Advanced Search

"Watchdog has bit" usually means really bad latency (>5 ms)

Running

halcmd show param *.tmax

in a terminal when LinuxCNC is running
will give more information

What is the host hardware and clock speed?

Also do you have all power management/turbo modes/speedstep
etc disable in the BIOS?  Really bad latency is quite often related
to CPU power management.

It looks like it may be ethernet latency  1  &  2  

I am unable to completely finish cutting out my parts. With every part I tried to cut this error stopped me right before the end of the process leaving me with almost cut out parts. Any insights into what is causing and how to prevent it this is greatly appreciated.
 

so im back at it. i found my conflict. it was the second realtek card that was giving me problems. im running an amd ryzen 5 with an rx 580 gpu. i know it overkill for what we are doing here. i am getting a joint following error and hm2/hm2_7i96s.0: error finishing read! iter=24429. forum.linuxcnc.org/27-driver-boards/4223...error-finishing-read. i have been playing with the settings according to this post but i am still getting the error. i set the servo thread to 200000 and the p to 500. also i changed the dpll to -100. any ideas what could be causing this? what settings can i play with that will get me in the right direction to solving this?

Your right it is a 7i76E.


The resistor network is OK but the markings on IC's are very faded so what should I be ordering?

(I am in the UK and going to have to pay a minimum order handling charge to Farnell so I may as well order them all at the same time).

And thank you for your help!

You must mean 7I76E if you have CR17

CR17 is a SMBJ5V0 (SMB package)
Its job is to prevent overvoltages on the the
7I76Es 5V rail.

If PTC4 was fried that sounds like you accidentally
connected 24V to the 5V outputs on TB3. PTC4 is a
1206 2A PTC MF-NSMF200-2

Most likely hidden damage would be to 5V parts
U10 (serial), U12 (step/dir 0..3), U15 (encoder) ,
U19 (step/dir 0 ..3),  and U24 (step/dir 4)

Also if 24V got routed to step/dir outputs, the 27 Ohm resistor 
networks near the step/dir outputs may have been damaged
(check with ohmmeter, damage =  high resistance/open)

Just tried your suggestion of removing S20000 from the preamble. It appears to have worked. The reason I say "appears" is because it was not a straight forward test as I had already probed the board with the VFD off. Then I needed to warm up the VFD/spindle after which I ran the file from beginning to end. The probe section worked as expected as did the actual milling.

Thanks again for the suggestions.

I have been using Linuxcnc for a few years and I have learned enough to understand how little I really know about its inner workings. I started with a parallel port eventually moving to a mesa 7i76E. The machine started with 3/8-16 screws and skate bearings and now has ballscrews and linear rails. I enjoy modifying the machine as much as using it.

Mesa boards and combos that work for analog +-10V drives, all 25 pin ones
7i97T
7i92TM + 7i77 <<< notice the M, it can be directly plugged into 7i77
7i92TM + 7i83 + 7i85 <<< notice the M, it can be directly plugged into 7i85
7i96S + 7i83 + 7i85 <<< needs a DB25 to IDC26 cable very cheap at Mesa shop
7i76E + 7i83 + 7i85 <<< needs a DB25 to IDC26 cable very cheap at Mesa shop
There are also other boards, but the 7i96S combo can be easily found, the first two should be available as it was noted by PCW, might also be sold out.

Hi all,

And PCW in particular,

I would like to add a SPINX1A to my last (fifth) step/dir interface on my 7i76e.  I also have a 7i85 attached on the first expansion connector.  

I assume I'll need a replacement bit file and I haven't seen anyone else request this exact combination yet.  Is this possible?

Thanks in advance!

rodw wrote:

You don't need pncconf or stepconf to build a working config joint axes or not, you just need a text editor and do some reading. 
My Mesa 7i76e was not supported in pcconf so I had no choixe 7-8 years ago when 2.8 was relesed and I survived.
You could build a xyza config and edit the result
 

Yes, I'm working on it, I'm editing a configuration I made using stepconf and I'll test it on the machine as soon as possible. When you have tested it, I will return with the result whether it worked or not. Thank you for your help.

You don't need pncconf or stepconf to build a working config joint axes or not, you just need a text editor and do some reading. 
My Mesa 7i76e was not supported in pcconf so I had no choixe 7-8 years ago when 2.8 was relesed and I survived.
You could build a xyza config and edit the result

Danke Aciera für Deine Antwort! Also ich habe das hin und her probiert mit den Vorzeichen. Es scheint tatsächlich auch bei gleichen Vorzeichen genau so zu gehen. Entscheidend ist nur erstmal, dass die Richtung stimmt und man auf den Schalter fährt und nicht von ihm weg. Ich hatte mich nach dem Talla83 Video #12 -Homing gerichtet, der verwendet entgegengesetzte Vorzeichen.

