5i25 - 7i76 support testers wanted

Thanks Chris, that certainly explains things. I was wrongly under the assumption that the folder was now included in the buildbot version.
My pncvonf is showing as beta 1 after I updated to the buildbot version.

The main thing I was hoping for was the connector and pin assignments to agree with that of the actual board. It was also a bit unclear of the status of the spindle outputs. I went back and researched the posts though and should be able to sort out the pinout labeling.

Thanks much
George

George
Sorry for the confusion.
I just pushed a fix to 2.5 buildbot that I think really improves the connector/pinout relationship.
The tabs now will tell you the connector number and the pin number represent the signal names in Mesa's docs and HAL numbering in linuxcnc.
The pins are also laid out on the page as you should find them on the connector.

If you get a chance let me know if it helps...

That is fantastic news! I will get the update from the buildbot and report back in the next day. As ususal, I am a little confused, the pinout info is in the 5i25 folder I believe? Must be you made connections to get that respository updated.

Thanks so much.
George

no the 5i25 folder stays the same.
The Mesa configuration page that you select the signals have numbers and the tabs should tell you the connector number.
The encoders doesn't tell you the connector number but you can look that up in the Mesa manual.
It's not perfect - just better - I hope

cmorley wrote:

George
Sorry for the confusion.
I just pushed a fix to 2.5 buildbot that I think really improves the connector/pinout relationship.
The tabs now will tell you the connector number and the pin number represent the signal names in Mesa's docs and HAL numbering in linuxcnc.
The pins are also laid out on the page as you should find them on the connector.

If you get a chance let me know if it helps...

Either I am doing something very wrong or your changes are not live yet and available for download on the buildbot. Here is what I did:

using this file from the buildbot (deb buildbot.linuxcnc.org/ lucid master-rt) I updated and installed using these commands from the buildbot page (
"sudo apt-get update" and "sudo apt-get install emc2" ) The terminal window showed the progress as it installed, so I have confidence that a update occurred.

Rebooted for good luck and ran pncconf, selected 5i25 with Xy-lathe, and 7i76x2, changed step-gens, encoders etc to reflect a single 7i76 board and cleared all the sanity check boxes

I got tabs for I/O connector 2 and 3 along with smart serial 0 after accepting component changes

connector 2 shows pins 24 -40 under its tab
connector 3 shows pins 0-4 on the left and 0-1 on the right (two entries for pins 0 and 1 ???? for the same connector????)
the smart serial 0 tab shows 7i76 tabs under it labeled tab0, tab1 and tab2 ( tab 0 and 1 have pins 0-23 listed, tab2 shows a single selection defaulting to unused analog output)

smart serial 0 shows tabs under it for Tab0, Tab1 and Tab2 for the 7i76 board

SO......... I do not see the changes you made to correct the tab/connector/pins to agree with the physical mesa board.

Am I doing something wrong?????? OR is it just that your changes have not gone live on the buildbot just yet????
(sorry for being a PITA)

Thanks,
George

well looks like the changes had not made to buildbot yet. I don't actually know how to check if it did...I should find that out.

as for two 0 pins on the same connector - remember the pin numbers are the COMPONENT number not the physical pin location on the connector.
look in the 7i76's online manual to convert this to the actual connector pin location.
I know thats a bit of a pain but you'll wire it once (hopefully !) after that if you decide to edit the HAL file you need to know the component numbers.

The changes I made where so the sserial's tabs (eg TABo) will say 7i76-I/O TB6.
then i also changes the pins layout to more reflect what you find on the card eg for TAB0 (TB6)
there are 16 inputs starting at 0 and 8 outputs starting at 0

Not a PITA at all. I wish it wasn't a PITA for you !!!
I do appreciate your feedback so I can fix this for others.

Ok, thanks, guess all we can do is wait and check the buildbot occasionally for new updates.

In the meantime, I started to play. I hooked up the mesa boards, and connected a walwart supply to the 7i76, nothing else connected to it.
I setup some limits, a x and Z axis and a few misc pins in pnccong and applied. I keep getting the same error message when I launch the config. I tried eliminating some defined pins in pncconf but the same error message pops up on launching.

Maybe you can shed light on the cause of the error, it reads "hal:131: parameter or pin: hm2_5I25.0.7I76.0.0.spinout-mixlim not found (2880 on next line)

I was just going to play with the boards on the bench and test with a meter to see if I could get outputs to work and sim a input with a resistor and a cliplead. Nothing is easy LOL

Any suggestions on the error message? Also, if you hear any news on your changes going live on the buildbot please make a post to advise.

Thanks,
George

cmorley wrote:

Has anyone built a config with the 7i76's spindle control?

I need to understand how SpinOut-maxlim SpinOut-minlim and SpinOut-scalemax are used properly.

WHOA !!! I am getting a error message showing Spinout-mixlim Is pncconf generating a typo perhaps???? I am wondering if it is being generated in the hal as mixlim and should be minlim.

I will experiment with it and change the spelling in hal to minlim and see what happens while I wait for a reply.

George


EDIT That was it !! I looked at the generated hal and searched for spinout-mixlim and replaced it with spinout-minlim and the error message is now gone. So it appears pncconf is generating a typo


PS that error message is gone, on the the next one LOL (hal 229 halui feed-override.counts was already linked to signal 'fo-count' )

EDIT Think I found that error message too. Did a find in the hal file for feed-override.counts and it looks like some things are being assigned to the wrong pins by pncconf (used twice), here is the code from the hal that I believe is the offender. It looks like a lot of things are mislabeled in the spindle override section using feed override variables. This one is going to be tougher for a newbie like me to fix.

# connect feed overide increments - MPG

setp halui.feed-override.count-enable true
setp halui.feed-override.direct-value false
setp halui.feed-override.scale .01
net fo-count => halui.feed-override.counts
# connect spindle overide increments - MPG

setp halui.feed-override.count-enable true
setp halui.feed-override.direct-value false
setp halui.feed-override.scale .01
net so-count => halui.feed-override.counts

DANG!! I fixed it !! pncconf was assigning spindle functions to halui feed override pins, I corrected all the feed-override areas in the spindle section to spindle-override and now EMC launches without error! Of course I am not nearly done, this is just bench testing to get myself familiar with linux, EMC and mesa boards, its all new to me within the last couple weeks.

Looks like I found a couple more areas for you to fix Chris. Am I a PITA yet? LOL.

hmm for some reason you are getting an older version of pncconf.
the 'mix-lin' was found and fixed already. Are you should you are using the buildbot version? something weird there.

as for the spindle override - you bet you found another error! thanks for the report.
instead of feed-override it should be spindle-override
I assume you selected both.

see you'll be a HAl expert in no time !!

by the way min-limit is the minimum output you want
max-lim is the maximum
max-scale is the number you want to represent maximum output.

eg if you want +-10 volt output :
min_lin =-10
max_lim -10
maxscale =10

for 0 to 10 volt
min_lim = 0
max_lim =10
max-scale =10

for 0 to 1000 rpm
min_lim = 0
max_lim = I'm not sure if this is scaled or not so could be 10 or 1000 (I bet 1000)
max_scale =1000

I might mention that to use MPG override options you need encoder input.
the 7i76 only has one encoder input it's usually for the spindle.
you could have override capability using a rotary switch though.