Machsupport Forum

Mach Discussion => General Mach Discussion => Topic started by: Raychar on December 21, 2015, 10:50:18 AM

Title: More inputs and outputs needed in Mach3
Post by: Raychar on December 21, 2015, 10:50:18 AM
Hi everybody,

For my cnc machine, parallel port 1 is used for 3 axis and other machine facilitates use. While I built an electrical auto toolchanger, it consumed all the I/Os of the parallel port 2. Now, I built a 5th axis as well,  I don't think there is another port for my use, is there any way to solve it?

Thanks in advance.
Title: Re: More inputs and outputs needed in Mach3
Post by: robertspark on December 24, 2015, 03:11:18 AM
UC300
Smooth stepper (Ethernet)
Cismo
Purelogic plcm-e3
Modbus (depends how fast you need the IO's read)
Title: Re: More inputs and outputs needed in Mach3
Post by: Raychar on December 24, 2015, 08:38:45 AM
Thanks guy, I have little knowledge in your mentioned devices, may I know the following:

>SSE -how many I/Os totally it has? Is the version for bi-polar stepper motor of it available in market?
>Modbus -After intalling it, will my current connections in two parallel ports be affected?

Your kindly advise will be much appreciated.
Title: Re: More inputs and outputs needed in Mach3
Post by: robertspark on December 24, 2015, 04:11:33 PM
I will do them one at a time for you over the next day or so (Christmas... other chores....)

UC300:
Manual here: http://cncdrive.com/MC/UC300%20datasheet/UC300%20users%20guide.pdf
Has the equivalent of 5 LPT ports (two are I/O Standard, 3 are Input only), also has 2 Analog Input and 2 Analog Output 16 bit.
- Controls upto 6-axis simultaneously.
- Works with Mach3 software via plugin.
- Upto 100kHz operation.
- 49 digital inputs and 36 digital outputs.
- 2 analog inputs and 2 analog outputs.
- USB connection to the control PC.
- LPT port compatible.
- Fast communication with data buffer for robust and stable operation.
Mach3 plugin only (no support (yet!) for Mach4), maybe in the future.

Personal comments: I would have bought one... but I had a USB Smooth Stepper already.... and I ended up with the dreaded earth loops issue ... and USB on a plasma cutter is not good for noise immunity.  Other than that it looks like a nice bit of kit.   I know someone with a UC100 who is very happy with it (but it is only one LPT port.... and you need more), and he uses it with a plasma cutter and has no issues.

Rob
Title: Re: More inputs and outputs needed in Mach3
Post by: Raychar on December 24, 2015, 11:07:47 PM
It is an very attractive product, thank you for your information. I will dare to buy one and try to use. But, for its software, I still hesitate as I don't know if it was fully matured and developed as Mach3. Is it really capable to have all functions performing in real time against what is in Mach3 having 10th second time lagging in some? Will it be be in most efficient stage when they are come along and work in combo set, I mean its hardware and software...?

As I remember, I came across the nose interference situation some years ago when I made my diy the edm wire-cut machine. Eventhought how shieldings made in the cables and housing case, the position DROs still have errors and very annoying. I am now busy in doing cnc auto-toolchanger, wireless auto-probing. Later on, when I have time, I will get back to try to resolve it....

Title: Re: More inputs and outputs needed in Mach3
Post by: robertspark on December 25, 2015, 04:30:55 AM
Smooth Stepper:

website: http://www.warp9td.com/

Two variants, USB (USBSS / USS) or Ethernet (ESS)
Product Comparison here: http://www.warp9td.com/index.php/products/#details

Note the rest of this is biased as I am a user of the ESS and active on the forum.

In short, ESS = 3 x LPT ports, USB = 2 LPT ports + a differential port (which can also be used as inputs).
Both have an expansion port.... which has not really been developed publicly since they were both launched / sold, there is an aspiration to provide an SDK, but that is my opinion only.

There is a lot on information on the official website.  The PDF manuals are some years old, but still relevant, and
The ESS has a more developed driver for Mach3 than the USBSS.
The ESS has backlash compensation, THC and M10Px/M11Px support, the USB does not.

Because both are FPGA they are reliant upon the plugin to provide the functionality, and both have good potential for really good development, but the development cycle is unfortunately slow (updates of plugins tend to be on about a yearly (roughly) if you have a look at the plugin release dates on the website).

