What's new
What's new

DMG Mori Celos any good ?

Panza

Stainless
Joined
Oct 23, 2005
Location
Lillehammer, Norway
I'm going to buy a new Y-axis lathe with 2 spindles.
I am looking at a DMG mori NLX2000SY. It has Celos control with Mapps (M730UM with 2 touch screens). I have not found a single positive review so I am kind of sceptical even if the price is great. Another option is a new Mazak 200MSY, which is more of a safe bet in the control department. Don't know if there is any other similar machines worth looking at ?
Any feedback on the M730UM control if very helpful.
 
What year is the DMG? How good is support for DMG Mori in Norway? How good is Mazak support in Norway? I think you are right in your opinions on a "safe bet".

Celos is a powerful Computer....the touch screen can have some delayed response time, between touch and action. There is a lot of page down, page over, page down. If the shop or the Machinist is particularly dirty I would stick to hard keys, instead of touch screen. We all have microscopic metal chips in our fingers-touch screens have always seemed dumb to me. But that's just opinion. BUT there is nothing outwardly wrong with Celos, it functions, it's just a little different.

R
 
I'm going to buy a new Y-axis lathe with 2 spindles.
I am looking at a DMG mori NLX2000SY. It has Celos control with Mapps (M730UM with 2 touch screens). I have not found a single positive review so I am kind of sceptical even if the price is great. Another option is a new Mazak 200MSY, which is more of a safe bet in the control department. Don't know if there is any other similar machines worth looking at ?
Any feedback on the M730UM control if very helpful.


I've mainly looked at NLX 2500's. Not run one of their current models with Celos.


"Peeps" here on Pm forum that have chimed in on CELOS for the NLX platform seem to like it... Seems fairly intuitive/straight forward.


Titan's shop bought one to go with their DMU 50 3rd gen.



^^^ you get to see a bit of celos here later on in the video.

__________________________________________________ ________________________________________________



I believe the biggest difference between MAZAK and a NLX is that MAZAK use meaty rolling element linear guides whereas NLX uses Boxed ways.


For me ---> read obscure nerdy applications ---> The rolling element linear bearings are better for spherical and curved profiles that ideally are cut to high form tolerances, same with tapers. I think Boxed way shine for automotive type parts for precise and robust straight cuts.



However: DMG MORI have come out with a "composite" lathe where it's Boxed way on Z but linear rolling element bearings in X. ---> ALX 2 - Universal turning from DMG MORI

^^^ ALX 200 (I talked to local DMG mori sales about it when it was just launched at IMTS but it's hard to get an opinion from those guys as they are always trying to up-sell on something, but seems more affordable than NLX perhaps. [Has integrated DD type motor head stock... Looks like they were trying to build a MAZAK lol]. Mitsubishi control on the ALX.

____________________

"Latest in-house made DDS (Direct Drive Spindle) turnMASTER distance between centers 300 / 500 / 1000 / 2000 type

Slideways on the X-axis for higher damping performance, and linear motion guides on the Y- and Z-axis for improved positioning accuracy*

Full closed loop control (Scale feedback) SmartSCALE as an option: Enhanced positioning accuracy"


______________________________________________________________________

* Never 100% confident by what DMG mori mean by Slideways versus 'Linear motion guides" ?
 
...... If the shop or the Machinist is particularly dirty I would stick to hard keys, instead of touch screen. We all have microscopic metal chips in our fingers.....

Kind of off-topic, but in regards to touch screens...

My Mori touchscreen is 22 years old now. The only trouble I have ever had with it is that it will not illuminate if the shop temperature is below ~45F. I try to make sure to not have coolant on my fingers before using it. I have no doubt that the 22 year old gasket sealing the edges appreciates that consideration.
 
I'm sure you're right. It's just a thing for me, I hate fingerprints on glass.

I have to admit I've never completely understood "Smudge screen technology"... Being in optics I too cringe at a glass platen with finger prints.


Personally I don't think keyboards that are a flat piece of glass 1:1 scale with a computer keyboard are that intuitive or ergonomic and I suspect if you type out the complete works of Shakespeare you'd end up with carpel tunnel syndrome. I think the mechanical action no matter how slight under your finger tips is really important. [I get that the CELOS glass keyboard can be cleaned off/ as opposed to stuff getting trapped around regular key recesses].

