[morecwbell001]

Hello- Several years back I purchases a Sieg type X2 Mini Mill and converted it to CNC. I haven’t run it in a long time, but became pretty frustrated when during longer cuts (1hr+) the machine would miss steps and ruin the parts (that had taken several other long cuts to get to this point). In researching this a bit, I thought that upgrading to adding rotary encoders would be helpful, but I’m open to all suggestions. Hoping to spend no more than $500-$1000 to get a more reliable machine. I'm looking for your guidance on what I could/should do please!

For reference my CNC upgrade was done using the following:
• X, Y, Z Ballscrews (CNC Fusion kit)
• NEMA23 270 and 381 oz/in steppers (X + Y/Z respectively from Automation Technology Inc.)
• Gecko Diver G540 4 Axis Driver (Automation Technology Inc.)
I’m not even sure the driver would accept encoder inputs.

Please provide me any suggestions!

Kelly

[he1957]

The Sieg X2 is a small mill with a 1/3 HP spindle motor so it's not intended to be too aggressive in the works requested of it. Some ideas of what you are cutting, how and the speeds and feeds along with the tooling being used would be useful.

First things might be to check if the machine holds position when doing air cuts over long periods, then perhaps to use lighter loads (lower travel speeds and lighter cuts).


Cheers,

HarryE.
===

[joeavaerage]

Hi,
I suspect you must be pushing the envelope of your steppers. You haven't specified what voltage power supply you are using for
the G540 stepper drivers? If its low, that is 24-36V you could try increasing it to the limit of the G540, 48V. That may improve the situation at speed
to avoid missing steps.

Even if you had encoders, the G540 can't handle all of the inputs. Even if it did, then what? You need some sort of smart driver to take advantage of the
feedback. And even if that were the case....they are still steppers, feedback DOES NOT MAKE THEM MORE POWERFUL, they are just as subject
to overload and losing steps as open loop steppers. You say 'but the smart drive will issue more pulses to keep up', but the only reason it lost steps in the
first place is overload, what make you think that extra pulses are going to have an effect when its already missing ordinary pulses?

To be honest, fitting encoders OR fitting closed loop steppers and drives does not offer enough advantages for the money invested. At best the steppers
will still miss steps as they do now but the smart drives would fault 'following error' when they do so.

I think you could really make a difference to your machine by fitting genuine AC servos.

https://store.dmm-tech.com/products/...ac-servo-motor
https://store.dmm-tech.com/products/...ac-servo-drive
https://store.dmm-tech.com/products/...-encoder-cable
https://store.dmm-tech.com/products/...-camp-hh-_-ssp

That is $321 per axis (60V,200W drive, servo and cables) which takes near to the top of your budget but it will achieve a result that steppers of any description
won't match. For an extra $10 a servo you could step up to 60V 400W servos. They will wipe the floor with any stepper!

Craig

[morecwbell001]

Thank you Craig and Harry!

To be clear on my issue- I'm not concerned so much about speed or power, but want to be able to walk away from my machine and know it didn't screw the part up. Ideally even being able to stop and fault out would be better than having to babysit it the entire time, or having to break down the cut paths into many small paths and starting and stopping the machine over and over.

A bit more info:
I am supplying the steppers with 48 volts

Craig- I like your suggestion to go with AC servos and would like to learn more. Can you help me understand what I would need above and beyond those four items? I used a UC100 Parallel to USB motion controller for Mach3 in addition to the items I listed, but this is an area that I'm pretty confused. I'm running Mach3 on a dedicated cheap desktop and it seems to work fine.

Kelly

[morecwbell001]

The Sieg X2 is a small mill with a 1/3 HP spindle motor so it's not intended to be too aggressive in the works requested of it. Some ideas of what you are cutting, how and the speeds and feeds along with the tooling being used would be useful.

First things might be to check if the machine holds position when doing air cuts over long periods, then perhaps to use lighter loads (lower travel speeds and lighter cuts).


