[grg_p]

Digging through my junk bins I found 72 tooth timing gear that will fit my ball screw. I think that with 5mm pitch and 5000 rpm, 22 tooth servo sprocket I’m hoping to get to 300 in/m. If my math is correct.

[joeavaerage]

Hi,
why bother having a reducing gear at all? You've got plenty of torque already and more than fast enough already. Just because you servos can do 5000rpm does not mean to say you have to run them at that.
If you programmed your controller to 1500rpm and direct coupled to the screw that would be 7500mm/min or 300in /min. Your servos have already got about 3.8Nm torque, rather more than my servos (2.4Nm)
and they are faster than I could possibly want, and I would suggest faster than you could possibly want as well, and your servos have got more torque than mine, so that would make your machine ridiculously fast.
You don't seem to realise that your servos (2.7hp) are in fact way too big for your machine. I have 1hp servos and they are marginally too big as well. 1/2 hp or 400W would in fact be enough.

Craig

[grg_p]

When I started this project 8 years ago I wanted x servo out the back instead of the front. I had already started my enclosure and don’t have room to run servo straight out the back. The servo’s are over kill but I got then for free so I’m going to use what I have and try to make it work. I don’t have drawings or someone’s work to copy so I know what I’m doing is wrong and I didn’t want to offend anybody. Sorry for this bad build. You’re right this probably won’t work so all I can say is I’m sorry. I won’t bore anyone with this anymore.

[joeavaerage]

Hi,
no need to be sorry, there is nothing wrong with what you are doing.

Your servos are outsize, your machine will be lucky to absorb 25% of the power they have....so dont sweat it. You do not need any gear or belt reduction, 1:1 would be fine.

Craig

[Titaniumboy]

Hi,
I use 1hp (750W) servos, rated at 3000 rpm, but can go 5000rpm.

How does that work? I have a servo spindle rated for 3000 rpm, but would love to run it faster if possible.

[joeavaerage]

Hi,
my Delta servos are rated to 3000rpm but their max is 5000rpm. The rated torque is 2.4Nm, and that is available right up to and including 3000rpm, but linearly reduces beyond
3000rpm. The power remains constant but the torque decreases.

3000rpm torque=2.4NM
5000rpm torque =2.4 (3000/5000)
=1.44Nm

This utilises a feature of AC servos called 'Field Weakening'. It is a very clever way to reduce the back emf of a motor and therefore get it to spin faster.

You may have seen advertised that a machine has a G0 or rapid traverse speed of 25m/min say, but a max G1 speed of 15m/min. This would suggest the manufacturer is taking advantage of Field Weakening.
He can certainly use it to make the rapids faster, but at somewhat reduced torque. When cutting, ie G1, however full torque is required, so the max G1 speed is usually rated speed at rated torque.

I have used this in my machine. The servos are direct coupled to 5mm pitch screws. At 5000rpm the machine goes 25/min, while at rated speed ie 3000rpm it goes 15m/min.
As it turns out 25m/min is too fast for my nerves so I deliberately de-tune them to 15m/min, and for most work I run the Rapid Rate Override at 50%, so down to 7.5m/min.
There is such a thing as 'too much of a good thing'.

Craig

[phomann]

Hi Craig,
Funny you talk about nerves. My little 7x14 lathe can rapid at 6m/min.
Watching the cross slide scream towards the chuck and stop with the parting blade 3mm from the spinning chuck tests the nerves.
I slowed the rapids on the Z axis to 2.5m/min.

Cheers

Peter


Sent from my iPhone using Tapatalk Pro

[Titaniumboy]

Thanks for your reply Joe. Rather than me hijacking this thread any more than I already have, where would be a good subforum to further discuss the idea of speeding up my servo spindle motor?

[joeavaerage]

Hi,
just start a new thread.

Craig

[ninefinger]

what kind of resolution do the encoders on the servos provide? The biggest remaining reason to gear down is to effectively multiply that resolution if needed, but somehow I doubt that is the case.
With a 5mm pitch - even 1000 steps per rev would be more than enough I think.

[joeavaerage]

Hi,
it very much depends on the servo.

Even el-cheapo entry level Chinese made servos have an encoder resolution (typically) of 10,000 count per rev. Twenty to thirty tears ago 10,000 count/rev would have been
considered state of the art, but these dayz 1,000,000 count/rev is common, and 10,000,000 count/rev is the new state of the art.

I use Delta B2 series servos that have an incremental 160,000 count per rev encoder. These are Delta's entry level servos these days. Like most modern AC servos they have
a feature called 'electronic gearing' which much as it sounds allows you to tailor the resolution to your requirement. For instance I have my servos direct coupled to 5mm pitch screws
and the electronic gearing set to 5000 counts per/rev. That means each Step from my controller rotates the servo 1/5000th of a rev and the axis advances 1um. 1um is plenty fine enough for
me.

Unless you have a very old school servo with a low resolution encoder then there is no need to gear the servo to gain resolution, all modern encoders have more than enough resolution without
that. You may of course be gearing to gain torque...and that is fine, but there should be no need for resolution.

Craig

[grg_p]

Ended up with 44 tooth and a 72 tooth for my x axis. 22 tooth was to small.

[binfordw]

Good to see you still posting.

How are you handling the belt tension?

[grg_p]

Belt tension was good so I think I can get away with out tensioner. I originally was going to do one but now I don’t see the need.

[binfordw]

Will be interesting to see how that works out.

[grg_p]

My enclosure which I had started before I moved down to my new work shop was going to set on concrete risers that I had made. But now I’m going to set it on I beam legs of my hoist. But when I set it on the beam it was too high. I would need steps for it. So I lowered it 10” by cutting notches in beam.

[grg_p]

Want to get Z ball screw as close as possible since base bearing and column bearing are not tided together. I will land top bearing when it’s setting on stand and can get it close as possible. It’s a lot easier to do mockup on the ground than in the enclosure. I had to anneal the ball screw ends to turn them, and the one end that came annealed from supplier was little short of what I needed. Went thru a few carbide bits getting it done. The bearing tightening nuts were an odd metric size so I had to order a die for them as well.
I’m getting ready to put base on stand and trying to get as much done on the ground as I possibly can. I ordered Masso touch screen and it is on its way. Need to start looking at wiring and servo cables and how drives match up with the servo’s. My overhead crane has been a life saver.

[grg_p]

Finally getting my mill base on my stand. It’s been a long time coming.

[grg_p]

I received my Masso touch screen and had to modify my mount. I originally was going to use computer tower and monitor with mach3. The screen mount was setup for old computer monitor. And it didn’t work with Masso screen. So I made new mount using old computer screen and stand. The original was too short. The new one is taller and pivots, slides up a down, also tilts. I like this one a lot better.

[grg_p]

My Y ball screw mount was going to hitting table so instead of grinding material off to make room for ball screw I made aluminum spacers to give me some clearance.