I have installed Mach4 build 1900 and Darwin 1.08 on XP. Driver install reported successful and DarwinTest.exe runs as expected.
I have copied mcDarwin.m4pw to the plugins folder and DarwinDLL.dll to Mach4 program folder. I have remoced sim.m4pw.

When starting Mach4 the error message in the image attached is displayed. I have tried restarting PC, re-installing Mach4 and removing the other plugins exept Lua but it dosen't help.

Does anyone have any ideas.

Alastair

hmm.. Have you tried running the 32 bit runtime for vs9.. ? I dont have any reports of not loading in
that way as yet. Download the VS9 runtime from microsoft..   vsredist_x86.exe , might help with that error,
though you should have gotten somthign about side-by-side config error if that was the trouble..

Art


 

Thanks Art, that got it loading!

Another question.
Is there a way to increase the width of the step and direction pulses as you can in Mach3 on the motor tuning page. I only had a short time to play once I got it loading but I couldn't see this in either Mach4 or Darwin. In Mach3 I have to set them to 3us to allow for the opto couplers on my breakout board.

Alastair

 

alastair:

  Sorry, at this point no. Darwin only puts out step pulses at 1 interrupt length in size. If I find this is too
limiting Ill look at stretching it, but initial polls showed most were fine at the default 1us that the printer ports usuallt
take by default.. Does it seem to lose steps without any stretching?

Art


 

In Mach3 without stretching it looses steps constantly and the motors sound terrible!

In Mach4 the motors will not move except for a small kick when you command a move in the opposite direction to the previous move.

I didn't have much time to experiment but the pulse length was the first thing that came to mind so I went looking for the setting. I won't get back to it until mid week but when I do get some spare time I will check the signals with a scope.

Alastair

alastair:

   That kick is a suggestion the step and direction pins are revered. Direction is held until a reversal, so if when you reverse you get a small kick.. thats the step seeing the direction change. Check your step and dir arent reversed..

Art

I read opinions from smarter people than myself. like Tweakie.  And I quote

"Art’s ‘Darwin’, parallel port driver, performed beautifully during the tests with no faults or bugs to report. It is very easy to set-up and the motion produced is just so smooth it is difficult to describe the difference from Mach3."

My experience is VERY different which makes me think I am doing some rather stupid things.      I am using Darwin 1.08 with Mach4 1888

Firstly I abandoned any idea of using Metric.   The results just confused the hell out of me.

For the life of me I could not get B axis (motor 4) selected as slave to Y axis.  Had to settle for A axis (motor 3)  Just could not figure how to get it to allow me to select B as slave.

Maybe that is a built in limitation of hobby version.  That would be a shame as I have one slaved axis and also a 4th axis (rotary).  That would blow Mach4 out of the water for me as a future option.

Dual axis homing I have not worked out yet.  Master and slave work well together up until I reference that axis.  Then they try to move in different directions.

However all that is small potatoes compared with the smoothness of operation (or lack of it) compared with Mach3.

25000 kernal speed.  Good driver test graphic.  Time in interrupt approx 7 micro seconds (is this my problem?)

No hardware changes. Start up Mach3 and all is sweet again, as it has been for years.

With Mach4 it sounds awful but does move and I don't think positional integrity is affected. (I need to check that)

Any ideas what silly mistake I have made?

Greolt

Hi Greg,

Thanks for the compliment - I am far from being smart - I am perhaps just lucky    I also had a lot of help from Art.  I have it working on WinXP and Vista but so far have had no luck at all with Win7.

If it's any consolation I could not get B axis working either (as you say it is probably a limitation).
I have not yet tried v.1888 but I am fairly certain the Metric / Imperial issue has been sorted.

I start loosing occasional steps, only on blended moves, when using a Pulse Count above 27000 but up to that the motion is beautiful so I am guessing it may be something relating to your PC but I have no suggestions as to what it may be.  Does 'Disable VBO Tool Path' make any difference ?

Tweakie.

Tweakie

I will try the 'Disable VBO Tool Path' thing tomorrow.

I am in a different time zone and off to bed now. 

Greolt

PS: this is XP

Hi:

>>"Art’s ‘Darwin’, parallel port driver, performed beautifully during the tests with no faults or bugs to report. It is very easy to set-up and the motion produced is just so smooth it is difficult to describe the difference from Mach3."
My experience is VERY different which makes me think I am doing some rather stupid things.   I am using Darwin 1.08 with Mach4 1888

     Goes to show experiences do vary. Could be machine, could be something else..time and testing will tell. Doesnt make you any dumber than I.. Im sure testing will show why things are screwy.. I havent had any limitations even up to 70,000 pulse frequency..BUT Im using servo's which can smooth timing discrepancies out of a stream. Its very possible that some problem or
another is creating trouble. Tweaky is the second to have grinding noise above a certain value.. Its something Im hoping time
and use will point to. It may be pulse width as one person pointed out. I only pulse at the default of 1us or so on the step, allowing the minimum time of pp port register changes to control the time. Im m3 I added up to 3-4 us to make the pulse
longer.. I may have to try that to see if it helps. (Tweaky: did you have to add time before to the pulse widths?)

>>Firstly I abandoned any idea of using Metric.   The results just confused the hell out of me.

   I used only metric, but I found any switch between inch and mm caused huge trouble for me, first you have to restart the program and something about the switch junked many things.. until latest version 1900 which seemed to fix that up for me.

>>For the life of me I could not get B axis (motor 4) selected as slave to Y axis.  Had to settle for A axis (motor 3)  Just could not figure how to get it to allow me to select B as slave.

   In M4 no motor actually owns an axis. Your "B" axis is simply motor 3. So youd map motor 3 to B axis in M4 and slave it. But Ive done no slaving tests so I cant confirm slaving works.

>>Maybe that is a built in limitation of hobby version.  That would be a shame as I have one slaved axis and also a 4th axis (rotary).  That would blow Mach4 out of the water for me as a future option.

   Darwin only handles 4 axis at the moment, I havent turned on the others, Ill list to do so, no reason for them to be off anymore..

>>Dual axis homing I have not worked out yet.  Master and slave work well together up until I reference that axis.  Then they try to move in different directions. 

  Probably has to do with the mapping, the homing can handle slaved axis, interesting that they move in different directions... ID have thought theyd home the same way they move, in reverse to each other. (Listed for checking)..

>>However all that is small potatoes compared with the smoothness of operation (or lack of it) compared with Mach3.
>>25000 kernal speed.  Good driver test graphic.  Time in interrupt approx 7 micro seconds (is this my problem?)

   Thats low compared to M3, and likely due to the decreased pulse width. Did your M3 have pulse width setting higher than default? Ill add a setting for that to see how stretching the pulse might help in smoothness for you. Mine is very smooth as was tweakies, but even Tweakie has roughness above a set number...that makes me thing pulse width..

>>Any ideas what silly mistake I have made?

  None. , you tested and reported, Im gratefull. I will add in a pulse stretch by making all steps take a 1us wait to make them longer, we'll see if that helps, and Ill check the homing to see why a slave might home in the reverse direction.. ( Thats ahandy report..). Ill also turn on the other two axis so 6 motors can work..

Thx
Art


Greolt