Updating to latest version of 2.8

In other words:
1. Add the keychain:
sudo apt-key adv --keyserver hkp://keys.gnupg.net --recv-key E0EE663E

2, Create: /etc/apt/sources.list.d/linuxcnc-buildbot.list and add...
deb buildbot.linuxcnc.org/ stretch 2.8-rtpreempt
deb-src buildbot.linuxcnc.org/ stretch 2.8-rtpreempt

3. Followed by:
sudo apt-get update
sudo apt-get install linuxcnc-uspace

Correct? Yes, I'm still feeling my way through the darkness w/Linux.

Correct? YES IT WAS!! Color me a happy camper. I'm now up and running LCNC v2.8 on Stretch with PREEMPT_RT...and it's only taken me 18 months to get here. Now it's on to PNCconf and the Mesa 5i25.

Thanks to all for hanging with me on the journey.

Leonard

As per your advice, threw out the old .ini/.hal and ran PCNconf. XYYZ config. New files attached.
I'm able to manually jog X and Y, but only with a single Y connected as one of the Y axis needs to be reversed. No output/movement on Z, but it works in axis. So:

1. I need to reverse joint travel on a stepper motor. I know that I need to set a "-1", but I'm not sure of the parameter.
2. Need to find out why 'Z' is dead.

Any suggestions/help would be appreciated. At least I'm communicating with the router! Yea!

Leonard

mac44mag wrote:

As per your advice, threw out the old .ini/.hal and ran PCNconf. XYYZ config. New files attached.
I'm able to manually jog X and Y, but only with a single Y connected as one of the Y axis needs to be reversed. No output/movement on Z, but it works in axis. So:

1. I need to reverse joint travel on a stepper motor. I know that I need to set a "-1", but I'm not sure of the parameter.
2. Need to find out why 'Z' is dead.

Any suggestions/help would be appreciated. At least I'm communicating with the router! Yea!

Leonard

File Attachment:

File Name: My_Yeti.ini
File Size:4 KB

File Attachment:

File Name: My_Yeti.hal
File Size:11 KB

Change the sign of the scale for joint 2.

Well, you know how to fix travel direction thanks to Bevins.
I can't see anything obvious wrong with your Z axis so I would check your settings closely. Then check wiring.
When I got mine going, I had the scale wrong and initially I thought it was not working until I saw the faintest of movement that a quick tweak to the ini file and it was off like a rocket!

So what are the Z axis parameters:
Ball Screw pitch
Stepper steps per mm?
Drive reduction?
MIcrostepping on driver?

OK, I have at least a semi-working router! I've moved it out of ICU and into a private room listed as 'in serious condition'.

(Backstory...I'd run PNCconf only to find that the plug attachment order to my G540 didn't match the actual outputs. X, Y1, Y2 all worked, but in different plugs from the expected, and the Z output was on a cruise somewhere in the Caribbean. In addition, the Y joints racked.

After some excellent advice (THANK YOU, BEVINS!), I was able to change the direction on Y2 and eliminate the racking, but Z was still dead. Cut to the present....

The dead Z axis problem appears to have been in the stepgen setup. A friend noticed that PNCconf had mapped Joint 0 to "hm2_5i25.0.stepgen.01" in the stepgen setup for Joint 0 across the board. Similarly, the mapping for the other three joints was advanced by one number.

Having the confidence of the ignorant (and a backup of the original), we changed the stepgen setup to match the joint numbers and the Z-axis came alive! After switching the other cables to what we thought should be the original XY1Y2Z layout on the 540, the Linux demo file ran like a charm with the observation that on the longer curved paths around the 'C's' sounding labored. It just sounded as is the router was working harder there than in other places. The physical connections now match what I originally envisioned as the original plug scheme on the G540.

However....isn't there always?

After the program ran and a suitable period of us doing the 'Happy Dance' around the shop, we took temperature reading on the back of the 540. The average temps between J3 and J1 were in the high 40's (C) and topped at over 55 deg. in the J0 (X) region.

I was a bit concerned about that (bliss of the ignorant, remember?). Using another hunch, we changed the SETPLEN AND STEPSPACE values from 10K/5K to 3K/3K. I let things cool and re-ran the program. This time temps around J0 topped at 43 degrees. On a hunch, I let it sit powered on and idling for 15 min. Temps continued to climb at a steady rate until they read ~51 degrees at J0. I had expected a slight rise in temps for a short while, but 15 min. seems excessive to me.

Questions:
1. Any idea on why the stepgen setups were advanced by 1?
2. I know that Gecko lists 80+ as a soft max, but should the temps continue to increase while the machine is a quiescent state?
3. Is what's been done have anything to do with the 'laboring'? Perhaps I've invalidated the TP setup?

In any case, progress appears to have been made. A hearty THANK YOU to Rod, Bevins, and all others who have helped me get to this point. Of course, any further help will be greatly appreciated.

Sorry this is so long, but I didn't want to miss any information that might prove enlightening. Current .hal and .ini attached. Previous editions are in an earlier post.

Leonard

THANK YOU, Mr. Pugh! I've added that to my database. Moral...don't forget to shut off the machine when done playing!

However, some type of external cooling for the G540 may be in order. I ran a program of about 4500 lines yesterday (a simple block with 50 holes) and monitored the temp. I terminated at 15 min. when the temp of the back of the G540 case reached 66.5 deg. C.

Will playing with (reducing?) the STEPLEN and STEP_SPACE have any effect on lowering the temps?

Thanks, again!

Leonard
PS: Sweet bike!

mac44mag wrote:

THANK YOU, Mr. Pugh! I've added that to my database. Moral...don't forget to shut off the machine when done playing!

However, some type of external cooling for the G540 may be in order. I ran a program of about 4500 lines yesterday (a simple block with 50 holes) and monitored the temp. I terminated at 15 min. when the temp of the back of the G540 case reached 66.5 deg. C.

Will playing with (reducing?) the STEPLEN and STEP_SPACE have any effect on lowering the temps?

Thanks, again!

Leonard
PS: Sweet bike!

I have a G540 on one of the first machines I did. It got hot along with the steppers but that's normal. What voltage are you running it at? 48 max I hope. I put my G540 on a heat shrink if I remember correctly.

Pncconf is used to get the initial config running. Then you shouldn't ever have to go back to it.
You'll get a wiz at configging these beasts....

Of course you do know that all this talk is fictional until pics show up...... It didn't actually happen unless it has been confirmed with pics and/or videos....lol

Finally, a chance to thank you personally for the reversing shout-out!

Yes, I'm running @ 48vdc, but as yet have no way to measure current draw. That's in the works. I notice that the article that MESA has up indicates a substantial decrease in case temp with just an external fan. That's probably my next step. And, I think I have a couple of old CPU heat sinks around somewhere that may get repurposed, although MESA didn't seem to find a lot of benefit there.

Thanks, again, for the help!

Leonard