I bought a USBSS but did not use it for years (made the usual mistake of buying the moving bits before building the machine....), when I came to use it I struggled with the ground loop issue that the USBSS is known for, and at that time the website was not very informative, it has improved drastically over the last year, then I found out that Torch Height Control (THC) was not available via the plugin and backlash compensation and I re-evaluated.... Ethernet SS or something else... after much deliberation settled on the ESS because of the ethernet connectivity (better noise immunity, longer transmission distance), M4 plugin being developed (or at least an official statement to that effect at the time), CS-Labs CISMO came close.... but cost was higher (although it does not need a breakout board..... in hindsight I probably should have gone with that probably because of things like threading which I may want to add to my lathe later, and i've probably spent as much on breakout boards (but there are cheaper ways to provide easy to wire terminals via LPT BOB and DB25 <> IDE 26 ribbon cable that what I did).

SS has a lot of potential, but is reliant upon the developer (as most motion controllers are especially if you want more than one LPT port equivalent)  ..... Pokeys has some good potential but it may not have enough IOs for your too.

Third choice came the PureLogic PLCM-E3 but it was about the time that Russia annexed Crimea, and there were some restrictions starting from the EU, their products seem to have since disappeared from ebay and one of their suppliers in the UK so I suspect that their products are being restricted by the current political climate.... but still a good product but only Mach3 support (at present) and it requires their own addon module for torch height control (extra cost from them).   Comments on the forum of their products has been good and any issues identified have been sorted very quickly (which has not been the case with some motion controller developers).


You asked two questions on the ESS:
>>SSE -how many I/Os totally it has?
Answered above.... 3 x LPT equivalent ports, with ports 2 and 3 configurable as all inputs or standard LPT type I/O arrangement.  (forget about the "expansion port" as its not available unless you set about to programme the FPGA from scratch).

>>Is the version for bi-polar stepper motor of it available in market?
Not sure if I am understanding this question... the ESS provides step and direction signals to 6 axis (+ spindle) stepper motor divers and the stepper motor drivers must control the stepper motors (serial or parallel wiring configured).

The ESS seems to be by far the fastest motion controller on the market as to how fast it can driver stepper motors.... although I'm not aware of anyone actually being able to use it that fast...(up to 4mHz!)... I'd be surprised if you were stepping a stepper drive over 100kHz (maybe servo drives.... never used them).


If you are interested in the SS, have a look at the website, and register and log on to the forum to check out the issues, as some elements (threading) have been an issue for some users for years, and also the common user issues are discussed (ground loops, look ahead, axis step rate / frequency, and velocity update frequency).  Would hate for you to become someone waiting for something to be developed or sorted out that you did not know about up front. 

If you have a basic milling machine or lathe or plasma or laser the ESS will more than likely meet your needs driving up to 6 stepper motors, on a windows PC with Mach3.   Potential for Mach4 is there with the initial driver release, and another one due soon (been imminent for a while now) ... but then there may be a timeout as they want to develop the USBSS plugin for M4, and then maybe other bits before catching up with ESS development on M4.


Title: Re: More inputs and outputs needed in Mach3
Post by: Raychar on December 25, 2015, 01:02:09 PM
They are very informative......

ESS seems to be better choice in functionality than USBSS. However, in term of ease of use in setting up and connection, usb one is better because it may be a bit difficult for an inexperienced one like me to set those IP address, net mask ....For an usual cnc milling machine user, the lack of functions of blacklash compensation is not a big deal. I used it, it cannot really bring a good precision...or THC, not need at all, so the choice between them is still in question...

I totally agree that the thought in buying a new electronic devices. Don't buy them in advance, buy them on demand and after going through the research. Software support and company background are of important. So is the success of Mach3.

I recall that I want to learn ARM microcontroller some eight years ago. I spent a certain amont of money in buying the development board, screen...but with limit amount of support from the company and forum, I end up in quitting the learning. Mach3 brings a lot of joys to diy guys...l can design the screen with buttons to control external devices like my electrical ATC, I can further program a microcontroller to do many actions...Today, many devices are cheap, I used to make BOB, axis DROs and stepper motor driver on my own. If I do them now, I will pay more in it than in purchasing them, but in the course of making, l learnt the technology....
Title: Re: More inputs and outputs needed in Mach3
Post by: BR549 on December 25, 2015, 02:15:13 PM
Just becasue you shielded them does mean you will not have ground loop problems . Ground loops are teh biggest probelms with CNC as you are interconnecting many differnce electrical schemes into one coorrdinated system. AND most boards are NOT designed to avoid ground loops and MOST have no idea what they are or what havouc they can create with electronics.

Most constantly confuse ground and common(DC) when they assemble a machine . They are 2 totally different critters.  

