Page 1 of 1

Some bugs

Posted: Thu Nov 22, 2018 5:46 pm
by wormball
Hello!

Here is my bug collection.

1. Sometimes, when it measures placement depth, it starts to move up eagerly and does not stop even when upper end switch is triggered. About a second after this it issues an error, and in corresponding depth cell appears negative value about -39 mm. So if i do not reset this value, next time it will try to pick the part 4 cm higher than it can, giving limit switch error. It seems that it behaves so only if the Z homing speed is more than 2000.

2. On the other hand, if i reset Z value using "R. these Z's" or "R. place Z's" button, not only selected Z is reset, but also all Zs below it.

3. When i panelize a board with more than 10 repeats by both axes, some copies receive duplicate names, e. g. C1_115 may mean capacitor on board 11, 5 or on board 1,15. Also not all the boards are placed in this case. I think additional underscore between numbers should be added, e. g. C1_11_5.

4. When i load the same file, "placed" checkmarks are reset on the screen, but when i try to place these, it says "all components placed".

5. Not properly a bug, but it resets "placed" checkmarks very slowly, i think there is no such hard computational problems in this operation.

6. Where it performs visual homing, both coordinates received by homing dot measurement should be negated to get the true carriage position, but only X is negated, so by Y axis we get error twice as large as residual coordinate difference.

7. When i reset the tinyG by button, the "pump on" flag is not reset, and it does not start the pump next time.

8. When "show pixels" is set on the "setup cameras" tab, it moves only about half of desired way when i click on the camera image on tabs other than "setup cameras".

9. When i open "setup cameras" tab, the "show pixels" checkmark is set immediately.

10. When i open "nozzles setup" tab, the "full speed by Z axis" checkmark is set immediately, resulting in too fast nozzle attachment/detachment and stalls.

11. Sometimes when calibrating nozzles it attempts to move with enormous speeds. And it fails, of course. For example:

Code: Select all

Unload nozzle #2
 -- zero Z movement command --
ReadyEvent: zero movement command
CNC_XYA_m, x: 65.236, y: 186.339, a: 0
==> {"gc":"G0 X65.67762343 Y186.339 A0"}
<== {"r":{},"f":[1,0,37,138]}
<== {"sr":{"posx":21.041,"posy":78.922,"posa":359.994,"vel":77.32,"stat":5}}
<== {"qr":31,"qi":1,"qo":0}
<== {"sr":{"posx":48.172,"posy":144.211,"posa":141.186,"vel":122522.95}}
<== {"sr":{"posx":65.678,"posy":186.339,"posa":0.000,"vel":0.00,"stat":3}}
ReadyEvent stat
<== {"qr":32,"qi":0,"qo":1}
 -- zero Z movement command --
ReadyEvent: zero movement command
So it tried to achieve 122522.95 mm/min, while program settings were 48 000 (X) and 40 000 (Y). Most likely it's tinyG bug, but interestingly i had not observed this in cases other than nozzle calibration.

12. On the other hand, also when calibrating nozzles, it starts to move with too slow speeds even when calibration is complete. And it reverts to normal speeds only when the software is restarted. It issues G1 commands rather than G0 in this case, so it is not the tinyG bug.

13. When i jog with F keys while nozzle is down, it says "attempt to move while nozzle is down". But when i jog with keypad, all is ok, so i crashed the nozzle several times this way.

14. When i pause and resume "place fast" operation, it re-measures the tape, but counts the same quantity of parts as if no parts were placed yet. For example, if i placed 4 of 100 parts, it will search hole corresponding to part 105, which is most probably absent. And if i placed 90 of 100 parts, it will search part number 191, which is well outside the working area. But it clearly does not need this parts cos it will place parts with number no more than 100.

15. If i stop the placement with stop button or esc key, it says "all components placed" and moves the carriage in the parking position. But if i stopped the machine cos it stalled, this parking movement will likely crash it. So i think it is not needed in this case.

16. Maybe not a bug, but i think that "test X" and similar buttons should move the carriage in only one direction at one push by similar reasons. If it stalls halfway moving forth and does not stall moving back, the gantry will hit the frame at the full speed. So i do not use these buttons to test maximum speeds and use manual coordinate setup instead.

17. When i load pick and place file, sometimes it emits pretty obscure message:

Code: Select all

---------------------------
Can't read CAD file
---------------------------
Error in file, Msg: Index was out of range. Must be non-negative and less than the size of the collection.

Parameter name: index
---------------------------
ОК   
---------------------------
It took me some time and effort to figure out that it does not like empty cells in csv files. I think it should at least say something like "error: empty cell in csv file" with row and column numbers.

Re: Some bugs

Posted: Fri Nov 23, 2018 8:56 am
by JuKu
Thank you, I appreciate these very much. I can't fix what I don't know is broken, and nowadays, I don't actually use my machine very much for real work. (I know I should...)

If you happen to hit bugs again, in some cases the full log file might be helpful (click inside the log window, ctrl+A to select all, copy and paste to a text document).

Re: Some bugs

Posted: Fri Nov 23, 2018 7:51 pm
by wormball
18. When i edit "Z0 to PCB" setting, it does not have any effect. So it can be changed only by measurement, not by editing, despite being editable. I do not remember now, whether it reverts to previous value after tab switching and whether this bug affect nearby fields.

19. Once i got "unhandled exception" error. However the software had been running several days and survived several hibernations. Here is the error message:
unhandled_exception.log
(7.05 KiB) Downloaded 539 times
20. After the mentioned error i found that not all the rows had been saved in the lpj file. I suspect that the absent rows are those which i moved up or down. However, i find moving rows in the text editor more convenient than in the software.

And here is the log from manifestation of the first bug:
liteplacerbreakstheceiling.log
(837.4 KiB) Downloaded 510 times
wormball wrote:1. Sometimes, when it measures placement depth, it starts to move up eagerly and does not stop even when upper end switch is triggered. About a second after this it issues an error, and in corresponding depth cell appears negative value about -39 mm. So if i do not reset this value, next time it will try to pick the part 4 cm higher than it can, giving limit switch error. It seems that it behaves so only if the Z homing speed is more than 2000.

Re: Some bugs

Posted: Fri Dec 21, 2018 6:06 pm
by wormball
21. When adding tape, it fills Y value not only in Y field, but also in "rotation" field. And it seems that it simply ignores this "rotation" field.

22. When setting actions for multiple parts, sometimes (often) it sets actions only for clicked part, not for all selected parts.

23. It seems it performs wrong rotation when placing parts on the back side. At least today i had to set 180 degree rotation for three different parts which as i remember were placed ok with 0 degree rotation when destined to the front side. Orientation was +X.