Wonder how easy it is to "Finger futz" versus mechanical keys ?

______________________________________________________________________________


* Anyone that has watched enough episodes of Star Trek (Next Generation) you can visibly see how awkward it was for the cast to even pretend to type on glass with your fingers.

** I know a couple of cinematographers where below a certain temperature their touch screen controls (for camera and monitoring) simply don't work. I'm not sure if they were vampires or not/ or had really hot or cold fingers ?
 
Not that it should be an issue getting that cold in a CNC shop but if it does then yes LCD touchscreens will get glitchy.

I worked as project manager for a computer company and constantly saw the laptop PM having to tell people in colder climes to not leave laptops in cars or take them out in the field during colder weather. They slow down and just stop working.

Cold enough will cause them to fail altogether. There are hardened application LCD's that can put up with this but they aren't putting those in machine tools as the assumption is I am sure that you aren't going to put a CNC machine on the deck of a ship or out back in the parking lot.

EDIT: As to the original topic on Celos and the complaint people have about lag. Heard that a lot on them. There should be NO lag in a LCD. But as I understand Celos is just a front end for whatever controller you like, Fanuc, Siemens, Heidenhain, your pick.

So it would seem right now the translator between the Celos interface and the actual controller needs some work if it's still laggy. You would figure that is something they can get ironed out if they think its an issue.
 
Not that it should be an issue getting that cold in a CNC shop but if it does then yes LCD touchscreens will get glitchy......

Yes, the only reason I found that out about my touchscreen was when a tree fell on my shop and broke the entire roof. Could not afford the cost to heat it when half the roof was a Harbor Freight tarp.
 
Not that it should be an issue getting that cold in a CNC shop but if it does then yes LCD touchscreens will get glitchy.

I worked as project manager for a computer company and constantly saw the laptop PM having to tell people in colder climes to not leave laptops in cars or take them out in the field during colder weather. They slow down and just stop working.

Cold enough will cause them to fail altogether. There are hardened application LCD's that can put up with this but they aren't putting those in machine tools as the assumption is I am sure that you aren't going to put a CNC machine on the deck of a ship or out back in the parking lot.

EDIT: As to the original topic on Celos and the complaint people have about lag. Heard that a lot on them. There should be NO lag in a LCD. But as I understand Celos is just a front end for whatever controller you like, Fanuc, Siemens, Heidenhain, your pick.

So it would seem right now the translator between the Celos interface and the actual controller needs some work if it's still laggy. You would figure that is something they can get ironed out if they think its an issue.

RE: Celos , I've heard they have been slow on tool changes etc. on mills / horizontals in that past (perhaps being extra safe). I know they had to do a ton of code migration for CELOS but didn't know there was an inherent latency in the interface ? (but maybe not an issue for a "Lathe" / turning center ?).

I wonder if they have ironed that out/ more now ?
 
What year is the DMG? How good is support for DMG Mori in Norway? How good is Mazak support in Norway? I think you are right in your opinions on a "safe bet".

Celos is a powerful Computer....the touch screen can have some delayed response time, between touch and action. There is a lot of page down, page over, page down. If the shop or the Machinist is particularly dirty I would stick to hard keys, instead of touch screen. We all have microscopic metal chips in our fingers-touch screens have always seemed dumb to me. But that's just opinion. BUT there is nothing outwardly wrong with Celos, it functions, it's just a little different.

R

The DMG is brand new. It is in stock so I suppose it could be made in 2017.
Mazak support is ok in Norway, not fast but ok.
DMG is about the same but that is what I have heard. No personal experience with them apart from the sales people which leaves a bit to be desired.
The Mazak QT MSY has touch screen too but that is reported to be fast.
The NLX is ways only while the Mazak is linear guides only. The mazak I have now seems rigid enough but I a saw a NLX cut at EMO and it was impressively smooth. I guess they were cutting mild steel and not 4140 HT. That probably makes for smoother cuts..
In the end if Mazak can't come really close to the DMG price that will decide.

Thank's for the video Cameraman.
 
Celos is a powerful Computer....the touch screen can have some delayed response time, between touch and action.

I am sure DMG engineered Celos around powerful processing hardware, but if they don't take the time and attention to detail required for stuff like touch input processing and screen draws, it will feel like a slow piece of shit.

