L3viathan's Steel Router

[eaglewolf]

[l3viathan]

Hah, I had to watch that for a while to figure out how it was relevant. I'm not sure if they'll be good enough , but i'm looking at either Vexta PK299-F4.5A or PK299-03AA (same motors, different wiring), and at least one of them double shaft. If the detent torque isn't enough to stop the Z from dropping when not under power, I'll have the option of adding a brake.

I'm not particularly impressed with the torque curve on the Vexta motors, but the Wantai and Long motors on Alibaba are lacking any torque curve info, and even the data provided is a bit inconsistent between sellers. I guess it's a case of, "better the devil you know".

[eaglewolf]

The problem with most companies is that they only show the torque curves, but not what controller/driver was used nor the voltage input so its like the torque curve is WORTHLESS unless you can match what their curves are.

Without all the specs, you are left scratching your head wondering if it would even work????

[ger21]

I'm not particularly impressed with the torque curve on the Vexta motors

With 75Volts, it's showing about 300oz of torque at around 1100rpm, which is imo, pretty good for a motor that size, and only being 4.5amps.

Vexta's are probably much better motors than most of the chinese motors you'll find everywhere else, but they often aren't available with specs that easily compare to the chinese models.

[l3viathan]

With 75Volts, it's showing about 300oz of torque at around 1100rpm...

That is something I was curious about. If the general rule is 32 * sqrt(inductance), that would put these motors at max ~50v. I imagine that 75v could result in overheating issues assuming the same current?

[ger21]

If the general rule is 32 * sqrt(inductance)

This is a rule that Gecko came up with. I believe that some drives are better than others at reducing motor heating.
Looking at the 48V curve, it drops the rpm by about 200-300? Still not too bad.

[l3viathan]

I got a vague response from Oriental Motor tech support regarding the use of 75v with their PK299. They have no information about the duty cycle of the motor at that voltage, and only recommend that it is kept below 100° C. I'm not a gambling man, so I think that is a non-starter. I'd prefer at least some safety margin, so unless someone has evidence that this configuration is stable without active cooling, i'll pass.

[eaglewolf]

In my experience yes the motors will get quite hot to the touch, but you have the advantage of an attached stage to assist in cooling and the hottest axis will be the one doing the most work the most amount of time

although I am Leary of 75V I also have enough experience to say that you should be safe enough at 75V unless you are running 24-7 or even 24-8 even :-)

[l3viathan]

I'm working on the design for my Z axis, to make sure I have the travel I want. This should give me a full 7" of clearance (same as the gantry) with about 50mm of tool length. I should also be able to touch the table with a 10mm tool length. This is as low profile as I can get in regards to the distance from the center of spindle to gantry face, it's about 7 inches with an 80mm spindle. I've seen this axis done the opposite way, but this gives me less moving mass, and more clearance around the spindle.

[eaglewolf]

hmmmm
small mock-up with a spindle-bit?

[louieatienza]

I got a vague response from Oriental Motor tech support regarding the use of 75v with their PK299. They have no information about the duty cycle of the motor at that voltage, and only recommend that it is kept below 100° C. I'm not a gambling man, so I think that is a non-starter. I'd prefer at least some safety margin, so unless someone has evidence that this configuration is stable without active cooling, i'll pass.

If the motor is wired bipolar series then you shouldn't have too much trouble with heat. Over-currenting would do more harm heat-wise. Usually with higher-end steppers like Vexta, Sanyo-Denki, Kollmorgen, Parker, Lin, they use insulators rated to over 300-500V. I run a small mill with small 320 in-oz steppers and use drives that pump out about 177VDC and I have rapids set at 200IPM, using 4mm ballscrews, and I'm not even pushing the machine hard at all. I even had a small desktop router that had 180in-oz steppers and ran them at 48VDC. And they made a 320V drive compatible for them too! I run my steppers bipolar parallel which makes them run a bit warmer, but not crazy hot.

[l3viathan]

If the motor is wired bipolar series then you shouldn't have too much trouble with heat. Over-currenting would do more harm heat-wise. Usually with higher-end steppers like Vexta, Sanyo-Denki, Kollmorgen, Parker, Lin, they use insulators rated to over 300-500V. I run a small mill with small 320 in-oz steppers and use drives that pump out about 177VDC and I have rapids set at 200IPM, using 4mm ballscrews, and I'm not even pushing the machine hard at all. I even had a small desktop router that had 180in-oz steppers and ran them at 48VDC. And they made a 320V drive compatible for them too! I run my steppers bipolar parallel which makes them run a bit warmer, but not crazy hot.

