Feature/joegwalsh/rts chip manipulation #8
Conversation
Splits the DF chip orienation function up by chiptype, Includes functionlity for COLDATA chips, which have different features. When the wrapper function for chip finding is called, this will run the COLDATA function several times taking the average This is to handle some inconsistencies with the sequence success Note, ColdADC has not been explitly implemented, I'm not sure if they have the same marks as the LArASICs but maybe this can be reused
Expands DF functions to recognize larger COLDATA socket
Code now uses main MoveChip function to handle operations considently, but still accessed through server with wrapper functions which call correct indices Will complain about incorrect inputs and will output consistently Will find direction at tray and adjust pick up suitably Will then use offsets at Up Facing Camera to correct socket position. Also accounts for socket drift on DAT Needs COLDATA pin handling implementing Should get COLDATA direction from text on chip though Pin analysis currently commented out, but is in the UFC chip orientation function and status returned in output Errors now handled with ErrorDictionary for consistency Outputs handled by LogResults function TODO: Include some recovery features for errors such as returning to tray, or requiring user input somehow
For some reason the Jump XY(...) function doesn't maintain handedness so added if statement to check which DAT tray Also some MSU updates to points and vision
Needed to resolve vision sequence conflict handling this is done in RC+ by importing sequences Also corrected a couple of minor types introduced in merge
tcontrer
left a comment
tcontrer
left a comment
There was a problem hiding this comment.
I made a few comments on things to clean up and double check. Could you also add brief functions descriptions to each of the new functions you have developed here?
QC/ChipTesting/RTS_Robot/Main.prg
Outdated
| ' Motor Off | ||
| ' PumpOff | ||
|
|
||
| 'Moved this to site selection so it always gets called |
There was a problem hiding this comment.
If this is all in SiteSelection now, can you remove the commented out version here? In general, should clean up main for general use
There was a problem hiding this comment.
I've removed the section that's now in SiteSelection. With regards to cleaning up Main more, I've added a Testing.prg file which can be added to the gitignore and then used to do local changes during testing without us all constantly changing main
| EndIf | ||
| Fend | ||
| Function DefineDirections As Int32 |
There was a problem hiding this comment.
The socket mezzanine and chip orientation are the same for each chip right now. Is that true, or does each site need to update these numbers from here?
There was a problem hiding this comment.
They're not quite the same, but I think this can be simplified down in a way which I can commit soon. I need to test with a robot first though.
| @@ -668,1167 +658,637 @@ Fend | |||
|
|
|||
| 'Function ChipBottomAnaly(chip_SN$ As String, fileNum As Integer, pallet_nr As Integer, col_nr As Integer, row_nr As Integer, DAT_NR As Integer, ByRef status As Integer, ByRef res() As Double) | |||
There was a problem hiding this comment.
Can we remove all of ChipBottomAnaly if it is no longer used and is commented out?
|
|
||
| VRun ChipBottom_Analy | ||
| Function PinsAnaly(id$ As String) As Integer |
There was a problem hiding this comment.
Can you add a brief function description?
| Print "Move chip from pallet(", src_pallet_nr, ",", src_pallet_col, ",", src_pallet_row, ")", | ||
| Print " to tray(", tgt_pallet_nr, ",", tgt_pallet_col, ",", tgt_pallet_row, ")" | ||
| status = MoveChipFromTrayToTray(src_pallet_nr, src_pallet_col, src_pallet_row, tgt_pallet_nr, tgt_pallet_col, tgt_pallet_row, 0) | ||
| status = MoveChipFromTrayToTray(src_pallet_nr, src_pallet_col, src_pallet_row, tgt_pallet_nr, tgt_pallet_col, tgt_pallet_row) ', 0) |
There was a problem hiding this comment.
Can you go through and check that the changes to functions here match the function calls in RTS_CFG.py?
There was a problem hiding this comment.
The input arguments for these haven't changed in RTS_Server.prg so as far as I can tell it still matches the integration code.
Adds some descriptions of inputs Removes some commented out code Cleans up Main.prg and adds Testing.prg which should be used after adding to .gitignore
Simplifies the offsets stored for each chip to a single offset The socket position code now returns the direction in which the chip should be inserted given these are the same for each socket Offset in SiteSelection:DefineDirection is now between hand U at the taught point, and the defined direction of the chip in the socket A set of Tutorial files are also added, with site name "TUT" This was for the sake of going through everything methodically from scratch at MSU and teaching new sequences etc. Will do some more testing to make sure not broken for different chip orientations Will then add some more user friendly wrappers for subfunctions i.e. GetChipFromSocket/Tray without all the necessary logging varaibles May add back in position offset handling at tray as well as socket
Had commented out a line I thought was redundant in going to socket point with a line in the socket alignment function which only moved to the XYU coordinate of the point to get the agl4 offset not to the correct Z height Now fixed so it doesn't just drop the chip form the camera position height
To avoid issues with git's own compression and character handling trying to zip up the project directly on Windows.
Removes some redundant MoveChip functions, people should now use the versions which account for potential alignment skews and corrects for these.
Breaks pick and place stages up into
GetChipFromTrayGetChipFromSocketPlaceChipInTrayPlaceChipInSocketPlease make sure you reimplement any site specific or chip specific code in the right places.
This is mainly an issue for the chip analysis with the UF camera which should now be done in the
UFGetChipAlignmentandUF_CHIP_FINDfunctions inRTS_tools.prgThe DFGet...Alignment functions should also be updated if necessary.
There are also some global alignment values which should be set in
DefineDirectionsfunction inSiteSelection.prgor in the RTS_tools.incNOTE: this has not been fully tested with COLDATA, and so there may be issues with the current functions, I've tried to head off any issues with the alignment functions but should be checked - Since the UF camera cannot easily identify the pin-1 corner, I've restricted offsets at the UF camera to be +/- 45 degrees, and we will rely on the DF camera to find the major direction (+n*90deg).
Also adds a new RTS_suberror
Next to implement
When there is an error, it should potentially move chip to the bad tray, not sure if this should be handled in the state machine or the RTS code. If in RTS code, need to allocate the bad tray position to place the chip in.
Also need to reorganize code as
RTS_tools.prghas become a bit unwieldy, but that should probably be a separate branchMight be useful to implement batch motion commands, so that the amount of back and forth when checking if trays/sockets are occupied before moving is reduced. I've started to put in the capability to do these group occupancy checks and then override them in the individual move functions but this may complicate logging and time stamping individual chips