Less Z travel during Job run => Easy Speedup?

[AnalysIR]

During a Job run, the LP travels back to Z 0 after pick-up and after placement.

Would it be possible to create a concept of a "Job Low Ceiling" (alternative Z 0 only during a job run & only for pickup/placement).

This way the LP would not have to travel back up to Z 0 twice for every component.

For example on my LP, there is no obstacle higher than ~5mm off the table, other than the nozzles. So if the LP moved up only 10mm instead of the 25-30mm, some travel time would be saved.

Obviously, this should not apply for nozzle changes or possible custom feeders. As it would be a configurable parameter, each user could set their own to suit.

If I am not mistaken, the time savings could be approx 1->2 seconds per component. So with say 100 components per job up to 1.5->2 minutes could be shaved off the run time.

I suspect there could be some issues with camera measurements made at Z 0 for pickups. However, it should be possible to calibrate for this or just do the initial 'Place Fast' at Z 0.

a related idea:
On the Z axis when the LP homes to Z, it goes to the limit switch and backs off a few mm and sets this position as Z0. Would it be possible to allow users to configure this back-off distance to make Z 0 closer to the table?

After some searching:

$xZB Homing Zero Backoff Distance to back off switch before setting machine coordinate system zero. Usually no more than a few millimeters

I see there is a setting for TinyG. I will play with this later on. However, I suspect this will require re-calibration....so it would still be nice to see a solution like the one above and a configurable parameter to config the Z backoff distance. (Might be useful to have on all axes)

UPDATE RE: Z Home back-off I mentioned above, I initially set the Z 0 position to be just above the top of the nozzle block...so I am sorted wrt backoff adjustment as I cannot set it any lower.

[JuKu]

Good idea! Added to the todo list.

[wormball]

I use mechanical delimiter for this purpose. However sometimes the liteplacer gets mad and bends my end switch.

[AnalysIR]

Interesting way to workaround a code change . This is probably fine for you, as your nozzle change rack is mostly below the table surface.

In my case (and many others) who have it on top of the table, the only obvious solution(?) is a software update that Juha placed on the TODO list.
In this case the Z0 must be above the top of the bolts(high point) in the nozzle rack to avoid potential collisions/crashes.

Out of interest, have you noticed any or much of a speed increase with your own adjustment

[wormball]

I had not measured the actual time difference, but it should be about 0.6 seconds per part given 8000 mm/min and 20 mm difference.

Also you may notice some spring that compensates some part of the gantry weight and facilitates Z movement.

Sinking the nozzle holder is not so hard, however in the current software you should set Z0 to PCB distance below the actual table level cos otherwise it refuses to change nozzles.

[AnalysIR]

tnx...that could shave ~5 minutes off one of my panels

The purpose/workings of the spring is not immediately clear to me. If I am correct, it seems to kick in only when the gantry is raised towards Z0?

[wormball]

No, i am talking about the spring that is directly behind the delimiter and pulls the Z assembly upwards.

> Username:AnalysIR
> Location:Dublin, Ireland
> Interests:All about IR remote Control Webshop: https://www.AnalysIR.com/blog/

I think you should change your domain name to analys.ir.

[AnalysIR]

OH, I see the long spring behind the assembly now...didn't notice that before.

(We did consider that previously, but as most of our customers are in North America, the .com domain is more accepted there and internationally. I suspect it is better for SEO as well. For Ireland the domain suffix is .ie & .ir is for iran. FYI: It is not available )

[wormball]

For Ireland the domain suffix is .ie & .ir is for iran.

I totally forgot about it.