Plating recording electrodes

439 views
Skip to first unread message

Balaji

unread,
Aug 18, 2015, 7:52:52 PM8/18/15
to Open Ephys
Hi,
I had a vague impression that I can pass current through the electrodes using the FPGA (isn't that how we measure impedance in the first place?). Is this true? Can I use this to plate my electrodes? Does anyone have a program that can do this that they are willing to share?
Balaji

Jon Newman

unread,
Aug 19, 2015, 11:48:54 AM8/19/15
to Balaji, Open Ephys
Hi Balaji,

RHDxxxx chips contain an onboard DAC and primitive voltage to AC current conversion circuit (series cap...) that can be used to pass somewhat known AC currents at electrodes while the corresponding time-varying voltage is measured through the same ADC that is used for recording. This allows you to calculate the electrode impedance for different time-varying current waveform. See pg. 30 of the RHDxxxx datasheet:


However, this circuit is unable to pass DC currents through the electrode which is required for plating. In order to plate, you can must tie an external, bipolar current source to the elec_test pin and then closing the switch that ties the elec_test pin to the electrode you want to activate (this provides a DC path from elec_test to the electrode, through the bath wire which is tied to GND or 3.3 volts to control the plating polarity). The RHAxxxx datasheet has a pretty good explanation that is basically the same as for the RHDxxxx chips on pg. 22


Intan and I have both created virtually identical secondary boards to do this. Unfortunately, my code is a disaster that is basically hacked together pieces from the Intan evaluation GUI and is specific to my plating board, which I would not recommend using now that Intan has created one that performs the same function. However, the Intan code to control the board is implemented in LabView or Matlab, and is quite expensive (something like ~2000 USD). It does allow you to do really cool things though, like automated closed-loop plating where the impedance measurements and plating are alternated until a desired electrode impedance is achieved. After I had created my design, Intan opened the design of their plating board. I forked it and made some minor modifications:

https://github.com/jonnew/impedance (in the folder impedance-pcb-intan)

I built the board in my lab without issue, but I'm not yet willing to pay 2000 dollars to use the Matlab control software. It would be great if we could incorporate a module for this board, and electrode plating/improved impedance measurement routines, as a components in the Open Ephys GUI (or maybe as a just small command line app, which is more my style...). I've been thinking a bit about how to do this, but frankly I've been pretty busy. If others are interested, we should talk about what the optimal solution would look like.

Something to should note: all voltages that interact with the RHDxxxx chips are limited to Vcc =~ 3.3V.  This include the plating voltage applied to the elec_test pin. Because DC resistance of electrode can be quite high, this severely limits the plating current that can be passed. This is why most stimulus isolation units can reach 100's of volts in order to pass a desired plating or simulation current.

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 post to this group, send email to open-...@googlegroups.com.
Visit this group at http://groups.google.com/group/open-ephys.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-ephys/45fbeb5c-d140-4b4f-baa4-c594e406fac7%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.



--
Jonathan Newman
Postdoctoral Fellow, MIT

Balaji

unread,
Aug 20, 2015, 3:55:55 PM8/20/15
to Open Ephys, balaji...@gmail.com, jpne...@mit.edu
Hi jon,
Thanks for a detailed answer. I'm excited to learn that someone got this working. And I'm excited to help in any way I can. 

Currently I am working my way through the datasheets to understand what all those extra pins mean. But I think I have a basic understanding of the issue at hand. I created a circuit hub project to make the PCB - although it is incomplete. Help identifying missing parts / verifying chosen parts or choosing better sources for parts is always welcome. 

I have a couple of questions to electroplating aficionados: 
1 .Use of lower voltages leads to slower plating right? So a 3.3 V source will plate 30 times slower than a 100 Volt source? Is that the only effect? Is there a minimal voltage barrier that I would need to cross to even enable plating? 
2. Is there any qualitative difference in electroplating when it is done slowly vs. done fast. Are the metals deposited in a different way? Is slow just as good/worse/better than fast plating?

I'd be glad to help with coding if it makes electroplating a solved issue within the community. 

-b

Brad Buran

unread,
May 9, 2016, 12:45:30 PM5/9/16
to Open Ephys, balaji...@gmail.com, jpne...@mit.edu
Hi Balaji,

What's the status of this project? Do you have a functional system working? I'm happy to flesh out the specs if needed.

