reach the inside without moving the machine out from the wall. One option is to use solenoids to operate the latches. I'll test the electromagnets to see if they are strong enough, and don't become hot, before I make holes in the housing to install them. I plan to operate the solenoids or electromagnets with a microswitch on the housing door, so that when the door closes, the coils are activated. That way I don't ever have to think about it.

[tahustvedt]

I bought this VFD a few days ago. It should be here in a week. It says it supports standard MODBUS protocol, and 10V control. I can't wait to play with it. Is MODBUS the recommended option, or should I just use the analog control output? I'll have to get an RS485 adapter if I want to use MODBUS.

Variable Frequency Drive VFD Inverter 2 2KW 2HP 220V 110VAC | eBay

[ger21]

If it's a HuanYang VFD, then this is probably the way to go.

A new Huanyang VFD plugin

[tahustvedt]

Yeah, That's a third possibility. I've done hours of reading about the plugin, but the one I bought seems to actually support MODBUS, at least according to the listing. It also has a pot on the control panel, which other HuanYang VFDs don't have. I guess I have to disconnect the pot if I want to use the 10V control option. Hopefully it will be clearer once I receive it. I guess I should focus on making a spindle mount instead of worrying about VFD control.

[tahustvedt]

I've started cutting the mounts. This one is cut from 1/2" 6061-T6. I will mill clamp screw holes holes before I split the halves at the lines. I sure hope the spindle really is 80 mm as it says in the specs. :P The other onw will be cut from 10mm or 15mm aluminum. The 15mm flat bar I have is bowed pretty badly, so i don't really want to use it for something this big because I'll have to plane it first.

Attachment 224866