Fix ReachTargetVoltage handling for too small resistance#2290
Conversation
25b0267 to
dbc993d
Compare
|
Acquisition continues after required sweeps have been acquired. It's been a while, but I recall it should stop automatically. |
We don't return voltages for all hardware types, so the current name is confusing. So let's rename it.
Missed in 85d6c5b (ReachTargetVoltage: Convert to V3 analysis function, 2020-11-20).
We want to only act specially on the very first sweep. But as we reset the index, we will land in this branch on every sweep if the resistance is too small. Let's prefer comparing to the number of acquired sweeps instead.
Missed in 85d6c5b (ReachTargetVoltage: Convert to V3 analysis function, 2020-11-20).
We know have 6 sweeps in the stimset, which triggers the finish condition as we need one sweep more than the targetVoltages wave. Also added is a test with multiple headstages.
That's the "new" approach we do in the modern analysis functions. ReachTargetVoltage was one of the first ones (2017) and does not follow that approach. We can fix that but this is IMHO tangent to the fix in this PR. In your pxp I see in the history some bug messages which need fixing. I tried to reproduce that here, but that did not work. So I would need a screen share for that. |
dbc993d to
10f81b5
Compare
|
@t-b I also get the error in main, so I guess we designed it such that the number of sweeps needs to match the size of DAScales (not an analysis param). |
We used to always use 100pA if the resistance is tool small, but the code obviously only wanted to do that only on the first sweep. Fix that.