[Autocraft]

Hello, I'm a newbie in everything cnc and I've been reading around the forums to learn everything I can.

I'm hoping to learn enough to convert our mill to cnc, make a cnc plasma cutting table, a cnc router, cnc tube notcher, etc.

Some quick questions tho. I've picked up a parker digiplan pdx13 along with the accompanying step motor, ball screws and linear guides. All dirt cheap too. They came from a factory machinery that is being sold for scrap. There are about 50 more units available if I want them but I'm still not quite sure I can make use of them.

So my question would be: Can I use them? I read in the pdx13's online manual that up to 7 units can be series together. They plug directly into the parallel port of a PC right? Does that mean I can have up to 7 axis if I so desire? How about compatibility with cam softwares?

Sorry for the questions. I know I can probably find the answers to some of them by reading more from the forums but the pdx13 units are very cheap but are only available for a limited time and I haven't learned enough to know if they'll be useful or not.

[Autocraft]

Um, no replies?

I've attached a picture of the unit just in case someone will recognize it.

Anyway, here's what I've learned from my research, please do correct me if I'm wrong.

The parker pdx13 units interfaces to the pc by the serial rs232 using tx/rx/ground and uses a command language called x-code to control the motion of the step motors.

On the other hand Modern cam software like mach3 uses the printer port to interface with the drivers. It uses 2 pins per axis to control the step and direction of the step motor.

Did I got it right? If so, I guess the next question would be how to make the pdx13 unit work with a cam software like mach3.

These are the solutions that I've thought of in my limited knowledge.

1) Make a program to emulate a printer port that's recognizable to mach3, read the signals and convert them to x-code and send it to the rs232 port to the pdx13 unit.

2) Modify the pdx13 units directly. Have them be able to read the step and direction outputs from the printer port of the pc.

Are these solutions feasible?

[Mariss Freimanis]

Any step motor drive serial port interface will be proprietary ($$$). See if it can be bypassed to use a step/direction interface instead. If so, you are in luck; Mach3 or anyother CNC program can run the drives. If not, seriously consider buying drives that can. Your alternatives using what you have will not be inexpensive.

Mariss

[Autocraft]

Any step motor drive serial port interface will be proprietary ($$$). See if it can be bypassed to use a step/direction interface instead. If so, you are in luck; Mach3 or anyother CNC program can run the drives. If not, seriously consider buying drives that can. Your alternatives using what you have will not be inexpensive.

Mariss

Well, I've just noticed from rereading the pdx manual that it has outputs for both the step and direction. I've attached the page for reference. Is there any way I can tap these output connections and use them instead as input for the step and direction signals from the printer port?

[Autocraft]

Well, I tried plugging the step and direction output ports to the printer port and tried controlling it with mach3. It didn't work, the step motor just turned erratically in both directions. I did however got it to work using the rs232 port and the x-code seems straight forward enough. I think it'll be possible to make a plug-in for mach3 along the lines of Grex to control it. My programming knowledge is quite limited but I guess this seems the most plausible way of doing it.

[One of Many]

I'd snag what you can and unload them for something better suited for your application. These drives incorporate the indexer and take commands internally for simple linear motion in a macro canned cycle. Rather slick if all you need is a preprogrammed sequence on a single axis and very little external support. Reprogramming via the RS232 is a snap as product sizes change, but the loop remains the same.

At a minimum, you may be able to set up different programs to simulate step and direction via triggers, but it looks a bit dicey and more commplex work than a standard compumotor setup as in a Zeta4 or S-6 microstepping drive. These are compatible with your same motors and can go for very reasonable prices on ebay.

Still a nice score if you can turn what you can't use into something you can, all the while helping someone else get what works for them.

DC

[Autocraft]

I'd snag what you can and unload them for something better suited for your application. These drives incorporate the indexer and take commands internally for simple linear motion in a macro canned cycle. Rather slick if all you need is a preprogrammed sequence on a single axis and very little external support. Reprogramming via the RS232 is a snap as product sizes change, but the loop remains the same.

At a minimum, you may be able to set up different programs to simulate step and direction via triggers, but it looks a bit dicey and more commplex work than a standard compumotor setup as in a Zeta4 or S-6 microstepping drive. These are compatible with your same motors and can go for very reasonable prices on ebay.

Still a nice score if you can turn what you can't use into something you can, all the while helping someone else get what works for them.

DC

It does have the ability to store 7 sequences ( w/c incidentally gave me a lot of trouble setting up the communications because the default setting is having the 7th sequence running causing the sequence commands to echo to the pc) but the indexer does accept real time inputs.

X-code is composed of simple ascii commands that the indexer seems to interpret in real time. A sample code would be

1A10 // set acceleration of the 1st pdx to 10steps/sec (2nd pdx would be 2A10)
1V20 // set velocity to 20 steps/sec
1D50000 // set steps to 50,000 steps
1H+ // set direction to clockwise
1G // execute motion

So I was thinking it's possible to interface the pdx's to mach3. We can set the acceleration and velocity using the mach3 values,then use the address number of the pdx for different axis like 1 for x, 2 for y, 3 for z, etc. The pdx also has inputs for limit/home switches and even triggers that can be controlled thru x-code.