Cheers,

HarryE.
===

Harry-
I'm not sure I remember exactly the background of if it was only with finish cuts or not, but I do remember it happened often on finish cuts. I think I would have been using a 1/4" ball end mill taking off .005"-.010" at pretty low speeds.

Kelly

[rcheli]

I have an X3 running 381oz-in steppers on the X and Y and Nema 34 on the Z all driven by G203V Geckos @ 48volts. I have never missed a step that I know of. I have same question as others have mentioned about what voltage you are running at? Also why do you have different steppers on the X &Y? I looked at the automation technologies website and it looks like the 270 oz-in stepper has an inductance of 3.6 which means you could run that one at 60 volts except for the 50 volt limit of the G540. Where as the 381 oz-in one has an inductance of 2.8 and could be run at 53 volts so is a better fit for the G540 50 volt limit. What I would do would be to change the 270 stepper for a 381 and up your voltage to 48 volts if you are not already there. Then while waiting for parts check the gib adjustment and look for any mechanical binding. Also be sure that your PSU can supply the required amps.

[he1957]

Humm, for an X2 .010" is a quarter mm (0.25) - that's a fairly heavy load if it's mild steel, sticky for aluminium if not using coolant or compressed air and the right cutter. The Sieg range about 10 years ago were a bit notorious for inconsistent casting/mounting holes so your mod kit may not align as well as it could. Check the table moves smoothly, especially when changing direction and/or entering curves/corners. Belt tension (if fitted) or gibs too tight?

See what max travel speeds you get with no load and/or try setting the acceleration/braking curves lower or higher (at finishing cut rates inertia may be causing lost steps).

Try an air cut, whole-table coverage pattern with some "interesting moves" and see if it holds origin.


Cheers,

HarryE.
===

[joeavaerage]

Hi,

To be clear on my issue- I'm not concerned so much about speed or power, but want to be able to walk away from my machine and know it didn't screw the part up

That is EXACTLY what this is about, your existing stepper DOES NOT HAVE ENOUGH TORQUE at the speed you are driving them. The attached pic is a pretty typical
example of a 23 size stepper. Note how the torque is reduced to less than half by 500 rpm. Thus your machine doesn't have to be going very fast at all before
the torque reduces to the point where it misses steps.

A servo on the other hand has nearly constant torque right up to it max rpm. The 200W servos I linked too in the previous post have a torque of 0.64Nm or abut 80 oz.in
but maintain that up to 3000 rpm. Peak (short overload) torque is 1.9Nm or about 200 oz.in. The slightly more expensive 400W servo has cont. torque of 160 oz.in and a
peak torque of 480 oz.in

Unlike steppers, even closed loop steppers, when a servo lags behind the servo drive really ups the current and therefore torque. There is a limit to how long it can do
that before overheating but none the less it is VERY MUCH more likely to RECOVER ITS POSITION than a stepper.

Modern AC servo drives/servos are as easy to drive as a stepper.

Servos of this description would mean that you can go faster, despite that not being your primary objective, but more importantly it will move with MUCH more
AUTHORITY and if it does lag behind (by a programmable error) it fault and stops rather than bugger up your part.


You would replace your G540 with a cheap parallel port breakout board like a C10 ($23 fromCNC4PC) and run step and direction wires to each servo drive. You will
also probably run another wire common to all servo drives to enable them, and probably one wire common to all servo drives back to the breakout board
and therefore back to Mach as a fault signal.

Craig

[109jb]

Suggesting $960 for AC servos for an X2 mill is in my opinion a lot of overkill. You can buy a brand new X2 type machine for about $600 and it just seems extreme to spend 1-1/2 times that much on drive motors. Oh, and that price doesn't include the power supply you will need for the AC servo/drive setup.

There is also nothing wrong with an open loop stepper system that is set up properly. I have a open loop stepper system on my G0704 mill and it works just fine. Tormach machines use open loop steppers and they work just fine. There are tons of open loop stepper machines both factory and converted that operate just fine. The key is setting them up right.