It is funny that some OEMs constantly have Com problems with USB and some NEVER have com problems with USB (;-)  

(;-) TP
Title: Re: More inputs and outputs needed in Mach3
Post by: beefy on December 25, 2015, 05:15:03 PM
It is funny that some OEMs constantly have Com problems with USB and some NEVER have com problems with USB (;-) 

(;-) TP

I think the USB Smoothstepper has created this impression that USB is automatically prone to problems with noise, etc, and ethernet is the fix. However I've came across a few threads in the past where people have had plasma machines (and other types) using USB motion controller cards and they've never had any such issues, and everything has been rock steady.

So it seems the individual designs of the USB determine whether it functions perfectly or not, and USB itself is not the root of the problem.

Keith.
Title: Re: More inputs and outputs needed in Mach3
Post by: robertspark on December 25, 2015, 05:41:52 PM
I don't think USB is the root of the problem... really it depends upon many things and one of which being .... type of plasma cutter (or at least the type of plasma start) as well as the proximity of cabling to the HF stuff and grounding of the table etc etc.

I know what my issue was and it was basically grounding the USB connector on the ESS to my driver box. (nope I have nothing against USB), but I did want THC and the USBSS did not offer it and neither did it offer backlash compensation (which although I don't use it, I have a few mills and a lathe which I was hoping to use the same driver box with).

I have an older hypertherm, and at the time was not sure if I was going to go with a bigger Chinese HF start unit (since decided to stick with what I have and wait for a bigger hypertherm when one becomes available second hand. as they offer better start characteristics (blowback start) and longevity of consumables and spares will meet my needs for the future, plus there are other things link machine control interfaces and RS485 interface on some models etc etc.

I really wanted to stick with the USB SS at the time as I got it working, but was disappointed that the driver was not being updated in the short-term in line with the ESS.  I have since found out that they have many more ESS units than USBSS boards in circulation.... hence ESS is having its drivers updated first to M4, then USBSS to M4.... then porting ESS M3 to USB M3... and time / speed of updates was my personal issue.

All I was trying to do was go through the various boards with more than 3 LPT equivalent ports that I knew of which is what the original poster seemed to ask for a bit of elaboration on my first post... plus add some of the bits I' found on them.

PMDX http://www.pmdx.com/MotionControllers do some nice motion controllers, driver updates seem to be as required, finger on the button from those lads, and they are usb... but just not enough pins for the original poster.
Title: Re: More inputs and outputs needed in Mach3
Post by: robertspark on December 25, 2015, 06:08:37 PM
CL-Labs CISMO
Website: http://www.cs-lab.eu/en/sklep-dzial-2,-,-CNC_Motion_control_systems_CSMIO_products_sets.html

3 x Offerings, really, in your case of I/O's you would be looking at the CSMIO/IP-S
If you scroll to the bottom of this page there is a comparison of all of the devices:
http://www.cs-lab.eu/en/produkt-2,2-CSMIOIPS___6_axis_Ethernet_Motion_Controller_STEPDIR_with_connectors.html

Costs, module options are here: http://www.cs-lab.eu/en/sklep-dzial-2,-,-CNC_Motion_control_systems_CSMIO_products_sets.html

Manual here: http://www.cs-lab.eu/en/upload/pdf/User_Manual_CSMIOIPS_EN_v2.1.pdf

personally I really like them, and nearly bought one, what slowed me down... cost... but then I did not account for my final break out board solution.

What I really like... ethernet (same as the ESS), and 24V input.... much less prone to noise and interference (especially when running a plasma).

I may still buy one depending upon M4 development of the ESS.  At the moment there seems little point to jump ship when a new M4 plugin is upcoming for the ESS.   Also suffering from "sunk cost effect" considering the ESS cost + bob costs.

What put me off (at that time).... torch height control... they are basically running it using the Mach3 THC plugin and using their analog inputs ... what you have to do is get (or build) an isolated voltage divider circuit to read your torch voltage in simple terms.   I have designed my own THC, just need to get around to finishing it, and now understand a whole load more than I did when I was shopping around for an USBSS replacement.  THC with motion controllers can be problematic a) it's not implemented, b) its rudimentary, c) requires other hardware from the same manufacturer.

One of the concerns with THC and using the M3 plugin is the M3 loop speed.... on a fast plasma cut M3 will lag quite far behind in mm before responding.

