Following error 6I25 - 7I77
04 Apr 2018 15:37 #108381
by jsskangas
Following error 6I25 - 7I77 was created by jsskangas
Hello
I updated my system from old 5I20 and servo cards to new 6I25 - 7I77 kit.
with previous setup I did run 30 000mm/min speed.
With the 6I25-7I77 my speed limits to 7592mm/min.
I traced this exact speed, there is no feed error before this.
MAX_OUTPUT = 500
ENCODER_SCALE = 8000
OUTPUT_SCALE = 500
OUTPUT_MIN_LIMIT = -500
OUTPUT_MAX_LIMIT = 500
I verified that these values are passed to pids and mesa 7I77 analogs.
There is no problem at servos they worked correctly with old mesas.
10V output value equals to 30 000mm/min speed, this is verifyed.
Mesa gives correct output voltages for drives.
7590mm/min equals to 760 rpm at servo and resolver toencoder emualtion in drives gives 20000 pulses per rev.
760 rpm equals to 12,7 1/s and this times 20 000 pulses this leads to 25.4 kHz
Im having really bad feeling that 6I25 is not quick enough to read encoder signals.
7I77 is running in serial mode 3 cause of 5Vdc MPG.
Wiring is same as with 5I20 + 7I33.
So no length difference or interference difference should be.
Any ideas what to check ?
I updated my system from old 5I20 and servo cards to new 6I25 - 7I77 kit.
with previous setup I did run 30 000mm/min speed.
With the 6I25-7I77 my speed limits to 7592mm/min.
I traced this exact speed, there is no feed error before this.
MAX_OUTPUT = 500
ENCODER_SCALE = 8000
OUTPUT_SCALE = 500
OUTPUT_MIN_LIMIT = -500
OUTPUT_MAX_LIMIT = 500
I verified that these values are passed to pids and mesa 7I77 analogs.
There is no problem at servos they worked correctly with old mesas.
10V output value equals to 30 000mm/min speed, this is verifyed.
Mesa gives correct output voltages for drives.
7590mm/min equals to 760 rpm at servo and resolver toencoder emualtion in drives gives 20000 pulses per rev.
760 rpm equals to 12,7 1/s and this times 20 000 pulses this leads to 25.4 kHz
Im having really bad feeling that 6I25 is not quick enough to read encoder signals.
7I77 is running in serial mode 3 cause of 5Vdc MPG.
Wiring is same as with 5I20 + 7I33.
So no length difference or interference difference should be.
Any ideas what to check ?
Please Log in or Create an account to join the conversation.
04 Apr 2018 15:48 #108384
by PCW
Replied by PCW on topic Following error 6I25 - 7I77
7590 mm/m = 1 Mhz encoder count rate at 8000 counts/mm this is faster than the multiplexed
encoders on the 7I77 can count with the digital filter on.
I would suggest
1. turn off digital filter on encoders
2. lower your emulated encoder resolution (do you really need 1/8 u resolution?)
encoders on the 7I77 can count with the digital filter on.
I would suggest
1. turn off digital filter on encoders
2. lower your emulated encoder resolution (do you really need 1/8 u resolution?)
Please Log in or Create an account to join the conversation.
04 Apr 2018 16:01 #108386
by jsskangas
Replied by jsskangas on topic Following error 6I25 - 7I77
Ok so My error was not taking considering that they are read in as 4x.
then its 1MHz.
Correct.
Yes I have practically a new japan made machine and I would like to use its accuracy and speed.
I can run 30 000 mm/min rapids. and accuracy around 0.004 mm.
I know that's more than I need but still....
I need at least 4 MHz reliable encoder count rate for each encoder.
5 axes + one spindel, is this possible with 7I77?
What settings I need to tweak?
then its 1MHz.
Correct.
Yes I have practically a new japan made machine and I would like to use its accuracy and speed.
I can run 30 000 mm/min rapids. and accuracy around 0.004 mm.
I know that's more than I need but still....
I need at least 4 MHz reliable encoder count rate for each encoder.
5 axes + one spindel, is this possible with 7I77?
What settings I need to tweak?
Please Log in or Create an account to join the conversation.