To put it in perspective, I just bought three 1600 oz-in closed loop steppers with AC powered drives for an RF-45 size machine for less than the 200W servos linked. I have the X and Y axes under power and they can run better than 200 IPM without issue. I really don't know the top speed yet because I'm testing with Grbl which is limited to 30kHz step rate.

So, what is going on with the OP's machine? Some things we know:

1. Machine has ballscrews
2. G540 drive (50V, 3.5A max)
3. 270 ox-in on X and 381 oz-in on y and Z (Why the difference? )
4. Using Mach3 with UC100


What we don't know:
1. Power supply voltage
2. Max feedrate
3. acceleration setting
4. Backlash compensation turned on?

On that last one, I've never used mach3, but I seem to remember talk of the backlash compensation causing problems presumably because it tries to instantly accelerate the axis on direction reversals.

From reading I also understand that the UCCNC software from the UC100 maker is very good software, most users saying it is better than Mach3, so you may want to consider switching to it.

The first things I would personally do is make sure you have a 48V power supply and that the G540 is properly set up for the motors. (ie proper current set resistors installed). If the current set resistors are wrong then you could have shorted your motors due to too much current. You may also want to get a third 381 oz-in motor, but should be able to experiment some first. Also make sure the gibs are properly adjusted and not too tight. Also turn off backlash compensation if on.

For setting up steppers, I generally do something like this:

1. Set acceleration low, like 3 in/sec^2 and make moves of increasing speed until the motor stalls. Do this several times to get the lowest speed at which stall happens.
2. set the max feedrate to 80% of the result of #1
3. now increase the accel gradually until the motor stalls trying to accelerate. Again, do this several times to find the lowest accel that causes a stall.
4. set accel to 80% of this value.
5. repeat the above for all axes.

I like to set all 3 axes to the lowest value for any single axis, but it really shouldn't matter. That should give reliable accel and max feed settings.

Now you can run some torture tests to see if you are still losing steps.

[morecwbell001]

Suggesting $960 for AC servos for an X2 mill is in my opinion a lot of overkill. You can buy a brand new X2 type machine for about $600 and it just seems extreme to spend 1-1/2 times that much on drive motors. Oh, and that price doesn't include the power supply you will need for the AC servo/drive setup.

There is also nothing wrong with an open loop stepper system that is set up properly. I have a open loop stepper system on my G0704 mill and it works just fine. Tormach machines use open loop steppers and they work just fine. There are tons of open loop stepper machines both factory and converted that operate just fine. The key is setting them up right.

To put it in perspective, I just bought three 1600 oz-in closed loop steppers with AC powered drives for an RF-45 size machine for less than the 200W servos linked. I have the X and Y axes under power and they can run better than 200 IPM without issue. I really don't know the top speed yet because I'm testing with Grbl which is limited to 30kHz step rate.

So, what is going on with the OP's machine? Some things we know:

1. Machine has ballscrews
2. G540 drive (50V, 3.5A max)
3. 270 ox-in on X and 381 oz-in on y and Z (Why the difference? )
4. Using Mach3 with UC100


What we don't know:
1. Power supply voltage
2. Max feedrate
3. acceleration setting
4. Backlash compensation turned on?

On that last one, I've never used mach3, but I seem to remember talk of the backlash compensation causing problems presumably because it tries to instantly accelerate the axis on direction reversals.

From reading I also understand that the UCCNC software from the UC100 maker is very good software, most users saying it is better than Mach3, so you may want to consider switching to it.

The first things I would personally do is make sure you have a 48V power supply and that the G540 is properly set up for the motors. (ie proper current set resistors installed). If the current set resistors are wrong then you could have shorted your motors due to too much current. You may also want to get a third 381 oz-in motor, but should be able to experiment some first. Also make sure the gibs are properly adjusted and not too tight. Also turn off backlash compensation if on.