Plus looks like they have issued their M4 plugin in line with the latest M4 update at this time (2803)

May not have implemented M10/M11 (yet either, but believe that it will be implemented in M4 plugin) older post here on it (but does not really apply to the original posters requirements from the above responses). http://www.machsupport.com/forum/index.php/topic,30020.0.html
Title: Re: More inputs and outputs needed in Mach3
Post by: BR549 on December 25, 2015, 06:43:46 PM
You also need to understand is the ESS has larger memory and I think A larger faster processor. THAT can make a BIG difference in what you can have programmed on the motion unit itself.  With an outside controller the controller has to be the Big Brain and do ALL of teh actual motion through pulsing ect AND do limits homing,probing. The PC just becomes the basic  planner.

Commercial CNCs have been using that scheme for decades Most also run Windows as the user interface and either use Ethernet or Fiber ethernet for teh Com link. Fiber link is also nothing new for CNC it has been around for a long time now and very stable. When the USB SS  was first being planned I talked to Greg about fiber link (;-). Other controllers had been using it for years even some DIY units.

The New UC400 ETH  from CNCdrives is reported to have a  large memory and a wickedly fast processor and ethernet with their OWN com protocol that makes the Com side almost real time in nature. Contrary to popular believe even real time controllers have a buffer (;-)

Also the old MsDOS 6.0  made a nice controller and it was real time as well.  There are STILL modern Commercial controllers that run MSdos as it's OS.

Just  a thought, (;-) TP
Title: Re: More inputs and outputs needed in Mach3
Post by: Raychar on December 26, 2015, 12:16:13 AM
So traditional or old things may not be no good, like LDPT. If I am not run out of I/Os, I would not intend to use the other additional plugins. Noise interference is very problematic and root causes are needed to be addressed in first hand, I would do it later ......My cases are..

My diy sinker edm machine releases a 600w power in a spark which is at high frequency. For the diy circuit boards, diy housing compartments, even the diy shelding cables and a large number of wrap wires inside, it is not unlikely has interference issue. It works in LinuxCNC on a desktop computer. Usually, computers are designed with better noise immunity which have multi-layer boards that includes a ground plane whereas my diy boards cannot have. However, the big spark also caused the computer screen disappeared a second, I didn't measure and know if the cnc portion, that is the motion path had the affect or not at that time.

My cnc router, on the other hand, is a completely European-factory made. The complete set came along with all cables, I only added a second hand old type desktop computer. It locates at my room. At the early time of usage, this happened. In the midst of running codes and cutting. Wenever I swiched on the room's ceiling fluorescent/neon? lamp, it went wrongly and the workpiece and cutting tool were ruined. From that time onwards, I got an habit to pause the running codes before doing any switching of my room's light. As I tolerate it, I don't do any further investigations and fixing that issue. The stepper controller box was a  well-built one with ac line filters, only small size Nema 23 stepper motors, a only 1000W Kress spindle, computer runs Windows XP on a AMD CPU.....who know what's the problem!
Title: Re: More inputs and outputs needed in Mach3
Post by: robertspark on December 26, 2015, 03:03:18 AM
UC400 ETH sounds interesting, there other boards seem well supported (although m3 only).  Can't see anything on their website.
Title: Re: More inputs and outputs needed in Mach3
Post by: robertspark on December 26, 2015, 04:54:32 AM
Raychar

Other options could be arduino or Pokeys u57 as well as modbus.

You have two LPT ports which you can consider as your high speed IO's now consider to free us some space on the two LTP ports if you cannot remove some of the lower speed stuff to use arduino or Pokeys u57 as a keyboard emulator to use oemtriggers to do some of the more mundane stuff like coolant pumps, fans or whatever else you have that only requires 10hz.

This would be similar to modus and using brains.

Have you got a pinout list of your two LTP ports of what you are using each pin for?

Rob
Title: Re: More inputs and outputs needed in Mach3
Post by: Raychar on December 26, 2015, 08:33:07 AM
Rob,

>As mentioned before, my second LDPT is for the toolchanger box use. Since there are a lot of hardware keys, Leds, limit switches and two d.c. motors to be controlled, a Microchip MCU is installed there. There is program, together with V.B. program in marcopump, they colaborate with each other to do logical functions/control on those hardwares. It demonstrated to work well. However,  in the very near future, I will make the version two. In that, less amount hardware keys will be used, each manual tool holder selection buttons still preserve but thet will be resided to Mach3 screen as virtual keys and leds.

For five tool holders which have their individual sensors, 5 inputs. Two motors, one is for door cover which will be sensed bytwo limit switches, another 2 inputs and 4 outputs(each motorneed to be controlled in forward and backward directions). The lock motor needs 1 input more for some special function as well. In that configuration, I intend to delete the MCU in the toolchanger box and to solely do the control by marcopump program...

