neuropixels
neuropixels copied to clipboard
cortex-lab
multiple probes
Hello, I have recently started to use Neuropixels (3B staggered) and am aiming to perform chronic, bilateral recordings, with 2 probes mounted on small motors (nanodrives) to have the flexibility of moving them separately. I have now encountered for the second time the same issue, which is that I get very bad noise once both probes are connected, whereas the signal is just fine for as long as only one of them is hooked up (suspecting some weird ground loops). Did anyone have the same problem before? Could it be problematic that the two probes, in particular their flex bases, are touching each other at the midline (hard to avoid because of the position of the targeted areas; attached a pic (sorry for the bad resolution, but to give you an idea how it looks like, probes glued to two nanodrives and still fixed to a stereotactic holder)? If you have multiple probes, you recommend having separate ground/references, or you connect them all together with one wire? all the best! lorenz
My setup was always like this: one reference on the skull (either skull screw or saline bath with silver wire in it); each probe has ground shorted to external reference; each probe has this gnd/extref line connected to the single reference point on the skull (so, the probes all converge on this point.
I would worry that if there is a lot of force from the two probes touching each other, then the noise might be coming from some strain on the circuitry. Is it possible for your experiment to angle the probes slightly so they tilt away from each other and avoid the contact?
Many thanks for your answer & advice! Re probes touching: problem is the flex bases are never straight but rather slightly curved. Tilting them may be one possibility, gluing them onto the motors the other way around another one (both a bit tricky). Will try. all the best, lorenz
Hello, when acquiring data with multiple probes (using SpikeGLX), what procedure/processing steps are you following to ensure the data is accurately synchronized (i perform long, chronic recordings, and even after calibration there will be significant error in the end)? I imagine you calibrate the headstages, record a 1Hz sync waveform in your experiments, and use TPrime posthoc? i would be grateful for your advice! all the best, lorenz
Hi -
I'm not familiar with TPrime, however, I would suggest that the only way to be certain for yourself is to duplicate a signal into the auxiliary channels for each probe, and align the timestamps of the digital events recorded in each post-hoc. Then spike times can be aligned using the same conversion. The signal that you use could be a 1Hz regular waveform, though it could be safer to use a waveform that has structure, e.g. randomly-timed flips or pulses that have meaning like a timing code - that way it will always be unambiguous which part of one recording matches which part of the other. This process will also work whether you calibrate headstages or not.
Nick
On Mon, Mar 2, 2020 at 3:25 AM lorenzfenk [email protected] wrote:
Hello, when acquiring data with multiple probes (using SpikeGLX), what procedure/processing steps are you following to ensure the data is accurately synchronized (i perform long, chronic recordings, and even after calibration there will be significant error in the end)? I imagine you calibrate the headstages, record a 1Hz sync waveform in your experiments, and use TPrime posthoc? i would be grateful for your advice! all the best, lorenz
— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/cortex-lab/neuropixels/issues/33?email_source=notifications&email_token=ABZ5IPZBFPLLBC7JIA77DYDRFOCTLA5CNFSM4KQJ4WY2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOENOYJWQ#issuecomment-593331418, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZ5IP52TZLJ45K24MHUXPTRFOCTLANCNFSM4KQJ4WYQ .
Many thanks for your answer & advice! Re probes touching: problem is the flex bases are never straight but rather slightly curved. Tilting them may be one possibility, gluing them onto the motors the other way around another one (both a bit tricky). Will try. all the best, lorenz
Hi @lorenzfenk
We have similar problems. We use two probes folded in a holder. If we use both probes one has noise . Everything is fine if only one of them is connected. The same method is used in a different Lab with no issues, but this is the third time in a row we have this issue. What did you do to solve your problem ?
Best, Andres
Hi @andresdegroot,
First off, I am really very sorry for the late reply, it completely slipped my mind!
Re the problem of noise when using multiple probes: I actually never had the problem after following @nsteinme advice. This means I avoid putting a lot of force on the probes when they are touching each other; and I shorten the REF/GND of all probes, so that i have a single REF/GND line that I then connect to a single reference point, in my case a stainless steel wire placed into the craniotomy.
Because I changed two things at once, it is not completely clear what has made the difference, but I strongly suspect it was the fact that I shortened all REF/GND wires that I then connect to a single reference point. As mentioned before, I didn't have any problems since I follow this procedure, with up to three chronically implanted probes, and even when they slightly touch each other.
Hope that is still of any use! All the best! Lorenz
Hi @lorenzfenk
Thanks for the reply. We tried all of it, but we still have the issue. We will look further. Hopefully we will find a solution.
Thanks Andres