For setting up steppers, I generally do something like this:

1. Set acceleration low, like 3 in/sec^2 and make moves of increasing speed until the motor stalls. Do this several times to get the lowest speed at which stall happens.
2. set the max feedrate to 80% of the result of #1
3. now increase the accel gradually until the motor stalls trying to accelerate. Again, do this several times to find the lowest accel that causes a stall.
4. set accel to 80% of this value.
5. repeat the above for all axes.

I like to set all 3 axes to the lowest value for any single axis, but it really shouldn't matter. That should give reliable accel and max feed settings.

Now you can run some torture tests to see if you are still losing steps.

Thank you to everyone for the good feedback. It sounds like I have a good amount of checking and testing to do. I'll give some of these things a try and see if I'm still having the issue. I very much appreciate the quick responses and useful feedback!

[cyclestart]

Several years back

Bit of a long shot here but the earliest run of G540's included some duds. If 'several' means many it's an outside possibility. The revision # is stamped on the board.

Hybrid closed loop steppers do some cool tricks. If they're of any interest I'd encourage you to do a bit of research.

[cyclestart]

And a good comparison of the options in this video.

https://www.youtube.com/watch?v=OrEwKZ8JWAU
The action starts around the 20:00 mark

[joeavaerage]

Hi,
great video. The closed loop stepper did better than I thought it would but still not a patch on an AC servo.

Simply put if you have the budget....why wouldn't you use servos?

Craig

[109jb]

Hi,
great video. The closed loop stepper did better than I thought it would but still not a patch on an AC servo.

Simply put if you have the budget....why wouldn't you use servos?

Craig

Simply put..... Because its not needed. Why spend $1000 when it could be a lot cheaper to diagnose it and find out what is really going on instead of just throwing money at it. It's the same reason I drive a 14 year old truck instead of buying a new one. I could pay cash for a new truck but I drive the old one because it gets the job done.

Lets look at some numbers:

1. You stated that the AC servo was 80 oz-in, 200oz-in peak
2. You posted a graph showing about 50% torque loss for a open loop stepper at 500 rpm

so lets multiply 381 oz-in by 0.50 and we get 190.5 oz-in at 500 rpm for the stepper, and a max of 200 oz-in for the AC servo. Pretty much the same. We know that the AC servo torque curve is pretty flat, so above 500 rpm, the servo has more torque, and below 500 rpm the stepper has more torque. And the stepper torque is available 24/7. The AC servo can only do that 200 oz-in for a short time, measured probably in seconds.

So below 500 rpm the stepper has more torque. If the steppers can't push the machine around below 500 rpm, then that means the servo can't either if it is direct drive. 500 rpm on a 5mm pitch screw is 98 IPM which is more than enough for a x2. Since the OP alluded to problems even at low feed rates, presumably much lower than the 98 IPM, the AC servo in a direct drive situation is no better and is actually probably worse. At low feed rates, the stepper should have a much higher available and reserve torque than the servo.

Servos are commonly belt driven at 3:1 or 2:1 so that the rpm capability can be utilized. However that adds more cost and adds complexity. It is great that these little servos can maintain their torque to 3000 rpm, but it doesn't mean squat if you can't even get to 3000 rpm.

This is why the OP need to first figure out what is wrong with the system as a whole. If the steppers aren't really performing as they should, there is a reason. If the steppers are ok, then something else is going on that needs to be discovered and addressed.

I just came in from about 8 hours of working in my shop, using my open loop stepper driven G0704, with one job that had a cycle time of about 1:15, and I didn't have one issue at all.

About 2 years ago that 14 year old truck of mine had problems. Sometimes wouldn't run right, other times would run fine. Sometimes wouldn't want to start. I could have taken $50,000 out of the bank and bought a new truck. ........Why wouldn't I since I have the budget for it......... That certainly would have settled the issue of not running right, but instead a little diagnosing by the shop and a $250 ECM solved the problem. Total bill was about $800, or $49,200 less than the new truck.

