Another simulation question

I am trying to establish just how much difference speed and accel settings make to the final engrave time so as to decide whether it is worth it to push the limits of my hardware. I only have a GRBL system with a Nano controller so my overall speeds are limited. However in the job attached (road signs) I am getting a through put time of between 80 mins and 82 mins even with doubling or trebling the X&Y accel settings. I am trying this on my offline copy of LB v 9.06. The only thing that makes a real difference is changing the line interval from the current 0.15 to 0.2mm. The reduces the time to 60 mins. So my question is: are these numbers from the simulator correct or is there really no point in asking my hardware to perform better?

UPDATE:
Ok I have done some more experiments and the sim figures do seem to be correct (of course Oz).
Changing the Y axis accel from 400mm/s^ to 800mm/s^ only gives a decrease in throughput time of 0.1%. Increasing Y accel to 3000mm/s^ only gives an decrease of 0.2%. So it’s pretty much not worth the risk of stalling the Y axis by asking it to accelerate any quicker. That’s a good thing!
Increasing the X axis accel from 1000mm/s^ to 2000mm/s^ gives a decrease of 2% so a reasonable improvement in % terms but still only going from 82mins to 80min, which in the grand scheme of things isn’t really worth worrying too much about.
The biggest gain on this job comes simply from increasing the overall engraving speed from 8,000mm/min to 10,000mm/min with a decrease of 19%. down from 82 mins to 67mins and that is worth having! So my conclusion is that I am not going to bother pushing the boundaries of my axis acceleration and risk lost steps, but just to keep the engrave speed as high as possible while still getting good results. I feel this is a safer option as lost steps are unlikely during normal fast travel.
Anyway, I thought I would share in case anyone else is thinking about investing in different stepper drivers and power supplies etc to get more acceleration. I know the type of job will make a difference if it has very short moves, but I’ll do some more tests

Thanks very much

David

road sign arrow old road sign font.lbrn (11.0 KB)

You might also gain some time by grouping things differently. For example, if all the left text is one group, and the number column is another, scanning by groups might go faster than scanning all shapes at once, because you won’t traverse as much white space in between.

There is a tradeoff though - short scanning moves don’t hit your desired speed right away, so if your speed is high, sometimes it’s faster to keep going than to turn around. Experiment with both methods in the simulation and see if you gain anything.

Thanks for your input Oz, I’ll carry on playing. But I did some more timing tests yesterday on smaller jobs, both engraving and cutting and the main thing that came out of it is particularly for engraving the Y accel numbers have very little impact as the moving time which is pretty much all acceleration is so low in relation to the stopped time while the X is scanning. Even on most vector cutting jobs I tried it was the similar. So I am not going to sweat trying to improve the max acceleration, especially of the Y axis with the increased risk of losing steps or a big cost in improved hardware. X is a bit different because the only consequence of increasing the acceleration on that axis seems to be wobble.
Question for you Oz - I was wondering if there might be mileage in having a macro button to change the $120 parameter to increase the X acceleration for engraving to reduce the time, but have another lower value for precision vector engraving where the wobble would be a bad thing? what do you think?

Cheers

David

Hi Oz, just thought I would let you know that your suggestion was spot on. I grouped the words on the left and a separate group for the numbers on the right and checked “fill all groups together”. Time went from 82mins to 59mins! Good call! thanks.

Cheers

David

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.