- Configuring LinuxCNC
- Advanced Configuration
- Maximum speed is clipped much below the capability of the servos
Maximum speed is clipped much below the capability of the servos
21 May 2018 20:33 #110985
by PCW
Replied by PCW on topic Maximum speed is clipped much below the capability of the servos
Can you post your entire (current) hal and ini files?
Please Log in or Create an account to join the conversation.
21 May 2018 21:00 - 21 May 2018 21:02 #110987
by Marcodi
Replied by Marcodi on topic Maximum speed is clipped much below the capability of the servos
Hello PCW,
We have a secondary problem, There is probably an error in the firmware of the 5i25_7i85sspx2. The stepgen03 doesn't exist on the cards. We guess there has been a human mistake in the numbering of the stepgen.
The PB3 card has 3 stepgens and 1 pwm signal but the PB2 card has only 2 stepgens and a pwm.
IO Connections for P3
Pin# I/O Pri. func Sec. func Chan Pin func Pin Dir
1 0 IOPort SSerial 0 RXData1 (In)
14 1 IOPort SSerial 0 TXData1 (Out)
2 2 IOPort PWMGen 0 PWM (Out)
15 3 IOPort PWMGen 0 Dir (Out)
3 4 IOPort StepGen 2 Step (Out)
16 5 IOPort StepGen 2 Dir (Out)
4 6 IOPort StepGen 1 Step (Out)
17 7 IOPort StepGen 1 Dir (Out)
5 8 IOPort StepGen 0 Step (Out)
6 9 IOPort StepGen 0 Dir (Out)
7 10 IOPort MuxedQCountSel 0 MuxSel0 (Out)
8 11 IOPort MuxedQCount 0 MuxQ-A (In)
9 12 IOPort MuxedQCount 0 MuxQ-B (In)
10 13 IOPort MuxedQCount 0 MuxQ-IDX (In)
11 14 IOPort MuxedQCount 1 MuxQ-A (In)
12 15 IOPort MuxedQCount 1 MuxQ-B (In)
13 16 IOPort MuxedQCount 1 MuxQ-IDX (In)
IO Connections for P2
Pin# I/O Pri. func Sec. func Chan Pin func Pin Dir
1 17 IOPort SSerial 0 RXData2 (In)
14 18 IOPort SSerial 0 TXData2 (Out)
2 19 IOPort PWMGen 1 PWM (Out)
15 20 IOPort PWMGen 1 Dir (Out)
3 21 IOPort StepGen 6 Step (Out)
16 22 IOPort StepGen 6 Dir (Out)
4 23 IOPort StepGen 5 Step (Out)
17 24 IOPort StepGen 5 Dir (Out)
5 25 IOPort StepGen 4 Step (Out)
6 26 IOPort StepGen 4 Dir (Out)
7 27 IOPort MuxedQCountSel 4 MuxSel0 (Out)
8 28 IOPort MuxedQCount 2 MuxQ-A (In)
9 29 IOPort MuxedQCount 2 MuxQ-B (In)
10 30 IOPort MuxedQCount 2 MuxQ-IDX (In)
11 31 IOPort MuxedQCount 3 MuxQ-A (In)
12 32 IOPort MuxedQCount 3 MuxQ-B (In)
13 33 IOPort MuxedQCount 3 MuxQ-IDX (In)
So stepgen 00,01,02 are existing and 04 and 05 aswell, but there is no 03 ( or 06 for that matter ). We believe there has been made an error in numbering. PNCconf also can not be used to select a third stepgen on the pb2 connected card. We are actually trying to get xilinx here to work, but the installation isn't going exactly as planned
Hope you can have a look and maybe we have found an important bug in this firmware.
We have a secondary problem, There is probably an error in the firmware of the 5i25_7i85sspx2. The stepgen03 doesn't exist on the cards. We guess there has been a human mistake in the numbering of the stepgen.
The PB3 card has 3 stepgens and 1 pwm signal but the PB2 card has only 2 stepgens and a pwm.
IO Connections for P3
Pin# I/O Pri. func Sec. func Chan Pin func Pin Dir
1 0 IOPort SSerial 0 RXData1 (In)
14 1 IOPort SSerial 0 TXData1 (Out)
2 2 IOPort PWMGen 0 PWM (Out)
15 3 IOPort PWMGen 0 Dir (Out)
3 4 IOPort StepGen 2 Step (Out)
16 5 IOPort StepGen 2 Dir (Out)
4 6 IOPort StepGen 1 Step (Out)
17 7 IOPort StepGen 1 Dir (Out)
5 8 IOPort StepGen 0 Step (Out)
6 9 IOPort StepGen 0 Dir (Out)
7 10 IOPort MuxedQCountSel 0 MuxSel0 (Out)
8 11 IOPort MuxedQCount 0 MuxQ-A (In)
9 12 IOPort MuxedQCount 0 MuxQ-B (In)
10 13 IOPort MuxedQCount 0 MuxQ-IDX (In)
11 14 IOPort MuxedQCount 1 MuxQ-A (In)
12 15 IOPort MuxedQCount 1 MuxQ-B (In)
13 16 IOPort MuxedQCount 1 MuxQ-IDX (In)
IO Connections for P2
Pin# I/O Pri. func Sec. func Chan Pin func Pin Dir
1 17 IOPort SSerial 0 RXData2 (In)
14 18 IOPort SSerial 0 TXData2 (Out)
2 19 IOPort PWMGen 1 PWM (Out)
15 20 IOPort PWMGen 1 Dir (Out)
3 21 IOPort StepGen 6 Step (Out)
16 22 IOPort StepGen 6 Dir (Out)
4 23 IOPort StepGen 5 Step (Out)
17 24 IOPort StepGen 5 Dir (Out)
5 25 IOPort StepGen 4 Step (Out)
6 26 IOPort StepGen 4 Dir (Out)
7 27 IOPort MuxedQCountSel 4 MuxSel0 (Out)
8 28 IOPort MuxedQCount 2 MuxQ-A (In)
9 29 IOPort MuxedQCount 2 MuxQ-B (In)
10 30 IOPort MuxedQCount 2 MuxQ-IDX (In)
11 31 IOPort MuxedQCount 3 MuxQ-A (In)
12 32 IOPort MuxedQCount 3 MuxQ-B (In)
13 33 IOPort MuxedQCount 3 MuxQ-IDX (In)
So stepgen 00,01,02 are existing and 04 and 05 aswell, but there is no 03 ( or 06 for that matter ). We believe there has been made an error in numbering. PNCconf also can not be used to select a third stepgen on the pb2 connected card. We are actually trying to get xilinx here to work, but the installation isn't going exactly as planned
Hope you can have a look and maybe we have found an important bug in this firmware.
