Ground/reference set-up with the new 64ch Hirose EIB & low-profile headstage

[Kevan Kidder]

Hi everyone,

My lab has been using the open ephys system for several years and I've came to question what I was taught/assumed regarding the use of ground and references. So i was hoping someone here could help us as we launch into our new projects. We are now switching to the 64channel Hirose EIBs and low-profile headstage (also using the OE acquisition board, OE GUI, and intan SPI cables). I've scoured through the available documentation and I havent found answers to several questions I have regarding these new hardware. Below I've made images with my questions embedded.  

We aim to record LFP and single units. My newest project aims to record from dHPC, vHPC, & mPFC using this Hirose EIB. With the old intan 64ch EIBs, we used to connect one of the grounds to the shell & the other to a skull screw inserted the cerebellum. This is one of the things that I feel was incorrect with our past grounding methods. 

Also I'd like to note that this figure in the published shuttle drive paper has ground wires that dont seem to match up with the current 64-channel EIB (probably because its the 32-channel in the image?). Which has added to my confusion.

Best,

Kevan Kidder

Mizumori lab

[Jon Newman]

Hi Kevan,

Good question and images. Thanks for that. 

1. On the headstage itself, go ahead and check if REF and GND are connected (they don't appear to be because R1, which is a zero-ohm jumper, is not populated). If so, then you dont need to connect anything between the REF and GND pins on the EIB. If not, then you probably want to. You can do this by putting a solder bridge or zero-ohm chip resistor across the R1 contacts on the headstage. The truth is that the REF pin on unipolar versions of the Intan chip almost always should be tied to GND. The reasons for this is a sort of long story but are well documented in the online course materials here:  https://ahleighton.github.io/OE-ephys-course

2. With respect to your question about if GND should be connected to the same place in CSF: If you connect _any_ ground contact to any place in the CSF then _all grounds_ all the way up to basically the center pin on your wall outlet are connected to that point. The reason for this is that the connection between all these nodes is copper and that has a very low resistivity compared to anything biological or any electrode in your circuit. So, all you need to do is connect _one_ ground point to one place in the CSF and you are good to go. So, why do people sometimes use multiple ground wires? There are two reasons and one is better than the other IMO: 

- Redundancy: one wire might fail for some reason such as bone regrowth,  oxidation, etc. In a muli-electrode circuit, ground is the only contact you really, really need. So it's nice to have a backup.

- Brain equipotentiality. Can we somehow bring the brain's average voltage to around whatever ground is? Yes, with one, good low impedance ground electrode. Does it work better with 2? I dunno but I'm very dubious because you are increasing the brain/copper impedance by at most a factor of 2 and this number pales in comparison to the brain's high resistivity and copper's low resistivity and I can't see it being worth the risk of infection, surgery/healing times, etc,.

So, your strategy: "we used to connect one of the grounds to the shell & the other to a skull screw inserted into the cerebellum" seems very reasonable to me. I maybe would not want to make it all the way to cerebellum though, just to CSF. Try not to let any blood get on the tip of the screw because it can form an insulating layer.

3. The three additional pins are the electrical and optical stimulation connectors for the next generation headstage (https://github.com/open-ephys/onix-headstage-64), which is compatible with this EIB. They are not needed for the low-profile headstage and can be glued over if you want to hermetically seal the drive.

- Jon

--
You received this message because you are subscribed to the Google Groups "Open Ephys" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-ephys+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-ephys/33594de6-c594-4577-b8dd-fd3e0e80e658n%40googlegroups.com.

--

Jonathan Newman

Postdoctoral Fellow, MIT

http://www.mit.edu/~jpnewman/

[Jon Newman]

increasing the brain/copper impedance by at most a factor of 2 -> *decreasing* the brain to ground impedance by at most a factor of 2

[Kevan Kidder]

Hi Jon,

Thank you for clearing this up for me! The reason I was worried about about the two groundings is I thought maybe that each side of the EIB was only contacting its respective ground. Specifically I thought this might be the case with the old 64ch intan EIBS in which I used two 32-channels headstages. Anyways we will continue on with one CSF ground. 

Best,

Kevan