Advanced Search

So I should probably start with my Dell first and then get a Mesa card once I feel up to it? Just to confirm, you are saying that you don't think it'll be an issue for making that part (assuming the jitter is not a problem)?

leexi wrote:

Please share your comments and experience.
 

1. This card is not clone. This is original chinese design. 
2. Never upload any firmware. The proper firmware is loaded by manufacturer.
3 You have to do "mesaflash --device 7i92 --addr 10.10.10.10  --readhmid". This shows card configuration.
4. There is in your hal file line like "loadrt hm2_eth board_ip="10.10.10.10" config="num_encoders=1 num_pwmgens=3 num_stepgens=5"". Functions ommited in this line are unavailable, but you can use it's pins as regular I/O. Of course you are limited by card design. If manufacturer designed some pin as output you cannot use this as input (and vice versa) because output circuit is one way.
5. As mentioned above dkms driver may help. Also other version of kernel may help. The "error finishing read" means the break of connections caused by driver/kernel bug rather than hardware problems, however check cable.  I experienced this problem even after a few hours of proper work. The good idea is leave opened LinuxCNC for as many time as you can (no matter you do something with, or not ).
6. I tried use PnCConf two times (the first and the last) and not recommend this to anybody. Better option is to find config similar to your needs and change something in text editor.

IIRC the original parallel port could switch up to 1,000,000 times a second. Assuming two switches per pulse, 8000 pulses per rotation, that's 3750 RPM, which is way more resolution than you need (steppers are 200 steps/rot, or 3200 with microstepping), and way faster than big steppers can probably run anyway.
When I was using steppers and a breakout, it was the steppers that were the limitataion.

  My only concern regarding using my parallel port computer with a breakout board is the delay (hence why I was considering the mesa card). Hypothetically, let's say my jitter isn't a problem. Will my experience making chips be the same as with a mesa card? I want to make some pretty complex parts and am concerned that any amount of jitter could make that impossible. I've attached an STL file of a part that I want to make for a friend. Could this be possible using a parallel port, or should I just spring for the more expensive card?

i will try it.. what i found in my ini... ich have - sign on some scales and outputs... i wired all up + to + and - to -.... but if you look into the pum560 sim file.. you will see that the TCP is upside down.. or is something wrong on my sim file??? thanks

Not sure but it seems a bit odd to me that you have a negative 'D-5' value and asymmetric angles for ALPHA-{3,4,5}:

When the example 'puma560_dh.hal' has:

and the melfa example has:

linuxcnc.org/docs/stable/html/motion/dh-parameters.html

After posting that image of the g code I believe I see it now. The % symbol is on line 5. So that would explain everything being off set considering the line is starting at 1...
I knew it had to have been something silly.

Update: That was it.... 

This is it.

If this isn't the version,  I don't know where it would be.

hi aciera!

i attached the ini hal and a second hal for the dh paramters.

possible i see somethnig not.. but genserkins with the MDH made me crazy .. hope i can fix it with the help of the forum!


Thanks for the quick answer!

Not much to go on. Please post your hal and ini files.

Also, please post the gcode

TheTinkeringMechanic1 wrote:

3.4.9

That is most probably the kernel version, not the GUI.
What user interface on screen?
It seems like GmoccaPy.
Upload a picture of the whole screen.

3.4.9

Hi guys!

I have gere a puma 200 robot arm from unimation.

All joints working
All brakes working
All encoders working
All ratios of the gears fit to encoder and measured angles anglea digital angle measurement tool.

After all this i change my kinematic to GENSERKINS

In joint mode all works.
In world mode i have issues.

There is one thing what make me headache... the TCP is upside down.

If i look to the puma560 vismach genserkins simulation.. its the same.

I set up the DH parameters.. i look to the puma 560 dh parameters in the simulation config and adapt them to my puma 200 parameters i get from the drawings.


In world mode Z looks ok... x y small radius?
But if i start the linux example i get errors of limits and kinematic errors.


Has someone the knowledge to look to it?

Thanks
Could someone help me please?