Hi,
I see a number of CNC users that do not think Mach4 is for them.
Yes there are still some outstanding issues like A axis not being able to maintain surface speed as the diameter controlled by the Z axis changes like Mach3 does.
Also Mach4 still cannot correctly display Toolpaths for 4/5 Axis simultaneous G-code as well as simple cylinders correctly like Mach3.
But both of these issues can be taken care of by the Mach4 programmers.
Apart from these issues Mach4 is an improvement over Mach3 in many ways especially machining very fine detail with small cutters down to a .05mm Rad tapered without any cutter damage using Hard brass.
Mach4 V3481 and HiCON are rock solid and can machine all 3 Axis G-codes perfectly, I have not tried 4 Axis as we need the above 2 issues included into Mach4.
We use Rhino/madCAM and Aspire to generate the G-code and both have no issues with Mach4.
Yes you may have to do a little Lua code to make it suit your choice of Controller Card and MPG, but that is no different to what you had to do with Mach3.
Both HiCON and ESS cards are in many cases a lot easier to set up in Mach4 than Mach3.
I cannot speak about the other controller cards as I only use these two.
Mach4 V3481 works fine with ESS and XP with 2GB Ram and with Windows 10 using HiCON.
Regards,
Mauri.

Hi Guys,
Just to add fuel to the fire, back in the Mach3 world you would have had to either use VB script (remember all the problems???) or Brains (Ever try to edit a long Brain???) to accomplish anything that was not included with Mach 3.

In Mach 4 you use Lua and the PMC. I don't hear a lot about people using the PMC but I just was able to connect 6 physical buttons on my lathe retrofit without know how to use LUA at all. LOVE IT!!!!!!!!!!!!!!!!!
It is a ladder logic program that compiles LUA script in the background.

My opinion is that Mach4 for a hobbyist is fine especially when you want to run a 3 axis machine or a 2 axis lathe. The hardest part for me was going from Mach3 to Mach4 I had all kinds of preconceived ideas about how it "should" work as opposed as to how it does work. Yes it has some idiosyncrasies but please don't tell me that Mach3 and LinuxCNC don't.

Mach3 has made a lot of parts for me over the years and I learned a lot making it do the things I want. The two major problems with Mach3 is that VB gets out of sync once in a while and I have broken my probe a couple of time. Problem two is that I get .XML file corruption after a few months of hard use and have to restore Mach3 from a backup. Once the space bar stops working as feedhold it is time to restore it. What else has stopped working properly that I don't know about in the mean time?

I have used LinuxCNC in the past and I am actually putting it on a machine as we speak because I don't have the electronics to do what I want in Mach4 and I currently do have all the electronics to do it in LinuxCNC.
I however don't like the look and "feel" of it but sometimes you gotta do what you gotta do.
I did however leave my self the option to upgrade to Mach4 in the future since I want to standardize on one control for all my machines.

I consider myself a high end hobbyist in that I have some large pieces of iron (quite a few of them actually) which are beyond what most would consider hobby size machines.

I am converting all my machines to Mach4 from Mach3 just because it works better for me.

Enough rambling! Time for more coffee and getting to work.

In Mach 4 you use Lua and the PMC. I don't hear a lot about people using the PMC but I just was able to connect 6 physical buttons on my lathe retrofit without know how to use LUA at all. LOVE IT!!!!!!!!!!!!!!!!!

Good on you Mike, that's awesome. Also a very good point I did not mention either. But you also either already knew or had to learn ladder logic. The question for the purpose of this topic is......... (and this is for everyone reading that wants to play along, not necessarily you Mike).

I know better but for arguments sake lets say you did not already know or understand ladder logic and had to take the time to learn it.

Would it be your fault for not already knowing ladder logic?
Would it be no ones fault and just part of the learning curve and even enjoyable as part of your choosen hobby?
Would it be Mach4s developers fault for providing the tool that gives you that option in Hobby?

