Messages

6781

General Mach Discussion / Re: Motion controllers - Take me to school

« on: August 31, 2017, 04:21:40 AM »

Hi Mike,
you have some very good advice from some seriously experienced CNCers. It would appear that they all agree your machine could be maximised
by the right steppers/supply and drivers without the expense of either servos or hybrid closed loop steppers. My own modest experience with open loop
steppers is consistent with what they are saying also.

One thing you did comment on in an earlier post was that the old PCs you have been using for a parallel port controlled machine have been unreliable
and you wish to do better.

If you were to use an external motion controller then your choice of PC is much much broader and new and reliable PCs come into their own.
Because your existing control unit is integrated to use a separate controller will require that you use separate drives for your steppers, but you had
already come to that conclusion and were not phased about making the investment. If you were to buy three good quality drivers, say 80V and 7-8A
capable you could, in the first instance, continue to use your existing power supply and steppers. It maybe that just by improving the drivers that your existing
steppers now behave acceptably. If not you can upgrade the power supply and/or the individual steppers until you have achieved the level of performance
you are after. The advantage of this approach is that its progressive. You will have to stump up for an external controller, probably a breakout board or two
and three stepper drivers initially. Thereafter you can spread your purchases to suit.

I'm not familiar with the AM882 that ger21 recommended but guess they are similar to Gecko drivers, that is to say 80V 7A capable and quality/reliable build.
They can be had for about $150 each. There are a few choices of motion controller in the 'cost effective' end of the market, the ESS among them at $180.
Add a breakout board for $50 say...a total initial investment of $680.

Craig

6782

General Mach Discussion / Re: 2 Axis Dispaly w Encoders

« on: August 31, 2017, 03:44:51 AM »

Hi,
that read out is made for a particular linear encoder. You will have to match the linear distance travelled per count to one count of your rotary encoder.

If for instance one linear scale count is 0.01mm then you will have to arrange a rotary encoder to match... lets say you have a 5mm pitch ballscrew
direct coupled to your stepper then the stepper will rotate (0.01/5)*360  which is .72 degree. Thus your one count is 0.72 degree or (360/0.72)= 500 count per rev,
therefore your rotary encoder must have 500/4=125 lines. I don't recall seeing an encoder with that unusual number of lines.

It may be some of the 'programmable' magnetic line encoders could offer the flexibility to specify some unusual number of lines. The only ones I've seen locally
have been worth BIG BUCKS, like $800 each!

To be honest I think trying to match a linear scale to a rotary encoder is going to be a real battle.

Perhaps ask the manufacturer of the readout if it is possible to 'reprogram' the readout , that I suspect would be a lot easier than the reverse.

Craig

6783

General Mach Discussion / Re: Motion controllers - Take me to school

« on: August 30, 2017, 03:16:00 PM »

Hi Mike,
yes you have some decisions to make.

Be aware that while steppers aren't great at speed they are very grunty when slow, in fact they beat the hell out of servos for torque of the same size.
You may find that to even come close to matching a steppers torque you have to buy a larger servo than you imagined. While you may be happy
enuf with lower thrust as a result of low torque you will absolutely miss the acceleration. Acceleration is way more important to cycle times than
high max speeds.

Given that this is a hobby sometimes you have to consider the fun/learning/satisfaction of doing a certain thing rather than its economic justification.
Having said that spending a big chunk of your budget on a certain project may well prevent you from taking on anther project that may have greater
value or applicability.

6784

General Mach Discussion / Re: Motion controllers - Take me to school

« on: August 30, 2017, 04:03:12 AM »

Hi,
garyhlucas  poses a very good question.

My own machine is probably even smaller than your Sherline, the X,Y,Z is 200mm/200mm/200mm. I use 5 phase Vexta steppers with low backlash
(less than 3 arc min)  10:1 planetary gearboxs. The gear reduction means they have heaps of thrust and great resolution (1um) but are slow,
G0s are 1200mm/min (60in/min). Given the size of the machine I don't really consider the slow rapids to be too much of a disadvantage.
Like garyhlucas my steppers never miss a beat either. If I replaced them with something faster I might be able to use some of the High Speed Machining
tooplaths talked up by all the major CAM manufacturers but that's about it. That's a lot of money to spend to achieve so little. Might make more sense
if this were a production machine but it isn't and will never be a production machine.

I have started to buy up components to speed up my machine... 34 size 5 phase steppers from Vexta at 900 oz.in. If I were to use half steps, entirely adequate
with the smoothness of 5 phase steppers and drive the ballscrews directly without the gear reduction I would have a resolution of 5um, still adequate and
based on published performance curves G0s between 7500mm/min and 12500mm/min. Vexta steppers and drivers are expensive and don't come up that often,
given that the benefits are fairly limited I've become quite tepid about shelling out all those bucks when they do come up!

