I use Dynomotion Kflop its $249, if you want analog 0-10volt go for the additional kanalog board for an additional $249. I've used it for both mill and lathe. Great software too but you have to learn C code. Used it for years, love it.

http://dynomotion.com/

Hi Craig,
UCCNC
PathPilot Mesa card
PathPilot LPT
LinuxCNC Mesa Card
LinuxCNC LPT
CommandCNC
CSlabs
PoKeys
Centroid Acorn
Centroid Analog

I could go on but what would be the point (;-)

Hi TP,
UCCNC                      UCCNC is not a feedback controller, neither are any of the UC**** boards ergo not PID capable
PathPilot Mesa card    No idea what PathPilot is.....is it realtime PID capable?
PathPilot LPT              See above
LinuxCNC Mesa Card   LinuxCNC is realtime capable and can be programmed to do realtime PID....if Linux is your thing
LinuxCNC LPT             See above
CommandCNC            No idea what CommandCNC is.....is it realtime PID capable?
CSlabs                       Yes CSMIO/A is realtime PID capable at 600Euro, shame about the buggy/lacking Mach4 plugin
PoKeys                       PoKeys has an on board API that allows for proportional control but cannot do PID
Centroid Acorn           A rather more expensive albeit complete solution....not Mach
Centroid Analog         See above

Two that should be on this list are:
Hicon                       With the activation to do PID $1200, good Mach4 plugin
Gallil                        Starts at $2000 for three axis, recently released Mach4 plugin.

In short there are only three boards that allow genuine PID with Mach, CSMIO/A, Hicon and Galill....all of then way
more expensive than an ESS.
LinuxCNC is capable if Linux is your thing...and that debate is as old as hobby CNC itself.

Craig

A useful list to know about, but what are the costs for a full implementation?
Dynamotion seems to be about $500 for the two cards needed.

How many of these handle the standard NIST g-code without lots of C programming? It's not hard to learn the basic g-code commands (G0 - G3), but learning C would be a demand too far for many hobbyists.

And just how 'full' are the various implementations anyhow? For instance, does the Acorn system offer as much as a Mach3/ESS one? And do any of them offer the equivalent of the ESS threading 'gearbox' which comes with the latest driver?

It would be a valuable service to us all to have this detailed in full. Sure, I use Mach3, but I am still interested in what alternatives there are, how they compared in functionality, and their full costs.

Thanks Craig for your info.

Cheers
Roger

Hi TP,
have been doing some research, PathPilot and CommandCNC are variants of LinuxCNC and are very probably PID capable.

Craig

Roger,
 Now that you have high confidence in the threading may want to play around with the
different thread cutting feed methods. You may even be more surprised on the results of
your systems ability to do threads. Example; Alternating flank infeed to do a acme thread
with a single point threading tool. HP requirement is greatly reduced!

Just for info:
Back many years ago tried to get slaving of the spindle to the Z axis.
We had enough problems just fixing the Mach3 threading when it was broke.
So never happened.
Why the slaving? Because back manny years ago a DOS controller allowed that, and using an mpg
one could just move the Z at varing feedrate along the stock, back and forth, to do
threading. It was a model makers dream since you could easily do threads way smaller
than 0-80's.

Matters not how good  a controlling device addresses threading. If all the other items
of the lathe system  ( which includes the operator ) is not applied correctly for the
task success will not occur. The nut goes on has it's place, but, meeting design intent
and spec are the name of the game.

Have fun,
RICH

Hi Roger,
on the Acorn website I came across this in relation to their step/direction controller:

•Spindle encoder input for CSS, Threading and Rigid Tapping

It would suggest that the encoder has A,B and I channels enabled to do rigid tapping. The controller is not PID capable overall,
I suspect it has a partial PID control of the spindle only in similar manner to the ESS.

Craig

Hi Roger,
Acorn do two other much more advanced models which are feedback capable:

AllIn1DC at $2250
OakCNC at $1600

Quirkily the AllIn1DC is not step/direction capable and therefore cannot drive steppers or AC servos. The CSMIO/A has he same
problem, sure you've got nice analogue outputs its just that you cannot therefore have step/direction.

The OakCNC cannot drive steppers either, according to Acorns feature list but can drive AC servo drives, I suspect the controller
is analogue output only and therefore the AC servo drive solution is based on analogue input rather that step/direction.

The much cheaper Acorn is $300 plus $150 for the entry level software other than the severely limited free software. It looks
to be a capable controller very much of the ESS ilk and approximately the same price as Mach/ESS.

The other two models are way way more expensive but they are genuine feedback capable controllers.

This supports my contention that 'genuine feedback controllers are way more expensive than open loop step/direction controllers'.

Craig

UCCNC  does a fine job rigid tapping and encoder threading for turn (;-) Does perfect guaged threads  Even does Peck tapping LH OR RH and uses BOTH channels of teh encoder plus index ABZ.

It is probably teh EASIEST controller to set up rigid tapping or threading. ALMOST plug and play.

Just a thought, (;-) TP

So why do I need PID again (;-)

Hi Rich

I have just reread your excellent Mach3Threading.pdf file . Some wonderful stuff about threading in general in there.

But ... it was written in 2009 with Mach3 version .032 and an LPT in the PC. This does not really represent what most new users will have today. In particular, the transition to an external pulse engine (such as the ESS) and the possibility of an electronic gearing is not considered. Is it time to consider updating it?

Also, I have to mention that I am not really happy with the m1076.m1s program version I have: 9/Apr/2015. Watching the threading very closely, it seemed that some passes did a significant cut while other passes did very little. I have looked AT the code in the program, but not in depth.

Cheers
Roger