Last edit: 21 May 2018 21:02 by Marcodi.
Please Log in or Create an account to join the conversation.
21 May 2018 21:48 #110993
by PCW
Replied by PCW on topic Maximum speed is clipped much below the capability of the servos
Sorry about the bitfile, Here's a fixed bit file/fixed pinout source:
You are probably the first person to try and use the second 5I25 connection with that bitfile
Your INI file still has 5 usec step and space times, these will limit you to about 100 KHz
or about 200 mm/s with a scale of 500 steps per mm
You are probably the first person to try and use the second 5I25 connection with that bitfile
Your INI file still has 5 usec step and space times, these will limit you to about 100 KHz
or about 200 mm/s with a scale of 500 steps per mm
Please Log in or Create an account to join the conversation.
21 May 2018 21:59 #110995
by Marcodi
Replied by Marcodi on topic Maximum speed is clipped much below the capability of the servos
Amazing!!!
Thanks for getting that fixed while i am still waiting for the xilinx download, and that would only be the start of the misery
We came to the conlusion because we could only find 2 dir signal on the card and the third position TB1 pin 5 was not generating a dir signal of +5V. So i guess uploading this new firmware will activate the 3rd stepgen on the P2 connector. This will probably sound obvious to you as yes, that is what i fixed, but to me that is not that obvious )). But i am very happy we can give it a new try tomorrow morning.
About the clipping problem. So, by setting our step and space times to 2.5usec we should be able to double our speed? Is that a correct assumption. Because you are extremely correct, the joint error comes at exactly 12000 mm/min. But that is too slow for the machine we have, it needs to run faster.
Thanks again for your help so far.
Regards,
Stephane and Janos
Thanks for getting that fixed while i am still waiting for the xilinx download, and that would only be the start of the misery
We came to the conlusion because we could only find 2 dir signal on the card and the third position TB1 pin 5 was not generating a dir signal of +5V. So i guess uploading this new firmware will activate the 3rd stepgen on the P2 connector. This will probably sound obvious to you as yes, that is what i fixed, but to me that is not that obvious )). But i am very happy we can give it a new try tomorrow morning.
About the clipping problem. So, by setting our step and space times to 2.5usec we should be able to double our speed? Is that a correct assumption. Because you are extremely correct, the joint error comes at exactly 12000 mm/min. But that is too slow for the machine we have, it needs to run faster.
Thanks again for your help so far.
Regards,
Stephane and Janos
Please Log in or Create an account to join the conversation.
21 May 2018 22:33 #110998
by PCW
Replied by PCW on topic Maximum speed is clipped much below the capability of the servos
Yes, reducing the step time and step space by a factor of 2 will double your maximum step rate,
that is the maximum stepgen step rate is 1/(steplength+stepspace)
You should also change these settings to about .01 (the .0005 value is appropriate for inch machines)
setp pid.N.maxerror .0005
(for N= x,y,z etc)
that is the maximum stepgen step rate is 1/(steplength+stepspace)
You should also change these settings to about .01 (the .0005 value is appropriate for inch machines)
setp pid.N.maxerror .0005
(for N= x,y,z etc)
The following user(s) said Thank You: Marcodi
Please Log in or Create an account to join the conversation.
21 May 2018 22:41 #111000
by Marcodi
Replied by Marcodi on topic Maximum speed is clipped much below the capability of the servos
I will make the changes and test it again.
Is there a way to change this clipping value in the code somewhere, to set it to a higher possible speed outcome?
Have a great night. And thanks again a lot for your help!!!
Is there a way to change this clipping value in the code somewhere, to set it to a higher possible speed outcome?
Have a great night. And thanks again a lot for your help!!!
Please Log in or Create an account to join the conversation.
21 May 2018 22:53 #111001
by PCW
Replied by PCW on topic Maximum speed is clipped much below the capability of the servos
The clipping warning just warns you that what you have asked for is not physically possible
(and will result in a a joint following error if you actually try to move faster than the timing allows)
That is the stepgen cannot obey 5000 ns step length and 5000 ns step space timing
constraints and generate a pulse frequency higher than 100 KHz because to do that it would have to
shorten one or both of those times
(and will result in a a joint following error if you actually try to move faster than the timing allows)
That is the stepgen cannot obey 5000 ns step length and 5000 ns step space timing
constraints and generate a pulse frequency higher than 100 KHz because to do that it would have to
shorten one or both of those times
Please Log in or Create an account to join the conversation.
- Configuring LinuxCNC
- Advanced Configuration
- Maximum speed is clipped much below the capability of the servos
Time to create page: 0.078 seconds