Hate on Apple, but Jobs and the folks there recognized very early on that a computer's instant reaction to inputs (mouse, keyboard, or touch) is absolutely critical not only for people to feel like they are working with something of quality, but also in the overall usability of the system. The original iPhone team has literally written books about the absurd lengths they went through (and continue to foster) so that the system reacts as instantly as possible to inputs. This kinda stuff is gospel now in the consumer facing tech world.

That a company the size of DMG Mori would basically ignore or be unaware of this basic tenant of UI design does not speak well to their leading the future of CNC machine controls.
 
I am sure DMG engineered Celos around powerful processing hardware, but if they don't take the time and attention to detail required for stuff like touch input processing and screen draws, it will feel like a slow piece of shit.


Personally! IMO! If you got used to it-you would. The problem with (guessing) .25-.5 second lag, is that the Operator doesn't know whether the "button" has been touched or not. So the automatic response is to press it again, but the lag in visual response of the screen, is not synched to the Operations I/O data.

If you follow what I'm saying. The Operator touches the screen-the data is processing in the CPU (yet the screen hasn't changed) press the button again and you are pressing the button on the next screen....NOT on the first one. Gaaaay!!!

R
 
We're satisfied with CELOS. We have three CELOS-equipped NHX machines with delivery dates spanning three years, late 2015 to present.

It's gotten progressively better. It's still not fast - there is still latency between screens, and you can't type as fast as you would on a smartphone, but it's gotten much better since the first version and you get used to it. I no longer see it as a problem, just something for me to whine about when I'm having a bad day.

The reality is that CELOS has saved me a lot of time. For every second I lose in latency, I gain minutes and hours elsewhere. I will make a video sometime early next year showing how we manage tool life. It's very slick and operator-friendly, and is not possible on most other controls.

As for the NLX, it's solid iron, largely unchanged from the original Mori Seiki NL models. Box ways on XYZ and linear guides on the subspindle. You can't go wrong with an NLX.
 
We're satisfied with CELOS. We have three CELOS-equipped NHX machines with delivery dates spanning three years, late 2015 to present.

It's gotten progressively better. It's still not fast - there is still latency between screens, and you can't type as fast as you would on a smartphone, but it's gotten much better since the first version and you get used to it. I no longer see it as a problem, just something for me to whine about when I'm having a bad day.

The reality is that CELOS has saved me a lot of time. For every second I lose in latency, I gain minutes and hours elsewhere. I will make a video sometime early next year showing how we manage tool life. It's very slick and operator-friendly, and is not possible on most other controls.

As for the NLX, it's solid iron, largely unchanged from the original Mori Seiki NL models. Box ways on XYZ and linear guides on the subspindle. You can't go wrong with an NLX.


Admittedly NHX machines are pretty damn awesome... So maybe Celos is not an issue to the fantastic work you do @Eric/OragneVice… But if that was happening on HAAS control one would never hear the end of it.


The NLX have made improvements (little by little) over the original NL's (in some cases).

But for more complex high tolerance/tight tolerance continuous profiles linear rolling element ways/ bearings tend to yield better results as you don't have to over come issue of static friction (stiction) on boxed ways.

That's maybe why DMG Mori came out with the ALX vs. the NLX.


The ALX is supposed to be cheaper than the NLX.

OP is looking at the NLX 2000 … The NLX2500 series usually is substantially more expensive than "equivalent" MAZAKs (different animal) especially for twin spindle and live tools + Y axis.


The NLX 2000 being smaller / slimmer ?? -ish maybe less $ than the vaguely equivalent MAZAKS.

_______________________________________________________________________________________________


I agree with the "Konig's" point about latency in the interface is a real "no-no" ; I've worked in VR systems for 27 years and we have to adhere to 60 (frames / second (stereo-3d) ) 1/60th of a second frame redraw rate and buffer swap; other wise folks have a bad time. Any real time system including GUI interface has to respond in a non-drunken way like the person running the machine, but maybe 4 Guinness-es would "Fix" the latency problem :D [I agree with Littlerob1 about latency and double or triple presses, not good for impatient individuals doing mission critical things.].

___________________


@OrangeVice
look forward to your next videos.
 
I can’t say right now but we’ll be receiving our NLX2500SY/700 in the next few weeks, can’t wait!

