What is the worst latency value for machine

More
21 Oct 2019 06:39 #148445 by Dinuka_Shehan
Please tell me the what is the highest latency value that machine could work properly for machining stuff
With lowest hardware configuration without Mesa boards?

Please Log in or Create an account to join the conversation.

More
21 Oct 2019 11:55 #148462 by tommylight
Up to 100.000 is very usable, since you can get 25.000 steps per second with double stepping enabled.
You can use the machine even with 250.000, that gets you 10.000 steps/s, but it will limit the maximum velocity quite a bit.
I do have some small machines with Dell laptops with latency over 300.000, but those machines use direct drive with belts and the maximum speed it about 1 meter per minute, so with a base period of 500.000 they can still do way more than what i need.
When using microstepping, that speed limit will get very low.
The following user(s) said Thank You: Dinuka_Shehan

Please Log in or Create an account to join the conversation.

More
07 Sep 2020 06:33 #181097 by shadab
Taking a look at other threads I believe that Latency is relevant only for parallel connector boards. PCIe ethernet etc etc boards don't have to generate the Steps to drive the motors so Latency test is not a true measure of their performance.

Please Log in or Create an account to join the conversation.

More
07 Sep 2020 11:12 #181130 by tommylight

Taking a look at other threads I believe that Latency is relevant only for parallel connector boards.

It is relevant, but not nearly as much.

PCIe ethernet etc etc boards don't have to generate the Steps to drive the motors so Latency test is not a true measure of their performance.

They do generate steps and PWM and read encoders by themself and that is the reason why the PC latency is not to much important as long as they can keep the timing of 1ms they are good.

Please Log in or Create an account to join the conversation.

More
09 Sep 2020 20:10 - 09 Sep 2020 20:16 #181502 by Todd Zuercher
There isn't really a hard rule that says this works and that doesn't, it is much more of a soft answer of it depends... (I know not the answer you were looking for.)

The latency requirement really can depend on a number of factors. If the machine has need of generating steps, or reading encoder pulses in software (such as when using a parallel port), then the limiting factor will be the needed accuracy of your minimum sampling period (Base-thread.)

If those aren't a factor for your machine (if those things are done in hardware outside of Linuxcnc), then things such as the maximum speed of the machine combined with the maximum acceptable error tolerance combine to determine what the acceptable maximum latency should be. For example if your machine is very slow and/or you don't need particularly great accuracy, you can probably get away with worse (higher) latency.

Conversely, if you have a very high performance machine with high accelerations and want the highest possible precision. Low latency and a faster than normal Servo-thread will be needed.
Last edit: 09 Sep 2020 20:16 by Todd Zuercher.
The following user(s) said Thank You: tommylight

Please Log in or Create an account to join the conversation.

Time to create page: 0.226 seconds
Powered by Kunena Forum