First of all many thanks to all involved with the upgrade.
what I miss (or to stupid to find) is a more global "mark as read"
 

Thank you!
I couldn't find a way to send PMs, so I posted my email. 
My gmail address is my spam/trash mail though, so no worries.

Yes, I am interested. It is here collecting dust.
I never used it except to reverse the protocol and to play with it.
It's the 2 axis 1080TC with control panel as displayed here.
 
How can I contact you? 
My mail is hansfbaier --+at+-- gmail {[dot]} com

Ive been talking to Aanny for years now! She helps as much as she can.
Also, id be interested in purchasing the controller for a spare if you're interested?

has anyone run into this I have kinda hit a wall and am looking for some insight

rtapi_shmem new failed due to shmget(key=0x48484c34): Permission denied
rtapi_init: could not open shared memory for uuid
HAL LIB: ERROR: rtapi init failed
hal lib: rtapi app main: Invalid argument (-22
Note: Using POSIX non-realtime
tpnod: dlopen: /usr/Llib/linuxcnc/modules/tpnod. so: undefined symbol: hal init
<commandline>:0: waitpid failed /usr/bin/rtapi app tpmod
<commandline>:0: /usr/bin/rtapi app exited without becoming ready
<commandline>:0: insmod for tpmod failed, returned -

Yes I had to buy their lathe controller to reverse the protocol as their tech support would not give me the protocol. It is thus still sitting idle here and I could sell it. Give the salesperson Aanny greeting from me when you order it

That is such a cool control panel build!
The GUI is gmocappy, I suppose?

Totally recognized that control panel!
Ive bought many things from SZGH tech, including a new lathe that should arrive in a month.
Thats so awesome you figured out the communication and connected it to linuxcnc!! Means ill have to order a couple panels to add to my existing linuxcnc machines

yeah I sadly left it at tmp then I needed to reset to isolate the cores so the screenshot is gone sadly. I only remembered after I started writing the post. sorry

You've put a lot of work into this project. Thanks for sharing it!It's a really cool project, and you've got me thinking about whether I should follow a similar approach for my upcoming project: building a control system for a Mori Seiki MV Junior VMC.I still have some spare circuit boards from my lathe project, and I was originally thinking of just adding a few more buttons and a larger display. But after seeing your project, I'm starting to reconsider my approach! 

servotune: free parameter and scope tool for Lichuan A5 (LC10P) and B2 drives, testers wanted

Hi all,

Lichuan's PC software for the A5/LC10P and B2 is a parameter editor and not much more. I wanted something closer to the tuning tools other drive makers ship, so I've started writing one. It's called servotune. It's free (GPL), and runs on Linux and Windows.

It talks to the drive through its RS-485 Modbus port (CN3/CN4 on the A5, CN2 on the B2) using an ordinary USB-RS485 adapter. It doesn't need LinuxCNC and doesn't touch the step/dir side. The drive can stay wired to your Mesa card while you watch it and LinuxCNC moves the axis.

What the first release does:

1. Scan and probe: finds the drive at any baud rate and format. A strictly read-only probe writes a report of which registers answer and how fast.
2. Parameters: every parameter by group, showing the drive's value, the default, the range, and when a change takes effect (immediately, at stop, or after a power cycle). You can edit one, or load an .lcd file from Lichuan's software and write only the differences after a review screen. It can also save the drive to an .lcd file.
3. Program a drive from a file: writes only what differs, reads everything back to verify it, and saves a report. The original values are journalled to disk first, so an interrupted session can be restored.
4. Scope: the drive's own position following error, speed against speed command, and torque, live. It has triggers, cursors, captures you can overlay to compare before and after a gain change, and CSV export. This is useful with step/dir, where LinuxCNC never sees the following error but the drive does.

One thing to watch if you script this yourself: .lcd files trim trailing zeros, so "P08-00=30" on a 0.1 Hz parameter means 30.0 Hz (raw register 300). servotune takes the scaling from the manual, not from the file.

The honest bits:

1. So far it has only been tested against its built-in simulated drive, not on a real drive. That's why it's tagged as a pre-release.
2. The scope polls over Modbus. Lichuan doesn't expose an internal capture buffer, so expect roughly 30-180 samples a second depending on baud rate. That's enough for following error through a move, overshoot and settling time. It isn't enough to see fast oscillation, and the scope warns you when that might be happening.
3. It doesn't drive test moves or offer guided tuning yet. That's next.
4. The Windows build hasn't been run on Windows yet.
5. The EtherCAT LC10E isn't supported.

How you can help:

If you have an A5/LC10P or a B2, please run the read-only probe and post the report, here or as a GitHub issue. It only sends Modbus read requests.

Each report answers something the manuals leave unclear: B2 register numbering, whether the A5 really runs at 115200 baud, and which monitor registers are 32-bit.

Downloads (Windows zip and Linux AppImage), source and docs available from my GitHub, linked under my profile.

Not affiliated with Lichuan, just a user for many

Hugh

andrax wrote:

A brief description of the script would be helpful.
What does it do?
How does the process work?

Without a detailed description, I wouldn’t test the script—my servos are too valuable to me to risk it.
 

use the source?
* reads parameters (SDOs) with 'ethercat upload' and displays them in an editable table
* writes SDOs with 'ethercat download'
* monitors following error with halcmd

I tried the windows-program from stepperonline (with a USB cable to the drive) and it does similar things, with something similar to halscope for plotting following-error etc, but this tuning-script is more convenient as it talks over ethercat to the drive while linuxcnc is running. No automation for actually finding the gain parameters, that still has to be done manually e.g. following the post by StepperOnline above.

I really like this new style! But i would also do all of the axis the same style, so no A and B in smaller font and in different columns, instead all digits lining up. Maybe make the G54 and DTG the same font size as X,Y,Z.

Is there anyway to test this already or was this just a mockup?