The next option for us was the Doosan Puma 2600SYII. Did’t get a price on a Mazak but sounded like they’re closer to Mori then you’d think. Mazak support in our area is not great from what I’ve heard. I was sold on the box way for us.


Sent from my iPhone using Tapatalk
 
Personally! IMO! If you got used to it-you would. The problem with (guessing) .25-.5 second lag, is that the Operator doesn't know whether the "button" has been touched or not. So the automatic response is to press it again, but the lag in visual response of the screen, is not synched to the Operations I/O data.

If you follow what I'm saying. The Operator touches the screen-the data is processing in the CPU (yet the screen hasn't changed) press the button again and you are pressing the button on the next screen....NOT on the first one. Gaaaay!!!

R

Yea, that was the mission-critical thing Apple discovered way back during the initial development of the Macintosh. They were creating a whole new paradigm to make computers accessible to everyday folks, and they quickly discovered that the huge part of that was that when users did something, the system had to react to their input *instantly*. Otherwise, folks get confused not knowing if anything happened, then they go fire off another command or click a button again. The timescale for this is measured in microseconds.

They went as far (and still do) to making the User Interface on a separate layer, where the graphical side of things is it's own high priority computing thread in the system and is disconnected from the underlying functions being called. A button can instantly change state, or a menu can drop down, or a new screen appear without the burden of the big lift happening underneath. All so the user feels like something happened.

This has been basic user design 101 since. Whole books are written about it ("The Design of Everyday Things") and entire niche fields of computer science deal with it. Yet DMG Mori, a multi-billion dollar global engineering firm in one of the more intense human interface design fields in existence, ignored 100% of all that.
 
<snip> The timescale for this is measured in microseconds.

They went as far (and still do) to making the User Interface on a separate layer, where the graphical side of things is it's own high priority computing thread in the system and is disconnected from the underlying functions being called. A button can instantly change state, or a menu can drop down, or a new screen appear without the burden of the big lift happening underneath. All so the user feels like something happened.

This has been basic user design 101 since. Whole books are written about it ("The Design of Everyday Things") and entire niche fields of computer science deal with it. Yet DMG Mori, a multi-billion dollar global engineering firm in one of the more intense human interface design fields in existence, ignored 100% of all that.

Yup, even Apollo guidance computer on the (Lunar Excursion Module/LEM) designed by HAL... (1966-71) Hal laning (J. Halcombe Laning) had what was called an 'Executive Waitlist" Operating System... Basically high priority tasks are executed first and low priority task executed later. J. Halcombe Laning - Wikipedia

These days CPU's are cheap as chips and wondered why the original MAPS style control could not be in a separate "box" / computer sneaked in on the back side (with cooling) and then have the Interface for that "windowed" (like an OpenGl window, via fiber optics) as a dumb-terminal onto the CELOS displays, that way all the original bomb proof speed and response of the MAPS control would be retained/maintained regardless of whatever peripheral "nonsense" or organizational software + industry 4.0 running concurrently would not get in the way?


I admit it's an "Odd" architectural mistake where interface response was not given the highest or higher priority ?


I'm sure as processors get faster and they (DMG MORI) get better at executing multithreaded code (from legacy sources) as Orange Vice reports things will continue to get better.


Perhaps the SW that runs on the CELOS for NLX / Lathes does not have such latencies as there's not that much computationally going on ?
 
These days CPU's are cheap as chips and wondered why the original MAPS style control could not be in a separate "box" / computer sneaked in on the back side (with cooling) and then have the Interface for that "windowed" (like an OpenGl window, via fiber optics) as a dumb-terminal onto the CELOS displays, that way all the original bomb proof speed and response of the MAPS control would be retained/maintained regardless of whatever peripheral "nonsense" or organizational software + industry 4.0 running concurrently would not get in the way?

CELOS is, like Mori MAPPS before it, a Windows based system on top of a 3rd party control (Mitsubishi, Fanuc, Heidenhein, or Siemens). The problem is either DMG Mori is writing really poor software, or (and I think this is the real issue), on a button press, the Windows layer is waiting for the control side to come back with a command execution complete signal before drawing reactions on the interface side.

What it should do is graphically execute the input (change the button state, or even move to the next window) while the control underneeth is executing the command. 99.999% of the time, there will not be a variance and it will feel like the whole system is operating in a much more responsive manner. If there is an error with the executed command, you can alert the user or throw up a message.
 








 
Back
Top