If the answer you come to is option 3, let me know what options and/or features you think we should disable in Hobby. Or maybe we leave them in and it just helps us determine what questions to ask in a litmus test to determine if you really are a Hobby (realize doing your customization may take considerable amounts of your time and effort) or Industrial (wanting customization but don't have the time to do it myself but also willing to pay for it) user.

If Mach4 Hobby isn't considered Hobby material I'm sure we can make adjustments to fix that.

Hi,

Yes there are still some outstanding issues like A axis not being able to maintain surface speed as the diameter controlled by the Z axis changes like Mach3 does.

Hm, is this an mach4 issue? Shouldn't that be calculated by the cam process?

Mach4 V3481 and HiCON are rock solid and can machine all 3 Axis G-codes perfectly, I have not tried 4 Axis as we need the above 2 issues included into Mach4.

Hm, that's the point, where I became unsure.
I bought an ESS, but when I started configuration, I was astonished by the update-frequency of 40Hz, which already is 4 times faster than with mach3.
But ...
My cnc will be a 4-axis gantry, where I start with CL-steppers. The stepper-drivers handles the closed loop, but I'm interested in moving CL up to mach4. But I think, 40Hz is way to slow to start with CL.
Next doubt: I want feed-override in hardware - no matter whether to use potentionmeter or mpg - and I'm quite sure, that 40Hz is to slow for it too.
Therefore I wrote, that I miss decent motion controller, that ist affordable (for me less than 500 bucks).
But that's not an issue for artsoft - it's just coupled to the decision mach4 or not mach4, as any software requires its own motion controller and each is incompatible to the rest.

Reinhard

django013,
I have tested many CAM programs in the past.
I do not know of any CAM program that can output surface speed and the A Axis using the Z Axis as the radius to control the speed.
If you want a controller that can in all conditions you need to speed the $$, HiCON cards can do it all.
Regards,
Mauri.

Hi Mauri,

I do not know of any CAM program that can output surface speed and the A Axis using the Z Axis as the radius to control the speed.

I think, the radius in the picture is just a helper information for human beings.
CAM, as well as motion controller has a path with starting- and endpoint. Cutting edge of the drill bit is the Z-Axis, so CAM and/or motion controller need to calculate a path with possibly 4-axis moving. The surface speed is the speed working out that path. The path may be radius based, but need not be so.

If you want a controller that can in all conditions you need to speed the $$, HiCON cards can do it all.

Hm, if I get the picture right, Hicon does not raise closed loop to mach4. Closed loop works at motion controller level.
I guess, same would be possible with ESS too. Not sure, if that makes any difference to closed loop at stepper controller level.
Closed loop at application level may have an impact at path optimizer.
... and I did not see any special input for feed override.
So I'm not convinced, that Hicon offers better solution than ESS does. Hicon is just way more expensive.

Reinhard

Have a look at Pokeys

Hi Reinhard,

So I'm not convinced, that Hicon offers better solution than ESS does. Hicon is just way more expensive.

Reinhard

The Hiconn is not way more expensive.
With the ESS you need to add breakout boards. The Hiconn has them built in.
The ESS and most other motion controllers are 5v I/O. The hiconn is 24V.

The Hiconn now comes in two flavors, the 6 axis high speed version and the 4 axis 125 KHz hobby version for significantly less. Both of the Hiconn's feedhold almost instantly.
If you are building a machine that would have run fine of the PPort then the hobby version should work great for you. Mach3 peaked at 100Khz and most computers wouldn't go over 50 KHz.
The 6 axis version does 4 MHz IIRC. Great for digital servos.

I just picked up the Hiconn hobby version for my lathe and I am looking forward to single point threading as soon as I get my encoder hooked up. Not sure how many other boards can thread yet.

I also picked up a Pokeys 57CNC board and I am very impressed. It will be used to run a small machine that will have a laser on it at some point. The initial price was very fair but I will keep an eye on what breakout boards and other items needed to be added for full functionality.

