Advanced Search

pgf wrote:

I've just implemented this for myself.

I already have autoprobe routines that set g54 parameters for X, Y, and Z. I augmented my autoprobe functions to log the x and y values of every completed probe, along with the g54 offsets in effect at the time.

When I have a rotated part, I touch it twice, along a surface that's parallel to its twisted X axis. That gives me two new log entries. I have a script that level devil  reads those entries, and calculates d = atan ( (y2 - y1) / (x2 - x1) ), after correcting the x and y values with the offset information.

The script prints the result in degrees. In fact, the script simply prints the necessary g-code: "G10 L2 P0 R $angle", which I then pasted into the MDI window.
 
Once you generate the G-code command with the calculated angle, do you have any specific procedures for verifying the machine’s response after you paste it into the MDI window?

Once you generate the G-code command with the calculated angle, do you have any specific procedures for verifying the machine’s response after you paste it into the MDI window?

Fortunately this one is a reasonably easy fix:

it will spit out an error advising that you need to work inside a virtual environment, since you probably aren’t using your LinuxCNC machine for Python development, just run the override flag. Apologies I am typing on the go otherwise I’d look up the exact flag you need to add.

Where are you getting the info?
No DC side relays nor switches!!!
Never switch the DC side, unless you're in the solar panel business.

tommylight wrote:

Seems a bit excessive and not very useful.
It is much better to use a single small industrial switch and 4 of automatic fuses inside the control box, you can switch on/off automatic fuses as they can handle a bit of torture although they are not made for it, but since they will be used during testing only, they will be fine.
 

Hi,

Regarding the automatic fuses, I understand that we can also use DC circuit breakers.

I am studying to understand how to select appropriate DC circuit breakers for different PSUs and loads.

I understand that: Fuse Rating in Amp = Load Current x 1.25

I am using the Mean-Well MDR-100-24 PSU whose output is 24VDC 4A to power the 8 proximity switches exclusively.

The current rating for each of the switch is 300mA. In theory, 4 of the switches could activated at the same time, the maximum Load Current is 4 x 300mA = 1.2A.

But because the switches are just inputs to the Mesa 7I96S, I understand that, in this case, they draw very little current, maximum about 20mA each.

The maximum Load Current is then: 4 x 20mA = 80mA.

An DC-rated miniature circuit breaker is:

- Current: 1A - 2A

- Voltage: greater than or equal to 24VDC; up to 60VDC or 125VDC

- Curve Type: Type B

Is this correct, please?

Thank you and best regards,

...behai.

 

Got some goodies too. 

Got the table modified last week.

Pushed a change to Master to use 64 bit integers for hm2 quadrature
encoder count values rather than 32 bit, Seems to work as expected.
(32 bit ints cause a wrap at  a count of 2147483648)

(looped back a stepgen at 30 M counts/sec to the encoder
input to test, still took a while...)

Hmm ok that confirms it. I did most of my octopus functionality testing on the octopus v1.1, I didnt like the buzzer on the pro. The octopus PRO has level shifters on the UART lines. When I jumper the TMC UART line to another pin, it configures fine. That is the only thing I can think of. But this overlaps with another issue Im having with another board, that also has a level shifting circuit on the UART line.

Scott, if you can provide input/assistance on this issue, it would be appreciated.

Im trying to support this mellow flyCDY3 board, but the uart isnt working. I tried changing the baudrate, but no luck. Thankfully my buddy mdroberts is much smarter than me, so he collected some data for me on the issue.

It looks like i cant upload the excel file he made, but here is what he said

Further to the TMC2209 read-back. I used the scope UART decode feature to confirm that the programmed values sent to the TMC2209 seem reasonable and consistent with the registers, but the TMC2209 doesn't put any data out after a register read is requested. The sequence after resetting the board is provided in the attached.

I just don't see any output from the 2209 when a register read is done.

You should remove the e-mail, spam incoming

Looks like Francisco got it.

Rob

Who's the lucky winner?

I've just implemented this for myself.

I already have autoprobe routines that set g54 parameters for X, Y, and Z. I augmented my autoprobe functions to log the x and y values of every completed probe, along with the g54 offsets in effect at the time.

When I have a rotated part, I touch it twice, along a surface that's parallel to its twisted X axis. That gives me two new log entries. I have a script that reads those entries, and calculates d = atan ( (y2 - y1) / (x2 - x1) ), after correcting the x and y values with the offset information.

The script prints the result in degrees. In fact, the script simply prints the necessary g-code: "G10 L2 P0 R $angle", which I then pasted into the MDI window.

andypugh wrote:

And it's true, I think. For example we really could so with a section on toolchangers. 

And Ethercat now it can be installed without breaching licenses
And how to optimise later distributions

Hi Giovanni !

Step #1     I misread that as python3-lxml being already installed with the install of Linuxcnc.      Wrong of course and caused problems

Step#3     I didn't know where to expand the zip file.      Later I seen it didn't matter moved to the gladevcp folder

Step #5   "sudo ln "   I misread that as a upper case I at first.    

Step #8   I misread   {XID}  as (XID)  which also caused problems

Misc:    aux_gladevcp  should be changed to gladevcp as already mentioned

Much of the problems were due to me printing out the instructions and the pdf being in a quite small font which lead to me misreading some things.   A larger font would help here.

Thanks again for all your work !!

Hi all,

I've changed my setup to use an ethernet-based Mesa board rather than a PCI slot one. As a result, I have the old 5I25 Superport board which I have no use for. $10 for postage and its yours. Send me an email if you're interested.

Rob Forbes