[ger21]

I was going to go the rack and pinion route but then I saw your concept and figured since I had the belts and pulleys anyway, why not give it a shot. I'm going to run my parts on the waterjet at work but wanted to get everyone's feedback on the design to see if there is anything I've overlooked.

Feel free to point out my mistakes .....................

Imo, to use this system with steppers, you really need to use some gear reduction to get decent resolution. I've got a 4:1 reduction on the one I'm building.

[zool]

If you look around it would appear that mixing different sizes and types of motors on the same machine is taboo. Regardless whether the axis is a heavy X gantry moving continuously or just the Z raising and lowering the spindle vertically an inch or so between cuts, on most machines everyone specs the same motor and linear mechanism on every axis. My thinking is that this is odd and financially wasteful. Every axis has different power and speed needs. In other words my penny pinching butt just can't comprehend putting a 400 dollar servo on perhaps a plasma cutter Z axis just because its also on the 22 foot long X axis.

On the Z axis for plasma and router cutting you can get by just fine with a cheap little stepper motor direct driving a screw. anything more is just vanity in my opinion. The weight of the axis virtually feeds itself. On the Y axis a direct drive screw and stepper again performs well at only 5 feet long. Where the ServoBelt really shines is on the gantry rails moving that long heavy X axis with lots of inertia. Now, If I could only buy 2 servos then that's where id put them - on the X. Then by using 2 inexpensive steppers for Y and Z it's possible to power this all up for under $1000 with simple direct drive mechanics on every axis and negligible compromise in performance. This is my current thinking.

The reasoning for one type of motor and one type of rail is: "SPARE PARTS".

Spare parts have to exist, and have to be inventoried, provided secure storage space [so they don't sprout wings], and accounted for, not to mention the opportunity costs of a break down [total cost of production stopping].

When the administrative costs, plus the potential loss of revenue are taken into account, the minor additional cost of a "too large" motor on the Z-axis as a percentage of the total cost of the machine is infinitesimally small. Same reasoning for linear mechanisms.

Best summary: Penny wise; pound foolish.

[Ken_Shea]

Gerry,
You have done a superb job on that set up.

Cant quite make out the GT Belt specifications and what is your upper tooth count or diameter?


Ken

[Ken_Shea]

Mike,
Your design has prompted a lot of interest, to share your design and help to the DIY group is quite generous of you.

Thanks
Ken

[ger21]

Gerry,
You have done a superb job on that set up.

Cant quite make out the GT Belt specifications and what is your upper tooth count or diameter?


Ken

Belt is 9mm wide 3mm pitch.
17 tooth and 68 tooth.

The T5 drive pulley is 14 tooth, for a 25mm belt.

[ger21]

Made some Delrin rollers, and have it almost complete. Not sure if I need it, but I'm going to add a frame on the outside of the belt to support the pulley and shoulder bolts against the belt tension. I'm concerned about the side load on the shoulder bolts?

Right now I've got about $170 into this. Add another $50-$75 for the T5 belt, and it's about $225-$250/axis. Not a whole lot more than an acme screw and AB nut.

[JD68]

I'm concerned about the side load on the shoulder bolts?

.

Gerry nice job!

As a first step how about counter boring the shoulder bolts? Maybe something like half the diameter of the shoulder screw? May be easier than adding the plate you mentioned?

JD

[ger21]

I don't have the capability to pull that off. I made this with my router, which isn't that precise.

I can make a bracket to support the shoulder bolts pretty easily.
But do I need to support the pulley shaft? It's a piece of hardened 8mm drill rod. I wonder if it'll flex at all under load? There's a bearing right behind the pulley and another about 1-1/2" behind it.

[diyengineer]

Made some Delrin rollers, and have it almost complete. Not sure if I need it, but I'm going to add a frame on the outside of the belt to support the pulley and shoulder bolts against the belt tension. I'm concerned about the side load on the shoulder bolts?

Right now I've got about $170 into this. Add another $50-$75 for the T5 belt, and it's about $225-$250/axis. Not a whole lot more than an acme screw and AB nut.

that loooks good! I think it will be more then strong enough!