> The first LDPT has some ports for coolant pump and spindle controls, when they are controlled by codes and run in M-codes, they are not supposed to resided to the external control by Arduino nor MCU. Pokeys is not my choice at the time being, it says it can be easily programmed graphically, personally, I like to do them by C in MCU or Arduino that I learned, at least, I know every inside them...
Title: Re: More inputs and outputs needed in Mach3
Post by: Raychar on December 26, 2015, 09:18:32 AM
For the noise interference, I also recalll that...

> when I designed a small electronic item which involved the using a MCU, I was told that different brands have different level of interference susceptibility. Even a substitution one which has exactly the same function and which can directly run same program code, they exhibit a slight difference in this level. A tester which output an in-rush cuurent was demonstrated and proved at that time. That told me layout inside the I.C. is of important. Of course better IC didn't resolve the problem all. I can say only 20 percent roughly, another may be the circuit, pcb layout designs....Later on, a small a.c. line filter (choke coil and capacitors) was added. By using the cheaper grade substituted MCU, the product passed the test and were shipped out.

> for a diy project some time ago, it involved the using a real time clock I.C., I run it in dry battery or d.c. and it posed no problem at all. But it failed in rectified d.c. (from a.c. line)...even though how well I made on the rectified circuit. The I.C. is not a new one and widerly used in a.c. operated products. This may tell me that a same type I.C. which is of different batch in production or there is minor malfunction inside, they may affect its interference susceptibility.
Title: Re: More inputs and outputs needed in Mach3
Post by: BR549 on December 26, 2015, 12:39:44 PM
I would consider moving all teh axis control over to a LPT port the use teh left over pins for other I/O control or another means of I/O (modbus,etc) for the non critical I/O.  Doing the Tool Changer as a PLC is always a good idea as then you just talk to teh PLC to request and confirm a ttol change and not use up valueable port/ pins.

Then you can consider moving to an outside controller as is mentioned that way with a controller like teh UC300 you have 5 ports plus analog in and out to have available.
Title: Re: More inputs and outputs needed in Mach3
Post by: marting on December 26, 2015, 02:00:55 PM
Hi to all.
I've been reading this post as I have build my CNC router 10 years ago with Rutex servo drivers and like you have had these problems, particularly interference. The most important thing is to ground ONLY at one end of the shielding, don't connect or bridge from another end and the most important is to ground at the power supply. This has been my experience both as a CNC hobbyist and as a professional at work.

I too am planning to build my second CNC and since Rutex (the place I bought my servo drivers) are no longer functional, have been looking at CNCdrives, from what I see and read, these look sturdy, however have not seen much posts on the forum of how others who bought are doing with them.
Title: Re: More inputs and outputs needed in Mach3
Post by: BR549 on December 26, 2015, 05:40:16 PM
I know several people using their servo drives and they love them. VERY dependable.

The CNCdrive  controllers are ALSO supported in their OWN CNC software Interface  > UCCNC , also very nice to use, very easy to setup. It is NICE when the Same people design and build both the Controller Hardware and Controller software . Hard to be able to make any excuses as to why things do not work properly.

(;-) TP

Title: Re: More inputs and outputs needed in Mach3
Post by: marting on December 27, 2015, 07:06:40 AM
I've been trying to get in touch with CNCdrive by email for the past week now but no answer or email receipt confirmation. I know it's the festive season and that they might not even be at the offices but to be honest I am starting to question if these are real or not. At least they could have done an "Out of office" notice.
Any idea what would be the delivery times from date of order and what is the after sales support looks like (i.e. is it ok)?
Thanks
Marting
Title: Re: More inputs and outputs needed in Mach3
Post by: robertspark on December 27, 2015, 07:14:39 AM
Some businesses in Europe may shut down from last Friday 18 dec, up until 4th jan.

You can buy from other suppliers.

Think they are in Hungary, so patience is sometimes required, especially around this time as the mail service goes slow too.
Title: Re: More inputs and outputs needed in Mach3
Post by: Raychar on December 27, 2015, 01:06:36 PM
> right! It makes sense to have all axis to be allocated to the two LPT ports.

> by the way, if a friend of mine want to start cnc milling, what I suggest himt to buy, Mach3, M4 or UCCNC(w/the drive controller). ..its seems M4 still has some imperfections in plugin portion &.., will it be solved in very near future?  As i haven't use it, it claims that its core structure/design is completely different from Mach3, is it really better? The first two softwares are universal type, the last one seems only applicable to its controller...