Impedance testing: frequencies and capacitances used

[Chris Rodgers]

I'm wondering about the "impedance check" function of the Intan headstages, as implemented in the OpenEphys GUI. I assume that when I click "Measure Impedance" in the FPGA node, it measures it at 1KHz since this is the standard frequency. Is that true?

But why does it keep changing the capacitance? There are messages like "Setting capacitance to 10pF". The fact that it takes longer to check the impedance when the capacitance changes makes me think that this is a mechanism for checking Z at frequencies other than 1KHz. If so, is that data accessible somewhere?

[Antonin Blot]

I have the exact same question. The part of RHD2000Thread.cpp doing the impedance measurement is a bit complex. I assumed the frequency is 1000 (set line 2136 if I understood correctly) and the impedance is measurement three times with different Cseries but some comments also refer to a "impedancefreqdialog.cpp" that I haven't found.

Can you confirm that you measure the impedance at 1kHz and that, for now, there is no way to change that frequency on the GUI?

[Jon Newman]

Assuming the code actually works, which is questionable given its complexity, the impedance check is simply using the on board current generator and the intan chip and measuring the resultant voltage on the native channels. Capacitance is calculated via phase shift. 

See page 29 of intan data sheet:

http://intantech.com/files/Intan_RHD2000_series_datasheet.pdf

Some potential reasons:

- current generation is simply using a DAC in series with capacitor, which works if we assume the capacitor is absolutely perfect. Given that these are 0.1 to 10pF, this claim is pretty questionable since its going to be competing with weird parasitics all over the place. I assume Reid did the necessary tests to make sure this was not an issue.

- The voltage across the microwire during a measurement is going to depend on its impedance and the current you inject. You would want to tune the injected current in order to fill the ADCs +/-5 mV dyanamic range by in order to get repeatable measurements. This involves changing the series capacitance and DAC waveform adaptively for each electrode until you got a large but non-saturating voltage back. Im unwilling to look through the code, but if it does not do this, it might be the reason for non-repeatable measurements.

--
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 https://groups.google.com/group/open-ephys.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-ephys/13dc766c-14d5-471d-b518-75c16bed1b30%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Jonathan Newman

Postdoctoral Fellow, MIT

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