[joeavaerage]

Hi
fine, stick with your steppers, when budget permits I'm going to servos.

Why do industrial machines all have servos.....when steppers are good enough?

Craig

[109jb]

Industrial machines have much different requirements that a hobby machine. Industrial machines have tables alone that weigh several times what a entire X2 weighs. They run upward of 800 IPM rapids or more to reduce cycle time. They have linear ways. They have high horsepower high speed spindles. Every second counts when talking industrial machines and production environments. An industrial machine is many many times more rigid and capable, requiring axis drives that are equally as capable.

None of that applies to an X2. It is a lightweight, slow speed machine that doesn't have enough rigidity or size to need more than about 50 IPM. Properly configured open loop steppers are more than capable.

Even in industrial machines the design requirements will dictate what is required. I will guarantee that if lower cost options will fulfill the requirements for a product, it will have the lower cost components.

...when budget permits I'm going to servos.

So let me get this right. You don't even have the servos you are proclaiming are the solution to all the OP's problems?

[joeavaerage]

Hi,
no, I couldn't afford servos at the time I designed and built my benchtop bed mill.

What I did was use Vexta 5 Phase Steppers with integrated low lash (< 2 arc min) 10:1 planetary drives. Even second hand they (and the 230VAC input drives)
were expensive, but at that time were still cheaper than available AC servos. I have been very happy with the steppers, in fact yesterday I spent all day making
circuit boards with my machine, one of the tasks for which I designed it.

With the gear reduction my machine has a great deal of axis thrust, over 1500 lb force at stall, which satisfies the other requirement that I wanted, namely the thrust
and rigidity to machine steels and stainless steels. Because of the gear reduction means my machine is slow, 1200mm/min or 60IPM.
The cost (off EBay) worked out to about $300 per axis. As I say I am very happy with them and I have had great service from them.

The DMM servos I linked too are still modestly more expensive again but were not readily available at the time I was designing and building my machine. Were I to build
my machine today I would use the DMM servos without a doubt.

One area where I miscalculated/misunderstood when I built my machine was I bought a 24000 rpm 750W Mechatron air cooled spindle from Germany. It a great quality spindle
and I am very happy with it....I used all day yesterday for instance. It is great for engraving and spinning small tools in aluminum but I had hoped it would work in steel.
High speed low torque spindles such as this are a poor choice for steel/stainless as I discovered ie my miscalculation/misunderstanding.

As a consequence I elected to make another spindle using a cylindrical ER25 Rego-Fix toolholder, P4 angular contact bearings being directly driven
by a second hand 1.8 kW 3500 rpm Allen Bradley servo and drive. I was able to buy the second hand servo and drive in New Zealand for $600 USD (equiv)
including shipping. I did have to buy/make cables and I also had to buy the servo drive setup software from Rockwell. Still all up the servo and drive cost
about $1000 USD (equiv). I duly made my high torque spindle and despite being home made and not without its faults I am none-the-less proud of it, and
it works an absolute treat in steel and stainless. I can swap between my highspeed spindle and my high torque spindle in about 15 minutes.

It is not as convenient as just having the one spindle that can not only spin fast for engraving but has high low speed torque for steels, but then again
such a spindle is well beyond my budget.

When I bought this servo it was the first 'modern' AC servo I'd ever really encountered and used. I was absolutely gobsmacked at the flexibility of the control
options, monitoring, setup assistance that is conferred by the drive. Quite simply it an order of magnitude better than open loop steppers.
Additionally now that I have some use of the thing I realize that the servo 'appears' to have very much more grunt than the specs would suggest.
Clearly that it not a scientific observation but the servo driven spindle seems to consistently outperform my best expectations.

Since then I have had some experience with Delta servos as an in service replacement for a DC brushed servos in a customers large plasma table.
The DC brushed servos had the loop closed within the controller and added a lot of complexity in addition to the cost of replacement of the nearly
obsolescent servos. The AC Delta servos were a breeze to fit and reduced the controller complexity considerably. The smarts built into an AC servo drive
means it is easy to use and integrate into your machine.

