bug: Thermocycler fails performing basic PCR program after lid closing/shifting temperatures #9175
Comments
More DataI built a new version of the Tc firmware locally from the opentrons-module repo which includes commits that allow the print out the thermistors values and uploaded it to the Tc in the event of failure. Here they are... ErrorResponse [line 43]: /dev/ot_module_thermocycler2: 'Received error response 'Error:Plate temperature is not uniform. T1: 2.0172 T2: 8.0703 T3: 4.2907 T4: 3.7157 T5: 8.7126 T6: 4.4932 T.Lid: 105.6111 T.sink: 61.8712 T_error: 0.0000 |
|
Thank you for the detailed recording and troubleshooting steps of this issue! The behavior you are seeing is not due to having the thermocycler turned off for a long time but due to the consecutive, multiple lid open & close commands (although, having the module start off from an ambient temperature could be aggravating the issue slightly). The issue lies in the way the lid's movement affects internal temperature reading. This behavior is documented in Opentrons/opentrons-modules#104. tl;dr of the issue is that every time the lid is moving, the thermocycler is unable to intelligently correct its block temperature. So each time the lid is opened or closed, the block/plate's temperature deviates slightly. It gets corrected immediately after the motion is completed. But if you have consecutive open and close commands, the thermocycler doesn't get any time to correct its plate's temperature; each consecutive lid command adds to the temperature deviation, ultimately making it large enough to result in a uniformity error. We don't have a fix planned for this behavior immediately as having the lid open and close multiple times consecutively is not something a protocol would typically need to do. That said, improving the lid functionality is something on our radar. |
This seems to be a different and more concerning issue. Would you mind creating a different ticket for it with the observed behavior and logs? Did this protocol also have consecutive lid open and close commands?
The |
|
Thanks for the thorough reply and all your hard work in particular addressing the Tc issues! I've been following your commits as I tried to debug. I'll try to adjust my protocols accordingly. I've been automating longer protocols which require multiple opening and closings of the Tc but will break them down. Though once this initially error happens and the Tc is "warmed up," I don't see this error happen again.
|
Yeah, I'll open a new ticket this week documenting the above mentioned error modality. Thanks again for you feedback on this. |
Overview
When running a simple PCR program the thermocycler (Tc) fails with a uniformity error after lid closing. Interestingly, this error only happens when the Tc has been turned off long enough for the lid and plate to cool down to the ambient temperature.
Steps to reproduce
*OT2 server version is 4.7.0.
*Tc firmware version is v1.1.0.
Documentation of error
I was able to reproduce this error 3 separate times using 2 different OT2s (A & B) and 3 different Tc's (A, B, & C). Below are 3 videos simultaneously showing the OT2 and app running in real-time up until they encounter the error. Below is also one API log which I downloaded and parsed only for the error below.
Videos
API Logs
Current behavior
mostly speculation on a brief review of a few tickets
I suspect that this behavior is the result of one, or both, already notes issues. Ticket #5602 documents this exact error where a plate uniformity error arises if the lid is opened or closed immediately after
set_block_temperature().However, the resolution to this bug is documented in ticket #5807 through the use of a
TEMP_THRESHOLDvalue calculated by averaging the last 10 polled Tc temperatures to determine a "fluctuation threshold". Given that this error only happens after the Tc has been turned off for several hours which is necessary for both the lid and plate to return to an ambient temperature that there's not enough history for it to poll such that the calculated "fluctuation threshold" is presumably to low. While not documented in this ticket, I have a similar issue where the Tc returns a plate uniformity error after holding at a temperature for 15 minutes and then shifts to another temperate afterward. For example, an RT-PCR reaction is first done by incubation at a specific temperature for several minutes followed by cycling through several different temperatures quickly for PCR. I would suspect that the acceptable TEMP_THRESHOLD` calculated from extended at a specific temperature would lead to such a small value that cycling out of that temperature would return a plate uniformity error only unless the Tc did not get to the next temperature quickly and without overshooting the target temperature.Expected behavior
For the thermocycler to be able to not fail when doing basic PCR protocols and not when shifting from 1 long incubation to another temperature.
The text was updated successfully, but these errors were encountered: