-
Notifications
You must be signed in to change notification settings - Fork 3
/
Copy pathTrouble shooting list.txt
114 lines (83 loc) · 6.26 KB
/
Trouble shooting list.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
=======
1/18/2022
Occasionally, the timestamp in the EVENTS file resets without logging a 'reset event' (event code 1). This causes headfixation data not being plotted correctly in the analysis script.
The cause of this issue is currently unknown and not replicated. When this occurs, this issue currently can be fixed by manually inserting a reset event (unix time 1 -1) to the data file at the location where the resetting occured.
====================== Resolved =============================
(Please move the items above to below once fixed).
9/1/2021
Feedback from motor should provide pole position to master Arduino. If feedback is not recived (e.g. when motor is broken), the master arduino should pause and give an error message.
9/1/2021
Matlab GUI occasionally loss connection with the cage but does not know it. As a result, it does not show trial data, even though the Autocage itself is running along. This can be checked by clicking the "Read" button on the GUI and see if parameters are updated (e.g. weight reading). If spotted, this can be fixed disconnect and reconnect the COM again.
4/23/2021
On windows 10 OS, Matlab 2021, GUI occasionally crashes.
It would be a very useful feature if we could adjust the pole postion at least along A-P axis (ideally also along M-L axis in future version).
5/17/2021
Mice lick a lot but the protocol is stuck at the stage 0 even after 2000 trials with the performance >95%.
The screws/nuts in the priston connectors might be loose such that mice could easily escape. Check msg and see if the headfixation interval is increasing as expected (25s
at final stage). The performance at msg should exceed 12/20 (60%) to proceed to a next stage.
3/20/21
A series of empty circles in a row in homecage GUI (not those programmed empty circles showing up hourly before headfixation stage) - no licking detection.
This is sometimes due to the default setting of maximum resistance on lickometer(s). Lower resistance usually fixes the problem (counterclock wise).
Also check that the battery voltage is >8 volt.
Mice could develop strong bias during learning even both lickports dispense similar amount of water, which can't be corrected by programms in time.
Try moving lickports back to center of lickports (70 by default).
SD card read/write errors in "error message" box.
There is a slide switch along the side of SD cards. Repeated removal/insertion of SDs could lead to misalignment of the switch.
Make sure the switch is pushed to the end of slot to fully "unlocked" state.
It's a good habit to click "set&reward" whenever a SD card is moved or a new mouse is started. An error message would show up if something is wrong.
During analysis, the first date when a new mouse starts might be wrong.
This is likely due to failure to reset the Arduino master card on the day1. In this case, the date of 1st trial (Trials.txt) could be used to replace
the first date in Events.TXT. Going forward, reset Arduino master card when starting a new mouse. Use the Unix time of first row in Events.txt as starting time.
8/20/2020
Without SD card, the system is running some sort of statemachine. It should be stationary
(this will be fixed in version 1.0 update)
8/20/2020
When plugging in SD card, user should hit 'reset', otherwise the system will run previously saved variables,
which produces a range of weird behaviors
(this will be fixed in version 1.0 update)
8/20/2020
There are a lot of cross talk between the two lickport, particularly when there is water (e.g. touching right port
triggers a transient <2ms signal in the left). We need to change either the hardware and software to make the lick
detection more robust.
1) Hardware change: increase the capacitance of the lickport circuit detecting contact
2) Software change: We need to change the Arduino_Task on how PortIn is triggered. Currently a transient change
will trigger. We need to alter the following portion of the code by adding more than one timestep to the history.
PortIn should be triggered
// Determine port entry events
if ((PortInputLineValue[x] == HIGH) && (PortInputLineLastKnownStatus[x] == LOW)) {
PortInputLineLastKnownStatus[x] = HIGH;
CurrentEvent[nCurrentEvents] = Ev;
nCurrentEvents++;
}
8/20/2020
Without plugging into GUI, plug in SD card with "final config files". Hit reset, program does not run normally standalone. However, it works
well when connecting to GUI.
Due to noise when power is plugged in, triggering false 'early licks'.
8/20/2020
For a new mouse, in protocol 0, the program runs well when connected to the GUI. However, when GUI is disconnected, the programm stops running.
8/20/2020
Motor position not updated on GUI when plugged in. Motor position should be updated whenever a cage is selected.
(this will be fixed in version 1.0 update)
8/20/2020
GUI saves the progress locally. GUI do not refresh when disconnecting. If connecting to a new GUI window using a previously used mouse name,
the program pick up the progress from the previous box connected to GUI. Progress should be tied to the SD card, not to the board. GUI should
follow SD card. GUI should have NO MEMORY
(this will be fixed in version 1.0 update)
9/10/2020
Version 1.0
If a cage is connect to matlab GUI. Then disconnect from the GUI and USB. Cage will be stuck.
When in this state, loading new programs onto SD card won’t update. The cage is completely stuck
8/20/2020
Documentation of Testing should include a brief instruction on how to upload Arduino program that include the
following information:
1) Where to install Arduino program
2) install Arduino Due board from "Tool"
3) they need to select Arduino Due, be plugged into the programing port, and select the correct PORT
4) The user needs to copy the provided library into the library of Arduino program.
5) A brief instruction on how to open RealTerm and connect to the Arduino (Note that when running RealTerm,
Arduino needs to be plugged into regular port, NOT programming port.
10/6/2020
1)The version 1.0 system run for a few hours and stopped running.
2)When clicking "close" button in HomeCage_GUI to close the COM port, sometime Matlab was not responding (stuck).
User has to reset the Arduino_Master in order to get out of the above issues.
it seems that the issues are casued by the commuication between PC and arduino.