In bipolar series, the inductance is usually much higher, which requires a higher voltage. The motors i'm looking at are 2.5mH in parallel/unipolar and 10mH in series. If the Gecko formula holds true, in series, the motors should run on 100v without issue. I looked at the Kollmorgen steppers, and the 320v torque curves are scary, they are practically a servo at that point (of course the price point is commensurate). What model steppers are you running at 177v in parallel?

[yellow68gto]

I am currently designing a similar machine and want to compare your z axis design with mine

Unfortunetly I have no rendering software. Just drawing on paper currently.

On my Z I am mounting my rails to the back of the spindle plate and my bearing on the gantry "Y" side of the plate.

The ball screw but is then mounted to the gantry "Y" plate
. I am using 40mm rexroth rails and a 40x5mm ballscrew.

I think my weight on Z is going to be considerably heavier then yours. Planning on a skyfire cnc bt30 spindle and servo drive motor. This assembly will weigh in excess of 65 lbs.

But this keeps my gantry "Y" plate smaller. And not hanging above my rails and in my mind supports the Z plate in a better fashion.

In your design I see a long unsupported plate that can cause flex and lots of flex when the spindle is raised up.
Weight is going to be heavier since I will be carrying the a axis servo, ballscrew weight, spindle and drive motor.
Hope this makes sense. I Just woke up.

Thanks
Ken
SPMACHINING

[l3viathan]

I am currently designing a similar machine and had questions about your z axis.

Unfortunetly I have no rendering software. Just drawing on paper currently.

On my Z I am mounting my rails to the back of the spindle plate and my bearing on the gantry "Y" side of the plate.

The ball screw but is then mounted to the gantry "Y" plate
. I am using 40mm rexroth rails and a 40x5mm ballscrew.

I think my weight on Z is going to be considerably heavier then yours. Planning on a skyfire cnc bt30 spindle and servo drive motor. This assembly will weigh in excess of 65 lbs.

But this keeps my gantry "Y" plate smaller. And not hanging above my rails and in my mind supports the Z plate in a better fashion.

In your design I see a long unsupported plate that can cause flex and lots of flex when the spindle is raised up.

Hope this makes sense. I Just woke up.

Thanks
Ken
SPMACHINING

There may very well be flex near the top (not sure how much yet until I run FEA), so I was thinking about making the gantry carriage wider than the z-plate, and bolting on sides to make it a channel instead of flat. The plates are 1" mic-6 for reference. All of my calculations when choosing motors has been with an assumption of a 65lb Z-axis. Reversing the bearings/rail, will put the longer plate at the spindle, which could run into clearance issues for some parts I plan to make with this.

I had considered a BT30 spindle briefly, but i'd have to make significant changes, and the cost would go up dramatically. I think your machine will be in a different class than mine..just your rails and ballscrew for the Z probably cost more than my entire machine

[yellow68gto]

I have gotten very lucky finding my components. I will have more into my steel order then into the precision mechanicals.

I am not understanding the clearance issues you mention. I understand the longer spindle plate, this is a nessecity on my design but i dont see how things would hang down any lower

[l3viathan]

I think the best visualization of the two concepts is in this post by Sapele. In the top configuration (which is what I think you are explaining), the tool length determines the maximum reach/clearance. In the bottom configuration (like I have mine), a larger depth can be reached with a shorter tool, assuming the spindle nose can fit into the cavity. I think it's a real balancing act; If I go with a stiffer Z-axis, i'll need longer tool stick-out. Since I plan to use a ER20 spindle, long tooling has larger ramifications. I think in your case, reversing the linear rails is better, because your BT30 will allow for larger/longer tooling (including tool holders) or at least more reach on average.

I do need to put more thought into my Z-axis, it might be possible to negate the clearance issue I was concerned about and have the linear rail mounted reverse from my design above. I'm thinking of the larger 5 axis machines that have very high Z clearance. They use a vertical box section instead of plate which would increase rigidity.

Thanks for the input, you've got me thinking!

[louieatienza]

In bipolar series, the inductance is usually much higher, which requires a higher voltage. The motors i'm looking at are 2.5mH in parallel/unipolar and 10mH in series. If the Gecko formula holds true, in series, the motors should run on 100v without issue. I looked at the Kollmorgen steppers, and the 320v torque curves are scary, they are practically a servo at that point (of course the price point is commensurate). What model steppers are you running at 177v in parallel?

