How to add a new Mesa card to pncconf?

it is the same as of 7i92 it coudt be loaded with multiple bit files that triggers the multiple options fail
therfor woudent it be good to get the DMESG readout for eth into pncconf and let the user decide on that values
ok i see also that the bitfile and dmesg going different as the pin file
as the bitfile refers the gpio to be special and the dmesg gets only the gpio LIKE the stepgens hardwar

A main tread opener you can use the same pncconf 5i25 file as the 7i92/7i93 is a 5i25 with eth in insted of pci/pciE

and OVERALL
Pncconf gives you only a basic Conf
you need to get into HAL/INI to get your system real state
so WHY not starting from bottom on a given 7i93 BASE conf as it is in Conf tread

WHAT are your datercards you pluged into the 7i93

So I find it hard to understand what you are getting at.
The discovery option is the idea to address custom firmware - as long as the firmware basic blocks are something pncconf already knows about.
There is actually a way to add firmware manually (it's tedious)

Pncconf can actually configure a very complicated config - it's just that most don't utilize it as so.

I could bore you with how it's hard but very useful when it does work - There is a great amount of hidden knowledge coded in to pncconf from over the years.

Chris M

I've got the config basically working as is, and I'm fairly familiar with hal. I agree with what Chris said prior though that pncconf can generate sophisticated configurations if used properly. For instance the machine I am working on is a XXYZ servo gantry with tandem X motors. Generating the config by hand could certainly be done but will be much easier with the capabilities built into pncconf (with another forum post to come on the topic of servo gantries via pncconf once I get to that point!). I've got two 7i40s and one 7i37TA, but all that is working properly.

Pncconf can also facilitate machine wiring configuration, allowing testing of I/O and open loop servo testing, which is why I am interesting in getting the 7i93 card specifically working, so that the right drivers load within pncconf to enable those features. Otherwise, the config generation works fine using other cards as a base and searching/replacing.

My goal for this machine is to generate the entire config via pncconf, with some minor changes after the fact done via automated search and replace such as `sed`. In a non-profit volunteer-based hackerspace like ours that seems like the most feasible option for long-term maintenance.

Thanks,

Scott

agree it isbest for mantainance
so as you almost there cm saiing it will benefit others going the same way
your mashine is also a comen one and becomes more popular as prices falling monthly

time will bring things together maybe somtime even my preferd 7i92 will be a pncconf choice

shasse wrote: I've got the config basically working as is, and I'm fairly familiar with hal. I agree with what Chris said prior though that pncconf can generate sophisticated configurations if used properly. For instance the machine I am working on is a XXYZ servo gantry with tandem X motors. Generating the config by hand could certainly be done but will be much easier with the capabilities built into pncconf (with another forum post to come on the topic of servo gantries via pncconf once I get to that point!). I've got two 7i40s and one 7i37TA, but all that is working properly.

Pncconf can also facilitate machine wiring configuration, allowing testing of I/O and open loop servo testing, which is why I am interesting in getting the 7i93 card specifically working, so that the right drivers load within pncconf to enable those features. Otherwise, the config generation works fine using other cards as a base and searching/replacing.

My goal for this machine is to generate the entire config via pncconf, with some minor changes after the fact done via automated search and replace such as `sed`. In a non-profit volunteer-based hackerspace like ours that seems like the most feasible option for long-term maintenance.

Thanks,

Scott

Scott, if you were to push this back to Linuxcnc, in my view, one of the useful bits of housekeeping would be to generate signal names in hal based on joint numbers not x,y,y1,z etc. Refer this example I shared a long time ago. forum.linuxcnc.org/47-hal-examples/34072...ranch-gantry-example

From version 2.8, the joints are allocated to axes by this ini file statement

Back when I was getting going, I changed the orientation of my machine a couple of times and it was always very confusing util I renumbered to the standard shown. If you did this, it would be much easier to change things around and it would allow the coordinates string to be entered by the user,not forced on him by a dropdown. I did raise this with Chris Morley at the time I wrote that config (which was when I tested revisions to pncconf for him).

In my view the joint axis (independent joints would be a better name) was a significant change to linuxcnc and at the time I did that config, neither stepconf or pncconf supported the proposed change and I doubt they would today.

I have only ever used V 2.8 and 2.9 on my machine and when I started there were a number of bugs I identified in the Axis and Gmocappy GUIs which have long been fixed becasue the GUI, had not fully supported the changes. In my view this is one of them too.

Otherwise, the config generation works fine using other cards as a base and searching/replacing.

I added an option to pncconf on the start page 'use INI substitution' then the HAL file will use [HMOT]CARD0 all over the place instead of the actual card name. The INI file of course will have this heading allowing you to easily change the name for different cards.

I don't like it but it may be handy for some people.

Chris