Fwiw, you can add an MB2 breakout board to an ESS for $150 and have 24V I/O, for a little over 1/2 the price of the Hicon.
And the Hicon can get really expensive when you start adding on the options.

I've never used either, but from what I've read over the years, I think the Hicon is better supported, and a more "industrial" solution.

Already posted about cost of upgrading to Mach 4 and thought it appropriate to post about TIME.

The first thing that comes to my mind is when and if one should upgrade at all.
Only the individual  can determine when, recall my undertaker comment, one will upgrade in
the future.

- If currently Mach 3 does everything you want it to do then no hurry
- If Mach 4 offers you something you need and desire that is lacking in Mach3 then maybe  start the transition.
- If you are a newbie to CNC, then chances are you don't know what you really need, the  learning time may be about the same, you
   won't be biased by something you don't  know about in the first place, Mach 4 may be the better choice.
 
Matters not if you are a hobbiest or commercial from a practical point of usage.


Some idea of time to convert was already posted and it is subjective. A few things come to mind as someone who only fooled with Mach 4 some three years ago. I'll keep track of of time when I start looking into Mach4 just out of curiuosity. But being a little wiser these days I will assume some things and accept some things as follows:

- Mach 4 hobby works for every command that is in the current manuals and accept the fact  that there are differences in the
  commands and they agree with books by Smid.

- The motion controller can provide for the instructions given it by Mach4 for every  command. Understand that I just don't have any
   tolerance to waite for "second party"  to  make their controller functional, said differentlt and candidly, have no tolerance for
   waiting years for some command to be implemented!

- Documentation must be available and complete for the above two comments relative to a basic mill or lathe. Currently there are 10 
  Mach 4 documents available for Mach of  which one probably needs to "study" 5 of them. Two are basicaly gcode programming
  and the time to transition should be relatively short for me. The gcode manuals  are not  meant to teach you how to write / program gcode and if  one deisres understanding surely     you will  find it in the Smid series of books. Smids CNC Handbook is 500 pages, so figure   around 30-40 hours to read it, then another 120 hours to study it, or take a two courses at a  tech  school for say 20 weeks! The TIME required to learn gcode is just a matter of degree of the level required to the what you need to do. And yes, if you don't use it you will loose it, but the refresh time certainly is shorter.

- I would say allow an hour for each for the Mach 4 manuals to get a "flavor" and can't really   say anything about an external  controller manual. If the status of completness is not there     then it could be a show stopper for me personaly at this time. Hmm ....the manual for Mach   3 mill took  6 months to do, and yes have done a few so I understand the investment of     time for a manual.
  BUT
 After all these years Mach 4  should have good manuals and there is NO reason not to!

- If you are more advanced and will desire screen customization additional  time frames     come into play. Any one who has done a
  custom screen, even a specific custom page using  VB in Mach 3, and knew how to use a screen designer, knows the effort required.
  I can only "assume" that the current screen designer is still rather intuitive and the transition won't take to much time. It will be just
  like learning a new CAM program since you need to  use the new tools that are provided.
  SO
  No use bitching about Lua ( I think it's still something they do in Hawaii ). 
  That's the tool necessary, and as Graig pointed out, one will need to understand Mach 4's structure.  Thus the user will need to invest
  time into knowing what they want to do, learn a new screen designer, study / learn the new structure, and lastly be able and willing
  to transition to a new programming language.  This is the one that reflect's back onto "Mach 3 doing what I want it to do". I
  personally will need to invest a fair amount of time just getting   Mach 4 Hobby to do replace what I can do in Mach 3.

I must say the following and mean no offense to anyone.
If one does not want to invest the TIME, good grief READ or STUDY, then please pursue a different hobby, since  a few parts of CNC are NOT  plug and play!


Just some thoughts about Mach 4, my other hobby is calling,

RICH