One problem is I live in the Philippines and ebay transactions in the past have been a pain in the ass. Shipping is also a problem. fedex is usually too expensive while the local post office is a den of thieves.


I've looked for zeta drive/controller on e-bay and they cost $200-400 each, very much out of my price range. Would have gotten some geckos for those prices instead. The pdx13 + step motor just cost me $10 so I'm kinda inclined to try make them work before I opt for the more expensive stuff.

[One of Many]

I read about that ability in the PDF also. What was not clear is if you could interpolate individually addressable drives within their language to get any syncronized motion between axis. Creating an interface in translation coding won't be a walk in the park either. As I understand it, mach does not put out code. Only step and direction signals to a parallel port based on a G-code file in its grip.

If you could tap into the circuit where the step and direction signals enter the stepper amp from the indexer portion of this drive, you may have a chance with mach.

Sounds somewhat like a galil multi-axis motion control card and its odd language. The difference being integrated multiple-axis. You may only be able to address the Digiplan code to one PDX at a time without interpolation as a simple option.

Your remote location does ad some challenges, but if there is a will.....there is a :wee:

DC

[Autocraft]

I read about that ability in the PDF also. What was not clear is if you could interpolate individually addressable drives within their language to get any syncronized motion between axis. Creating an interface in translation coding won't be a walk in the park either. As I understand it, mach does not put out code. Only step and direction signals to a parallel port based on a G-code file in its grip.

If you could tap into the circuit where the step and direction signals enter the stepper amp from the indexer portion of this drive, you may have a chance with mach.

Sounds somewhat like a galil multi-axis motion control card and its odd language. The difference being integrated multiple-axis. You may only be able to address the Digiplan code to one PDX at a time without interpolation as a simple option.



DC

Hey, I've just skimmed thru some Galil codes, and you are right, it's pretty similar to x-code albeit a little more complex in its functionality. Mach3 has a plugin to support galil controllers right? I guess that makes my attempt to interface the pdx with mach3 that more plausible. Like you said it won't be a walk in the park but it's not that many lines of code and I can use the grex and galil plugins as base.


I'm sorry for all my questions but what do you mean by interpolation? From my limited knowledge and understanding, when you daisy chain the pdxs together you have to set unique address for each by changing its hardware settings. Then when you send a command it echoes thru all the pdxs with the units only interpreting the commands that are specifically addressed to them (i.e. pdx 1, 2, 3 all receives the command 1A10 but only pdx1 will change its acceleration to 10steps/sec) then the command echoes back to the pc. Is this what you mean by interpolation?

Your remote location does ad some challenges, but if there is a will.....there is a :wee:

Hah, tell me about it. Old lathes and mills that would have been sold for scrap in US or Australia are still being sold for substantial amount of money because no one's moving to the new fangled cnc stuff. On the other hand hi-tech industrial machines from bankrupt industries are selling at almost scrap prices specially the electronics because of the afore mentioned unfamiliarity with any machinery past the 1950's.

[One of Many]

Interpolation is handled by the control to take code that intends to do an arc and mix multiple axis motion in terms of distance and velocity to achieve the proper ratio to follow that path.

As I read the PDX data sheet, I do not see that as an option. Sure, you can create accel/decel profiles, but once the go command is given to individual drives, they do not synchronize that profile to follow the profile under control of a common cpu that looks ahead to mix those signals or code together in order to maintain that path.

Even with the Galil to Mach3, there is a common CPU in the Galil controller and I suspect an interpreter type of post processor within Mach3 to convert G-code into something the Galil motion control comprehends. With this scenario, the brains control the limbs. With the PDX, individual brains are in each limb, but do not communicate amongst themselves in order to cooperate in a controlled manner.

Possibly making contact with a Parker applications engineer might be the best bet to find out if there are programming features to support what you are attempting to use these drives for?

I am afraid my speculation and conjecture isn't being much help. My intentions are not to discourage your efforts. I am just pointing out glitches and possible road blocks you might find along the way.

DC

[rasta]

hi;
have you been able to modify those dries to take step & dir?
regards
Mariano

Interpolation is handled by the control to take code that intends to do an arc and mix multiple axis motion in terms of distance and velocity to achieve the proper ratio to follow that path.

As I read the PDX data sheet, I do not see that as an option. Sure, you can create accel/decel profiles, but once the go command is given to individual drives, they do not synchronize that profile to follow the profile under control of a common cpu that looks ahead to mix those signals or code together in order to maintain that path.

Even with the Galil to Mach3, there is a common CPU in the Galil controller and I suspect an interpreter type of post processor within Mach3 to convert G-code into something the Galil motion control comprehends. With this scenario, the brains control the limbs. With the PDX, individual brains are in each limb, but do not communicate amongst themselves in order to cooperate in a controlled manner.

Possibly making contact with a Parker applications engineer might be the best bet to find out if there are programming features to support what you are attempting to use these drives for?

I am afraid my speculation and conjecture isn't being much help. My intentions are not to discourage your efforts. I am just pointing out glitches and possible road blocks you might find along the way.



DC