Remora - ethernet NVEM cnc board

Yes I can see this log finished ## Entering IDLE state
thats all.
loadrt remora-eth finished successfully independlently on ip address or card is connect.
IP config pc is.
ip 10.10.10.11
mask C (255.255.255.0) 8
def gat 10.10.10.1
ping is ok.
M

I mean the output on rs232 when you turn the power on.

Me too,
power up card and I can see log on rs232. all is ok. log finished ...
Creating a std module
Registering interrupt for interrupt number = 58
Creating NVMPG module
UART1 DMA configured

## Entering START state

Starting the BASE thread
Registering interrupt for interrupt number = 25
        power on Timer 1

        timer started

Starting the SERVO thread
Registering interrupt for interrupt number = 28
        power on Timer 2

        timer started

## Entering IDLE state

---

pc side
cmd
linuxcnc remora-nvem.ini ok board is on net. but all move finished with following-error.
cmd-feedback is zerro for joint


I tried change IP in remora-eth.c or turn off card but result is same. linuxcnc start ok.

I would expect a message "can't connect"

rtapi_print("ERROR: can't connect: %s\n", strerror(errno)); //remora-eth.c

I want to know if there is some debug information after remora-eth connect to card?
Or what is wrong.

M

If you don't see the base thread objects in the rs232 output then your config on the nvem2 is not correct. Then it won't give a feedback, so you will get a following error.
Base thread object

X - Joint 0 step generator

Creating a std module
Creating Pin @
port = GPIOE
pin = 15
Creating Pin @
port = GPIOE
pin = 14

You could try to flash the pe-compiled firmware then flash the config as explained by spdbmp927 forum.linuxcnc.org/18-computer/44828-rem...ard?start=170#240666

Or maybe you had not the right branch? This firmware should work after you upload the config:
github.com/scottalford75/Remora-NVEM/tre...json_config/Firmware

royka wrote:

If you don't see the base thread objects in the rs232 output then your config on the nvem2 is not correct. Then it won't give a feedback, so you will get a following error.
Base thread object

X - Joint 0 step generator

Creating a std module
Creating Pin @
port = GPIOE
pin = 15
Creating Pin @
port = GPIOE
pin = 14

You could try to flash the pe-compiled firmware then flash the config as explained by spdbmp927 forum.linuxcnc.org/18-computer/44828-rem...ard?start=170#240666

Or maybe you had not the right branch? This firmware should work after you upload the config:
github.com/scottalford75/Remora-NVEM/tre...json_config/Firmware

I see step gens on rs232 output.
I tried flash precompiled and myself compiled firmware and them flash config. All is ok. card is running. 
I tried flash all tree branches (main, devel, features/json) but I can't connect to card from pc side.
Is it normal to start remora-eth (loadrt) without nvem card? There is no check for connecton?
my problem is on pc side I thing. network is ok  because mesa on 10.10.10.10 is runnig. 
M
 

Hi,

You will know if you have a connection to the board when LinuxCNC will come out of E-Stop. This is the check for a connection. What is your output when you run ifconfig?

scotta post=241746 userid=24854Roy, FYI I'm still looking at the following errors. Something really strange going on somewhere as the errors are consistently the same magnitude. I could test by increasing the max following error so that the program would keep running. The same magnitude following error spike on different joints. Makes me think it's something to do with DDS accumulator over flow that Rafal observed in the other thread. Anyway, Thanks for your help and patience.

Scott, Thanks for sharing that i was wondering id it was just mine. I am getting following error also. I have recorded what appear to be jumps in position FB from the NVEM that are consistently ~420, or ~60 counts. I have seen only these two values, in both directions, on all 3 axes. I decoded some packets to see what I could see and got some cool graphs like this:
 
Very weird. I modified my Remora-eth.c so that it does not update the accumulator on large sudden jumps. Instead I have it modeling feedback for the cycles in question to try and minimize position loss. Im going to test it out later today, but maybe it can help alleviate position errors until we can find the problem.
If there is anything I can do to help track this down let me know.
 

scotta wrote:

Hi,

You will know if you have a connection to the board when LinuxCNC will come out of E-Stop. This is the check for a connection. What is your output when you run ifconfig?

Thanks scotta,
i had no connect stop to 12V, now it is ok. 
after estop and machine on rs232 say 
## Entering RUNNING state
and now  I am able to change output pins .
Martin

 

spdbmp927: I have the same problem. But it looks like it's related to working on a laptop. High latency. Whenever he writes me a joint error. If I slow down, that's fine.

Hi Domi, when using the NVEM board latency of the PC is not an issue. The servo thread runs at 1khz in which a bit of jitter is not an issue. If the following error disappears when you slow down the motion then it would point to a configuration issue where the joint max velocity is not higher than the axis max velocity. So it cannot speed up enough to recover the following error.