Brad

Balaji S

unread,
May 10, 2016, 1:14:57 PM5/10/16
to Brad Buran, Open Ephys, jpne...@mit.edu
Hi Brad,
You will find Jon Newman's design in the circuit hub project. However, I balked at the initial cost of the build - especially given that I didn't know how functional it is going to be. But if more people are willing to give it a try - I'd like to get a functional version of this going. 

I believe that Jon has a working prototype but doesn't have code that runs with the setup. I can code this up with help. 

Jon - what do you think is the best way to proceed to get this functional?
-b


Brad Buran

unread,
May 11, 2016, 11:27:05 AM5/11/16
to Balaji S, Open Ephys, jpne...@mit.edu
Thanks. I will have to review this and see how well it will fit our needs. I should be able to help code this up (possibly as a plugin to the Open Ephys software), but it probably would be a long term project (I have a few other projects I need to finish up first).

Something to note is that this electroplating system will only work with the 128-channel headstage because that's the only one that supports the electroplating input. It's probably possible to redesign the lower channel count headstages to support electroplating.

Regarding the cost, it's quite cheap per unit once we order more than one. CircuitHub may be willing to facilitate cost-splitting. Currently OHSU where I work does not allow costs to be shared with other universities.

If I decide to purchase the 128-channel headstage, would you be interested in cost-sharing on CircuitHub to order a prototype electroplating board? Cost is $337 per board if we order two.

Brad

Balaji S

unread,
May 11, 2016, 1:38:36 PM5/11/16
to bbu...@alum.mit.edu, Open Ephys, jpne...@mit.edu
Thanks. I will have to review this and see how well it will fit our needs. I should be able to help code this up (possibly as a plugin to the Open Ephys software), but it probably would be a long term project (I have a few other projects I need to finish up first).

Good. I'm getting more comfortable with the open-ephys software design - i'll be able to help. 

Something to note is that this electroplating system will only work with the 128-channel headstage because that's the only one that supports the electroplating input. It's probably possible to redesign the lower channel count headstages to support electroplating.

I hadn't realized this. This is more a fact about the lack of access to the 'elec_test' pin in the intan 32 channel PCB designs. The pin is still there on RHD2XXX chip. A potential solution would be to redesign the PCB to include access to the elec_test pin. Not sure how much work that would mean.


Regarding the cost, it's quite cheap per unit once we order more than one. CircuitHub may be willing to facilitate cost-splitting. Currently OHSU where I work does not allow costs to be shared with other universities.

If I decide to purchase the 128-channel headstage, would you be interested in cost-sharing on CircuitHub to order a prototype electroplating board? Cost is $337 per board if we order two.

That would be fine with me.
-b

Brad Buran

unread,
May 11, 2016, 1:48:10 PM5/11/16
to Balaji S, Open Ephys, jpne...@mit.edu

On Wed, May 11, 2016 at 10:38 AM, Balaji S <balaji...@gmail.com> wrote:
Something to note is that this electroplating system will only work with the 128-channel headstage because that's the only one that supports the electroplating input. It's probably possible to redesign the lower channel count headstages to support electroplating.

I hadn't realized this. This is more a fact about the lack of access to the 'elec_test' pin in the intan 32 channel PCB designs. The pin is still there on RHD2XXX chip. A potential solution would be to redesign the PCB to include access to the elec_test pin. Not sure how much work that would mean.


It shouldn't be hard to redesign the PCB designs for up to 32 because they provide QFN package for the RHD2XXX chip. For the 64 channel design one would have to find a chip on wire manufacturer to handle the format (IDAX can handle this).

I think Intan would be happy to assist with validating any design that's developed. Would be interesting to design one with multiple connectors for various designs but the cost of the Omnetics connectors, particularly the high pin count ones, would rapidly inflate the cost of the board.

Brad

Jonathan Newman

unread,
May 17, 2016, 7:51:25 PM5/17/16
to Open Ephys, balaji...@gmail.com, jpne...@mit.edu, bbu...@alum.mit.edu
Since Reid shared his design for the electroplating board, I just made them myself instead of using my own untested design.

Software -- I have not had time to look into creating the plating software

Adding electroplating input: circuit is very easy. It just requires adding a very simple set of analog switches to the headstage


In this diagram:

Bath goes to bath ground
Ref goes to ref pin on intan chip
Elec test pins go to elec_test pins on intan chips.
T1, T2, T3 digital lines control weather plating circuits are open or closed and the direction of current.

WRT cost on circuit hub -- if you can just put the PCB together yourself it will be like 1/10th of the price.

Jon
Reply all
Reply to author
Forward
0 new messages