04 Apr 2018 16:31 - 04 Apr 2018 16:34 #108388
by PCW
Replied by PCW on topic Following error 6I25 - 7I77
4 MHz is above the 7I77s encoder specifications but I would try
1. Turning off the encoder filter bit as I mentioned before, something like:
setp hm2_5i25.0.encoder.00.filter false
2. You dont need 8000 counts per mm to retain .004 accuracy 4000 or 2000 should be fine
If that does not work there are some other tricks that may help
(1 MHz count rate = 8000 counts per mm * 7590/60)
1. Turning off the encoder filter bit as I mentioned before, something like:
setp hm2_5i25.0.encoder.00.filter false
2. You dont need 8000 counts per mm to retain .004 accuracy 4000 or 2000 should be fine
If that does not work there are some other tricks that may help
(1 MHz count rate = 8000 counts per mm * 7590/60)
Last edit: 04 Apr 2018 16:34 by PCW.
Please Log in or Create an account to join the conversation.
04 Apr 2018 16:52 #108390
by jsskangas
Replied by jsskangas on topic Following error 6I25 - 7I77
Thanks
Disabling filtering did the trick.
Now can safely run 30 000mm/min
Jussi
Disabling filtering did the trick.
Now can safely run 30 000mm/min
Jussi
Please Log in or Create an account to join the conversation.
10 Oct 2018 17:20 #118642
by jsskangas
Replied by jsskangas on topic Following error 6I25 - 7I77
So 7I77 is 1MHz capable.
Is this all axis combined?
Or can every 6 channel take 1MHz simultaneously?
Is this all axis combined?
Or can every 6 channel take 1MHz simultaneously?
Please Log in or Create an account to join the conversation.
10 Oct 2018 18:39 - 10 Oct 2018 18:39 #118646
by PCW
Replied by PCW on topic Following error 6I25 - 7I77
All 6 encoder counters (or say 12 if you have 2X 7I77s) can read up to ~1 MHz simultaneously
Last edit: 10 Oct 2018 18:39 by PCW.
Please Log in or Create an account to join the conversation.
10 Oct 2018 18:43 #118647
by jsskangas
Replied by jsskangas on topic Following error 6I25 - 7I77
Thanks.
It seems that I am loosing encoder pulses while running without filtering and above 1 MHz.
It seems that I am loosing encoder pulses while running without filtering and above 1 MHz.
Please Log in or Create an account to join the conversation.
10 Oct 2018 21:03 #118658
by PCW
Replied by PCW on topic Following error 6I25 - 7I77
Any significant asymmetry in the A/B signals will also limit the maximum count rate.
If you only need 3 axis of encoders, a config that does not use multiplexing is possible and will read up to 10 MHz or so
Also you may be able to bump up the mux sample rate a bit but ideally you would want to adjust the skew and the skew adjust is currently buggy (fixed but fix not pushed to master yet)
If you only need 3 axis of encoders, a config that does not use multiplexing is possible and will read up to 10 MHz or so
Also you may be able to bump up the mux sample rate a bit but ideally you would want to adjust the skew and the skew adjust is currently buggy (fixed but fix not pushed to master yet)
Please Log in or Create an account to join the conversation.
10 Oct 2018 21:13 - 10 Oct 2018 21:14 #118660
by jsskangas
Replied by jsskangas on topic Following error 6I25 - 7I77
Hello
I need 5 encoder inputs at least.
I'm running full 5 axis.
I did some testing just moment ago.
Now Im running 2000 counts per mm, 24000mm/min max speed.
filtering is on, but I still experience drift at least in X axis.
I need to investigate this pit more.
I already verified that it is not a mechanical problem, cause re-homing fixed axis drift.
Luckily I has abosolute resolvers in servos, so I can check driver position against linuxcnc position.
If these two differs , like I suspect, then I need to scope encoder line between servo driver and 7I77.
If not, then its pit more expensive case...
I need 5 encoder inputs at least.
I'm running full 5 axis.
I did some testing just moment ago.
Now Im running 2000 counts per mm, 24000mm/min max speed.
filtering is on, but I still experience drift at least in X axis.
I need to investigate this pit more.
I already verified that it is not a mechanical problem, cause re-homing fixed axis drift.
Luckily I has abosolute resolvers in servos, so I can check driver position against linuxcnc position.
If these two differs , like I suspect, then I need to scope encoder line between servo driver and 7I77.
If not, then its pit more expensive case...
Last edit: 10 Oct 2018 21:14 by jsskangas.
Please Log in or Create an account to join the conversation.
Time to create page: 0.096 seconds