Page 1 of 1

Report on Job stats

Posted: Fri Sep 13, 2019 1:39 pm
by Oaklands
(Hopefully I didn't miss this in the docs, if it already exists)

It would be useful to get an optional report of the job stats, at the end of each Job, as follows:

- output to file only, to avoid cluttering the GUI
- Number of parts placed
- list of parts placed (footprint, tape ID etc without any potentially confidential details)
- # nozzle changes, with nozzle #
- Total duration of Job HH:MM:SS.mmm
- CPH number
- Option to enable/disable this reporting feature via a popup context menu(right click) on the Job tab.
- Opt in option to automatically upload stats to LP, might prove useful for future roadmaps?
- Bonus option would be timestamps at when each part is placed - allows better analysis of making a change impacting only one component.

Reason for this is that I saw in the docs that CPH in the range of 500-800 cph can be achieved (expected?). However, in my tests I am seeing around 250 cph on average - using place fast on tapes..with 1 nozzle change & duration of about 15 mins. So If I make a change, it would be nice to be able to quickly see if there was a performance boost. Might also be useful in comparing different software releases...such that a release with performance improvements can be published with a number for a particular reference job or jobs along with the PC spec.

I understand that we are not looking to have a 'Pro' machine and that the tape position, motor config, vision etc etc all impact placement pick & place job run-times.

Finally, the times I am currently achieving for test jobs is acceptable to me, even if slower than expected....but always a nice bonus if I could get between 500-800 cph or more. :)


PS: I am running on a 10 year old laptop which may slow the vision a bit. However, I did notice that when I upped the run-time priority of the LitePlacer application I got about a 10% job time boost (Obviously, if I had this feature I could have an exact measurement!). That is something that could be built into the application as an option to apply at run-time (needs admin privileges). Alternatively, can be done manually thru the Task Manger.

Re: Report on Job stats

Posted: Sat Sep 14, 2019 9:57 am
by JuKu
There is not a status log, but I agree, there should be. This is now on my “accepted fixes and improvements” list.

For speed, you want to look at the motor settings. On my machine, I see almost no accuracy penalty between 4 and 8 microsteps, but maximum speed increases. Acceleration is more important than speed in most cases. If you use slack correction, slow move speed matters. I haven’t done a good analysis, but it seems that fast slack correction actually improves accuracy as well. This kind of makes sense; the idea is to have the slack (if there is any) always on the same side, and if the correction move is fast, it makes it tighter.

Re: Report on Job stats

Posted: Sat Sep 14, 2019 10:19 pm
by wormball
For speed:

- Determine your optimal motor currents and speeds by XYZA experimentally. Mine are about half the maximum current by XY.

- You may want to use 1.8 degree motors. https://liteplacer.com/phpBB/viewtopic. ... 6142#p9379

- The Z axis range is quite excessive for smd placement, and the gantry is pretty heavy, so Z movements take a long time. So i restricted the upper Z limit about 20 mm above the table instead of 40 mm stock (It would be great if there will be such software setting independent of mechanical end switch position). However there was possibility to crash the nozzle into the nozzle holder, so i sunk the nozzle holder below the table level (in this case you need to set "Z0 to PCB" value below actual table level, cos otherwise it will refuse to change nozzles).

- You may also attach a spring to the Z assembly in order to (almost) compensate its weight so you can choose larger Z speeds (i did not tried this yet).

Re: Report on Job stats

Posted: Sun Sep 15, 2019 7:25 am
by JuKu
> otherwise it will refuse to change nozzles

Good idea. The z motor is powerful enough to cause damage if it crashes badly, so I’d rather keep the precautions. But a tick box “I’ve sunk the nozzle holder” for different limit for nozzle changes is worth doing.