I did however need a decent high torque spindle for machining steels and did get around to spending $1500 or so on the parts to make one, it does get
used and represents a better use of my budget than faster steppers/servos. Yes garyhlucuas raises a very interesting question, do you really need to improve
on what you already have?.

Craig

6785

General Mach Discussion / Re: Motion controllers - Take me to school

« on: August 29, 2017, 09:21:00 PM »

Hi,
in broad terms you are correct but the finest resolution that can be achieved by a stepper remains the same.
The closed loop part means it correct its own errors.

AC servos on the other hand can offer greater resolution and speed but at a price. If you wish to get best resolution
at max speed then the humble parallel port may not be fast enuf for your purposes. In that case an external motion
controller is indicated, for instance and incidentally my perference is the Ethernet Smoothstepper by Warp9.

Whether you run Win7,XP or Win8 or 10 32 or 64 bit is immaterial to Mach if you are using an external controller.
I run a dual core Atom single board computer for my machine controller, hardly enuf computing power to get out of its
own way, but it works fine and is of course new. With an external controller using new PCs is well in order.

Craig

https://warp9td.com/

6786

General Mach Discussion / Re: Motion controllers - Take me to school

« on: August 29, 2017, 03:04:05 PM »

Hi,
I'm guessing that your existing arrangement is one large board with three stepper drivers  and a LPT interface?

To run separate drivers as you are proposing would require a breakout board. It job is to interface with the computer and provide some protection
for it so that whatever you hook up to it can't wreck your PC. Then two wires (signals) from the breakout board would go to each of your drivers.
In this situation the PC is still the 'motion controller' normally called a parallel port. You could use a separate motion controller if you wish but you
don't have to.

Is there a reason you wish to go for closed loop steppers?. When all said and done they are still steppers, they can't go any faster or magically get more
grunt by running with a closed loop drive irrespective of the advertising hype. If your existing steppers are prone to losing steps then closed loop drivers
can correct that by inserting extra steps to make up for the ones the stepper couldn't quite make. You are paying a premium when for just a little bit
more you could have genuine AC servos with all the speed advantages they bring. If your existing steppers don't lose steps then why bother with closed loop?

Craig

6787

General Mach Discussion / Re: C6 and C10 boards Isolation?

« on: August 29, 2017, 03:51:29 AM »

Hi,
you say that you have separate 5 and 12V supplies, what type are they? If they are buck regulators which are quite common then they won't be
galvanically isolated. Can you post some info about the supplies?

Craig

6788

Mach4 General Discussion / Re: Mach 4 UC100 HDBB2 Spindle Speed setup (last piece to the puzzle)

« on: August 29, 2017, 02:36:01 AM »

Hi,
kool, glad you got it worked out. Because I don't have a UC100 I believe part of the plugin is unavailable to me to edit.

The smoothstepper plugin was like that, if you don't have it plugged in and communicating the plugin didn't show up. A more recent release
has changed that, not that it makes much sense to edit the firmware when the device isn't plugged in but its good to see it so you can plan/scheme/
understand.

Craig

6789

General Mach Discussion / Re: Mach3 4th axis assistance Is Mach4 better for this?

« on: August 28, 2017, 06:56:12 PM »

Hi,
phew! 8000 deg/min-much better. At 10 microstep I come to 33.33334 steps per dregree at 6:1 like you.

Tkanks ger21 for clearing that up!

Craig

6790

General Mach Discussion / Re: Mach3 4th axis assistance Is Mach4 better for this?

« on: August 28, 2017, 03:12:18 PM »

Hi,
not quite sure what you are asking, sounds like several issues, sorting one by one is likely to result in less confusion.

If you are using the same motion controller (ESS) then it doesn't make any difference at all between Mach3 or Mach4. Mach4 is no faster,
more accurate or anything else, its advantage is the modular modern programming style which results in fewer bugs and WAY better
customisation ability. Likewise you can have fans up the yahzoo, a CPU with muscles on its muscles and it won't make your machine
run any faster. Your problem is not computing but your stepper/driver combination.

Suggest you start by checking each axis one by one. Put ridiculously slow speeds and accels and build up until you find the point where you
start losing steps or stall, then back off, significantly if your being conservative.

I can't work out what your A axis gearing is from your description but sound like the stepper has to spin WAY too fast to achieve the resultant
speed you want. Either change the gearing or fit a servo. A stepper loses authority beyond about 500 rpm down to about a 1/4 of its grunt at 1000rpm.
I find it hard to believe you can spin a stepper at 8000 rpm, aside from anything else the rotor must be due to explode with centripedal accel!

Craig