Yes, but at less current, about half the current as bipolar parallel. That's one of the reasons why bipolar parallel get hotter faster, the stepper stays current driven at higher speeds more or less proportionately to bipolar series. I'm running Parker E-AC drive and HV-233 steppers. I did have Lin Engineering build me three new steppers wired bipolar parallel internally but with double the torque rating (600in-oz series, full step). I'm pushing the limits with the smaller steppers, so the bigger ones should give me more headroom and top speed. I don't run them at the highest rated current, rather I back down a couple percent to further aid the motors running cooler. Also the steppers are mounted to proper motor plates, which are mounted to the cast iron structure, further dissipating heat. On all my machines I use dedicated motor mounting plates, of which the standoffs are mounted to, not the stepper itself.

[l3viathan]

Yes, but at less current, about half the current as bipolar parallel. That's one of the reasons why bipolar parallel get hotter faster, the stepper stays current driven at higher speeds more or less proportionately to bipolar series. I'm running Parker E-AC drive and HV-233 steppers. I did have Lin Engineering build me three new steppers wired bipolar parallel internally but with double the torque rating (600in-oz series, full step). I'm pushing the limits with the smaller steppers, so the bigger ones should give me more headroom and top speed. I don't run them at the highest rated current, rather I back down a couple percent to further aid the motors running cooler. Also the steppers are mounted to proper motor plates, which are mounted to the cast iron structure, further dissipating heat. On all my machines I use dedicated motor mounting plates, of which the standoffs are mounted to, not the stepper itself.

I took at a look at the spec sheet for those HV-233 motors, and they have a pretty impressive torque curve. Both the motors i'm looking at and the ones you use both seemingly ignore the 32 * sqrt(inductance) or the looser 25 x rated voltage rules of thumb. Are they just higher quality or use innovative designs that make these "rules" irrelevant?

[louieatienza]

I took at a look at the spec sheet for those HV-233 motors, and they have a pretty impressive torque curve. Both the motors i'm looking at and the ones you use both seemingly ignore the 32 * sqrt(inductance) or the looser 25 x rated voltage rules of thumb. Are they just higher quality or use innovative designs that make these "rules" irrelevant?

It's a "rule of thumb" when estimating what's needed, if you're at a starting point. I'm pretty sure Parker Hannafin's engineers know what they're doing, and these steppers are designed with beefier spindles and bearings, thicker baseplates and end plates, better stampings, better more efficient rotor design, better insulation... They run cooler longer but are able to take the heat, up to 170degC (I wouldn't recommend doing THAT!). Also Marriss does not list preference for stepper motor so I'm sure there is a safety factor involved with someone potentially getting cheap eBay steppers, which have crappy base casting, poor case stampings, cheaper wire, etc...

Again, voltage is the least of concern as long as the motors don't get hot enough to fry. The ones I mentioned are made by companies that sell them to machinery makers, and engineers use the data sheets provided by the manufacturers for their guaranteed performance and reliability. Like I mentioned these better steppers have insulators rated up to 500VDC, so in theory as long as you can keep the motor cool, you can use pretty much any voltage that your drives can handle.

What you DON'T want to do is over-current. That will cause too much heat too fast. I remember just newly installing my GeckoDrive G540 on my machine and making up cables with the current-set resistors. I never bothered to test the RadioShack resisters I bought, and one axis stepper kept heating up fast. Upon checking the resistance was over 10% what it was rated, putting it well past the 3.5A limit of the G540, so the drive would not go into current standby mode.

[l3viathan]

Time for an update. I've been waiting on parts to be delivered, and I've made a few design changes. The initial machine shop I found, cannot actually do the work, as they don't actually have a machine big enough. Rather than go back to pouring epoxy bedding, I've altered the design by using a two section column, so that the column pads and table rail mounts are co-planar and can be Blanchard ground instead. This is probably not what Pippin88 had in mind with a single plane base, but thanks for the idea! I have a few places nearby that have machines more than large enough for grinding, i'm just waiting on quotes.

I'm still trying to figure out an optimal design for my Z axis, but I think i'm fairly close to having a design I'm happy with, and looks rigid enough in FEA tests.

I'm also working on building a power supply, which I'm sure many here would consider ill advised. I have all the parts and I've made more than a few PCBs, and it will save some of my budget for other things. I still have to test the transformers I have to make sure they have a usable ratio (somewhere near 1.55:1).