Advanced Search

Attach the hal and ini files here so we can have a look.

@geebee
Make a new config, start over.

Welche servodrives hast du genau ?

Gruß bassti

You have:

FF1 = 0

in your ini file

This should always be 1

LinuxCNC is very good at pointing out failures, so if you are not getting latency warnings while using it, it is not a latency issue.
-
When i replied about the low current, i was going to add that to much current can also do that, but it has to be way over the rated motor current to cause such issues.
-
In general,
-if the motor gets warm in 3 minutes, or hot in 5 minutes of being on = way to much current
-if it gets hot in 15-30 minutes = all should be good
-warm is if you can keep your palm on the motor
-hot is when you can keep your finger on it, but not the palm
-to hot is if you skin remains on the motor = the motor is cooked
-
keep in mind that current ratings on cheap drives are "peak", while on good expensive drives it is "RMS", but lately cheap drives have both ratings written on them, so use the "peak" values while testing, then decide if you need to add more current if the motor skips.
Always use lower current if not sure.

The stepper motor withstands temperatures of 120 degrees Celsius.
This is, of course, the maximum parameter and during normal operation it should be cooler.
On the other hand, if the motor is cold, it means that the current is too low, resulting in less shaft torque.
Without a thermometer, this can be assessed by stating that the engine should be hot, but should be able to be touched by hand.

Sörry for eventually annoying anyone. Some reading I certainly did, it is just a lot of stuff. I have used several motors both of the same and different types and two different drivers. I can limit current at the tb6600 and certainly did. The cs-d508 does not provide any switch to do so. Its datasheet mentions overvoltage and overcurrent protection. Apparently the current nevertheless is too high, not too low. The motors heat up quickly when not moving. I lost a second one even though I have become more careful after the first one burned. I certainly do know how to wire a 2 phase stepper. Mistakes might happen, but I will not repeat them every time I reconnect the stuff. I have been working with steppers and arduino but not yet on a professional basis.

So at the moment, I suggest its more likely a latency issue. It is gonna take me a while to go through the stuff pcw has told me. After that, I will let you know what I did and what were the results.

Cheers

The stepper motor pins can be checked without any tools.
Disconnect the motor from the controller.
Jump any two pins together.
Turn the motor shaft by hand.
If you feel a clear resistance, the pins belong to the same coil.
If the shaft rotates freely, the pins belong to other coils.
Shaft vibration instead of rotation is a typical symptom of confused wiring.
However, if the engine stands still and squeaks, it is a matter of settings, either the controller or the program.

geebee wrote:

The motor vibrates when it is supposed to move but I cannot see or feel, respectively, any movement.

Causes for that might be wrong motor wiring to drive, or to low current set on the drive.
Check the motor wiring to drive first, there are strict rules on wiring winding A to output A, and winding B to output B.
Then read the manuals and set the drives to correct current for the motors you are using.
-
Was this with the step timing set to something extremely low like 2ns or similar? Why would you change stuff at random?
Friendly advice, stop messing with things and expecting them to work, do a bit of reading before actually wiring and running stuff.

geebee wrote:

 Please note that according to the advice from tommylight I set the ip adress of the mesacard to 10.10.10.11 instead of 10.10.10.10.
 

It clearly states to set the PC network attached to Mesa to 10.10.10.11 or anything else except 10.10.10.10 as this is the actual Mesa card IP.
You have set it correctly as your ping results to 10.10.10.10 prove, so all is good.

The ping times show that the host is set to a 10.10.10.11 IP address and the 7I95T is set to 10.10.10.10

With and Intel MAC you should disable IRQ coalescing:

man hm2_eth

For testing, you can temporarily disable IRQ coalescing with this command:

sudo ethtool -C eno1 rx-usecs 0

( replace en01 with your Ethernet device name if different )

For testing general host latency,  when a base thread is not used (like with Mesa cards), 
you should run the latency test with the base thread disabled:

latency-test 1ms

or

latency-histogram --nobase -sbinsize 500

Also you should go through all the standard latency reducing methods here including the isolcpus grub setting:

linuxcnc.org/docs/html/install/latency-test.html


Just to get things going you should make sure the step time is 5000
and worst case you may need to increase the servo tread period to say 2 ms
(2000000 ns)




 

 Please note that according to the advice from tommylight I set the ip adress of the mesacard to 10.10.10.11 instead of 10.10.10.10.In opposition to the ones above the ping results are now with mesacard on the Intel adapter, not the Realtek.

--- 10.10.10.11 ping statistics ---
60000 packets transmitted, 60000 received, 0% packet loss, time 59999ms
rtt min/avg/max/mdev = 0.004/0.005/0.103/0.001 ms
geebee@debian:~$ sudo chrt 99 ping -i .001 -q -c 60000 10.10.10.10
PING 10.10.10.10 (10.10.10.10) 56(84) bytes of data.

--- 10.10.10.10 ping statistics ---
60000 packets transmitted, 60000 received, 0% packet loss, time 59999ms
rtt min/avg/max/mdev = 0.139/0.253/0.368/0.010 ms
geebee@debian:~$ 


 

You are pinging the host PC, not the 7I95T
try:

sudo chrt 99 ping -i .001 -q -c 60000 10.10.10.10

( It will run for 1 minute then print timing statistics )

 

I changed the network adapters: Mesacard is now Intel 82579V, internet Realtek RTL8111/8168/8411
ping times now are 0.019 - 0.022 ms compared to approx. 0.03 ms before (see attachment).

No change in behaviour. The motor vibrates when it is supposed to move but I cannot see or feel, respectively, any movement.

sind die Glaasmesstäbe noch "original" an der Maschine? Dann sind dass ggf analoge, es gibt konverter (die habe ich auch) hat bei mir aber nie ganz sauber funktioniert. Habe die dann gegen neue getauscht. Weiß aber nicht was das nen Haus weg hatte die Konverter oder die Stäbe.
Ich würde die Endschalter der achsen nicht zusammenfassen. Macht die Fehlersuche leichter.
Ich persönlich würde auch mehr Ausgänge machen, wie reglerfreigabe und ggf Schütze für die Stromversorhung der Hochspannung an den Treibern.

Bzw das ist ja eine die schon ne steuerung hatte, da würde ich mir die Liste der IOS aus den Manuals / Stromlaufplänen ziehen.