Ja, ich habe Search_Vel auf 5 (=300mm/min bei meinem Setup) und Latch_Vel auf 2 (=120 mm/min) eingestellt.. Vielleicht ist das noch zu schnell, das werde ich an der Maschine noch endgültig einstellen. Momentan ist meine 7i76e noch auf dem Tisch zum Testen, bis ich ein Basis Setup damit habe.

I would check the 7I76E input you use for ARCOK  in halshow
to see if it switches as expected when the torch is fired.

Note that ARCOK signals from relay contacts may need a larger
load current than the 7I76E inputs sink (wetting current) so you
may need to add an external resistor (say 1K 2W) from the 7I76E
input to GND to provide this current.
 

I am unable to get a successful arc okay signal from my primeweld cut60 plasma cutter to qtplasmac. I wired the arc okay signal using 24v power from the 7i76e board and sinking it to one of the inputs just like a limit switch. When I go to cut out a part qtplasmac fires three times and says no arc okay signal was found. Is anyones wiring of arc okay different than mine, can anybody please share how you connected your arc okay signal? Thank you all for amazing help

Hallo zusammen!           
(Debian 12, LCNC2.9.0 Pre1 und eine 7i76e mit gemeinsamem Home/Ref Switch auf X)
Sorry, ich hatte Projektpause bis jetzt und komme nun auf mein Homing Problem zurück. Ja, ich habe sowohl die EN Hilfe zur INi und Homing durchgelesen, auch die DE Version. Und diverse Postings zum Homing, wo mein Problem nicht auftaucht.
linuxcnc.org/docs/2.6/html/config/ini_homing.html

Auf die Gefahr, mich tw. zu wiederholen:
Beide schlagen ja vier Varianten vor, wovon bei mir mangels Index/GMS nur die beiden ersten ohne Index in Frage kommen. die erste ganz simpel: Achse fährt drauf auf den Schalter und drüber, dann zurück bis der (Ab-)Schaltpunkt erreicht ist und dann auf Home Position. Kann man machen, ich will aber die zweite Variante mit zwei Anfahrten.

Was ich haben möchte, ist Zeichnung 2, wo auf den Schalter gefahren wird, dann zurück und wieder langsamer drauf, um von da aus auf die mit Home definierte Home Position zu fahren (bei mir +10)

IN der Zeichnung von (...ini-homing.html) werden Search Vel positiv und Latch_Vel ebenfalls positiv vorgeschlagen. Das kann aber nicht funktionieren, weil die Richtungen entgegengesetzt sein müssen. Ich habe alles probiert, erst Search_vel positiv (geht nicht weil Home mit 10mm vor dem Schalter aus der pos. Richtung angefahren werden muss, also Search_Vel sollte ein Minus Wert sein, dann gegenläufig Latch_vel ein Positiver Wert.

Was passiert bei mir: Ich habe jetzt noch mal alle Werte, die das Homing beeinflussen, ausprobiert mit Vorzeichenänderung usw. auch der Vorschlag von Dir, tommyligt mit den +/- 10 Home_Offset Werten. Nichts hat mich weiter gebracht. Bei egal welchen EInstellungen fährt die Achse auf den Schalter, wenn der erreicht ist bleibt sie stehen und zeigt an dass die Home Pos. erreicht ist. Keine zweite Anfahrt, immer sofortiger Stopp.
Hier mein Auszug aus der .INI:
[AXIS_X]
MAX_VELOCITY = 125.0
MAX_ACCELERATION = 1500.0
MIN_LIMIT = 0.0
MAX_LIMIT = 700 [JOINT_0]
TYPE = LINEAR
HOME = 10.0
FERROR = 1.0
MIN_FERROR = 1.0
MAX_VELOCITY = 125.0
MAX_ACCELERATION = 1500.0
# The values below should be 25% larger than MAX_VELOCITY and MAX_ACCELERATION# If using BACKLASH compensation STEPGEN_MAXACCEL should be 100% larger.
STEPGEN_MAXVEL = 156.00
STEPGEN_MAXACCEL = 1750.00
P = 1.0
I = 0.0
D = 0.0
FF0 = 0.0
FF1 = 1.0
FF2 = 0.0
BIAS = 0.0
DEADBAND = 0.0
MAX_OUTPUT = 0.0# these are in nanoseconds -
DIRSETUP = 10000
DIRHOLD = 10000
STEPLEN = 4000
STEPSPACE = 4000
STEP_SCALE = 640.0
MIN_LIMIT = 0.0
MAX_LIMIT = 720.0
HOME_OFFSET = 10.000000
HOME_SEARCH_VEL = -25
HOME_LATCH_VEL = 25
HOME_FINAL_VEL = 10
HOME_USE_INDEX = NO
HOME_IGNORE_LIMITS = no
HOME_SEQUENCE = 2
Ich wäre sehr froh wenn mir hier jemand helfen könnte!
Und danke schon mal für jeden Hinweis!
VG Arno