[zool]

Gerry:

Absolutely brilliant!!!

You've really out done yourself this time. Of everything I've seen on this great site, your invention [and inventiveness] is the best yet.

W.

[ger21]

It's Mike's invention, all the credit goes to him. :cheers:

I just added a belt reduction to his design to work with a stepper, rather than a servo. It did take me quite a while to figure out how to do it, though.

[mcphill]

Sorry if this is posted somewhere in the thread, ger21, but I could not find it... What belt are you planning to use with your setup (brand and part number) for the main drive belt? Are you staying with the "belt on belt" rack system?

[ger21]

Mike uses 32mm T5, but I'll be using 25mm T5. I couldn't find an inexpensive online store that I could get 32mm pulleys.

I did find a source yesterday for inexpensive belts,
T5 pitch Polyurethane Belt

Not sure if I should get the kevlar, but I'm guessing it's more flexible than steel cords.

[mcphill]

Thanks for the info and link. I bet the Kevlar is more flexible, and less stretch... I would go that way...

[diyengineer]

Mike uses 32mm T5, but I'll be using 25mm T5. I couldn't find an inexpensive online store that I could get 32mm pulleys.

I did find a source yesterday for inexpensive belts,
T5 pitch Polyurethane Belt

Not sure if I should get the kevlar, but I'm guessing it's more flexible than steel cords.

I went with Kevlar on my belt drive. It was roughly the same price as other alternatives, and doesn't really stretch. Super strong.

[ger21]

Just ordered the Kevlar belt.

I re-read the whole thread this morning, and saw Mike say that all the loads are transferred to the shoulder bolts (in my design), so possibly 100 lbs on the two inside shoulder bolts. That concerns me a bit, so I came up with this.
A bracket that ties the two inside bolts together, with two tapped holes to mount another bracket. The second bracket will house a bearing to support the drive pulley. This needs to be in 2 pieces, because the pulley has to move to tension the belt. I'll probably end up modifying this before I make it, as I don't want to interfere with the 6 screws that I need access to when I tension the belt.

I also read that there may be some backlash between the belt and pulley, but it can be reduced (or removed) by turning down the OD of the pulley so the belt seats into the grooves more. I'll check when the belt gets here, and do this if needed.

[zool]

Instead of two pieces, why not make just one, and pocket the areas need for clearance on the inside next to the pulley.

This would spread the force over a larger area as well as over the four bolts.

Instead of the hole and slots as on the orange plate, make the slot where the pulley attaches so the pulley could be adjusted that way. Adjusting with one bolt is a lot easier than adjusting with two bolts.

Also: Fewer parts = fewer problems.

[HomerSimpson]

ger21,

I am concerned about the fact that I don't have a gear reduction. I started to design my parts to accommodate it if I needed it but it got too complicated. I'm using the G540 with 10X microstepping which gives me around 0.001" per step--not great but it is just wood I'm machining and that seemed like a tight enough tolerance. I've noticed that the guy who runs cncrouterparts.com does a 3X reduction on his rack and pinion system so I may end up regretting this. Oh well--more time in the shop. There are worse ways to spend my time.

[ger21]

Depending on who you ask, you may or may not be able to rely on microstepping for resolution. Even though I'll be using a G201X with 10x microstepping as well, I tried to design to achieve .002" or better with 1/2 stepping, which is said to be much more achievable.

Another important consideration is how much force you'll get, especially at higher rpm's. Off the top of my head, I'd guess that a direct drive nema23 would only give you about 10-20 lbs of force once you got over 500 rpm?? But maybe with direct drive, you'll never spin it that fast?

[HomerSimpson]

At 500in/min the motor is only rotating at 200rpm and this stepper has more torque capacity at lower speeds. I don't think that will be a problem. The only thing I'm worried about is resolution. Why do you say that you can't rely on microstepping to maintain resolution?

I'll keep you posted on how it works out and you do the same. I'm interested to see how your gear reduction system works out.

I'm using 3mm HTD 9mm wide belts, which I thought I read that you are doing the same. I also bought my stuff at econobelt.com after much searching.