Delta and DMM are cost effective brands of good quality and considerably cheaper than US and European brands. There are still cheaper Chinese brands
but I'm hesitant to recommend them.

I take and understand your assertion that $1000 worth of servos is unwarranted for a $600 machine. Likewise I understand your homily about your truck.
If there is a fault with OPs machine or setup then it makes sense to address the fault rather than 'waste' money.

I note that you seem inclined, if you were to upgrade, to go to hybrid closed loop steppers. An intermediate solution costwise from the cheaper open loop
steppers to the more expensive AC servos. I personally regard the cost of the closed loop steppers to be unwarranted for the performance increase over
open loop steppers. The scientific evidence for the advantages of close loop steppers is pretty thin. I was therefore surprised at the outcome of the video
you linked to.

Personal experience tells me that 'modern' AC servos really REALLY kick anus......and ARE worth the premium (with Delta and DMM brands at least).

The question to whether OP would be overcapitalizing his machine were he to fit AC servos is as much as anything a matter of opinion. My guess is that OP
would, as I have become, so enamoured with AC servos that he would pleased no matter the investment and would almost certainly retain those same servos
for other 'larger' projects down the road. Just my opinion...

Craig

[cyclestart]

There's some confusion here over who posted what.
109jb isn't pushing hybrid steppers for this application and I posted the video. The real reason i posted was the bit about early G540's.
Fwiw my X2 is near hardware identical to OP (Sorry:long user name) and it's been OK. Admittedly it's not run hard.

[joeavaerage]

Hi,
my apologies, I did not thoroughly check whom had posted what.

To whomever posted the video link I was pleasantly surprised that the hybrid closed loop stepper performed as well as it does. The published torque/speed curves
would not suggest such an apparent improvement over a closed loop stepper.

Having said that I would not have expected a servo to perform as well as it does based on the specification alone and yet that is exactly my experience. I should
not be surprised then that closed loop stepper specifications somewhat understates their performance in the real world.

Open loop steppers still retain a significant price advantage over either of the closed loop solutions. Steppers, more generally known as reluctance motors, exhibit
the highest torque density (Nm per kg motor weight) than any other motor design at slow speeds. The problem is that at modest speeds they lose their authority.
The exact speed at which they become subject to losing steps is highly dependent on the machine, the cutting forces, the driver and its power supply. It is difficult in
advance to predict the effective performance envelope of a stepper. The video suggests that a closed loop stepper extends that envelope despite scientific
evidence to the contrary.

Servos on the other hand are well defined. While they have less torque ( at low speeds) than a stepper of the same size the overall performance envelope
exceeds steppers. An AC servo is an example of a Permanent Magnet Synchronous machine and have the highest power density (kW per kg motor weight)
of any conventional electric machine.

Excepting AC servos, synchronous AC machines are not that common in every day experience however all the electricity available to us on the public power grid
is generated by synchronous generators, albeit DC excited rather than permanent magnet armatures. The capability, reliability and ruggedness of synchronous
machines is very much the subject of vast swathes of technical engineering literature. That the same flexibility and ruggedness should be found in small synchronous
machines, aka our AC servos is no surprise.

It is the flexibility and ruggedness of AC servos that impress me greatly and I am delighted that quality cost effective brands such as Delta and DMM are bringing the
prices down to within my range. Such was not the case as little as five years ago, or at least it has improved over that time.

Craig

[109jb]

Cyclestart is right. I am not pushing any kind of new hardware for the OP until it is determined that the existing hardware won't fulfill his needs. There is another recent thread with a person having essentially the same problems as described here, missed steps even at low feeds, etc. The problem turned out to be that he was using the wrong dip switch setting on his open-loop stepper drivers. A few clicks and as that person described, he has more performance than he needs. The G540 that this OP has may just have the wrong 10 cent current set resistors installed. We don't know yet.