From: Joerg Hansmann (info_at_jhansmann.de)
Date: 2001-12-17 00:26:20
Hi Doug,
----- Original Message -----
From: Doug Sutherland <wearable_at_earthlink.net>
To: <buildcheapeeg_at_yahoogroups.com>
Sent: Saturday, December 15, 2001 2:19 AM
Subject: Re: [buildcheapeeg] Re: Necessity of ir isolation
> Hi Joerg,
>
> > Are any URLs / publications available that cover the above
> > mentioned effect - e.g. show that properly prepared and
> > attached EEG electrodes can get several meg ohms under
> > certain conditions (deep meditation)?
>
> I'm not sure if this is useful,
IMO it is a very good article...
> but the following article
> talks a lot about electrode impedance.
>
> Scalp electrode impedance, infection risk, and EEG data quality
> http://www.csi.uoregon.edu/members/ferree/pubs/EEG2001.pdf
Some key excerpts and my comments:
(BTW.: the article is from :
"Clinical Neurophysiology 112/3: 536-544 (2001)"
so it is really up to date when they talk of "modern EEG amplifers". )
"...
For normally distributed data, the absolute impedances and their possible mismatches will be related because a distribution of
higher impedance values will also tend to have higher mismatches, e.g., a set of scalp electrodes with 1-5 kOhms impedances will
have mismatches of at most about 4 kOhms , while a set with 10-50 kOhms impedances will have mismatches of at most about 40 kOhms .
Note also that as sponge electrodes dry the impedances drift up to 50-100 kOhms , but if all the electrodes dry together then the
mismatches remain at most about 50 kOhms .
Many modern EEG amplifiers have input impedances consisting of a resistive
component on the order of 200 MOhms. In our amplifier, this resistance is in parallel with a capacitive component on the order of 10
pF, providing a reactance of 265 MOhms at 60 Hz.
..."
10 pF is a very good value (low capacitance = high impedance = good ).
I think the input stage of the modularEEG should be revised for low
common-mode capacitance (guard-rings and traces, guarded cables to input connectors, etc.).
Assumed we meet 10pF at all, some things become obvious:
Unshielded (more exactly: unguarded) electrode cables with 50pF
(just estimated) and more will instantly decrease the input impedance
at 60Hz (or 50Hz) to 53 MOhms.
Also each electrode disk (e.g. 11 mm diameter) has an additional
capacitance (skin as dielectric )
So I conclude, that most of the commercial EEG specifications
with many GOhms or even TOhms input impedance are only true for
DC and very low frequencies. However where it would be really
useful (at 50/60Hz line frequency) they all decrease to
10MOhms..265MOhms.
(see pictures of MICROSIM simulation with
DRL and shield-driver deactivated:
"modularEEG-input_impedance_common_mode.png"
"modularEEG-input_impedance_diff_mode.png"
BTW:
The HF-blocking Cs: C24 and C25 should be changed from 1nF to
100pF otherwise the diff-mode impedance will drop below 10MegOhms
at too low frequencies.
)
As a result the potential divider effect
(
see "the potential divider effect" (Huhta and Webster, 1973; Pacela, 1967)
in "3.1 Influence of common mode voltage" of
www.biosemi.com/publications/artikel3.htm
)
diminishes the datasheet values of 120dB and more CMRR of the input
instrumentation amplifier to very moderate levels:
see MICROSIM simulation results:
"modularEEG-CMRR_50k mismatch_noDRL_noShield.png"
I have modeled theINA114 with infinite CMRR (idealized).
The influence of the potential divider effect results
in a very poor CMRR of about 60dB at 50/60Hz (see probe cursor)
This is far away from the 120dB datasheet value of the INA114
and is completely OP-AMP independant.
The only way to get a better over all CMRR is the
DRL and the shield-driver. (perhaps more simulations
in the next posting ...)
"...
To make numerical estimates, we assumed Zin = 200 MOhms. Assuming that with scalp abrasion Z1 and Z2 are at most 5 kOhms, the
maximum
signal loss is 0.0025%, which is completely negligible. Assuming that without scalp abrasion Z1 and Z2 are at most 50 kOhms, their
maximum signal loss is 0.025%, which is an order of magnitude larger but still completely negligible. Some older differential
amplifier systems have input impedances of closer to 10 MOhms. Even in this case, assuming electrode impedances up to 50 kOhms,
the maximum signal loss is 0.5%, which may still be negligible for most purposes. Thus signal attenuation is expected to be
insignificant without scalp abrasion, even when modestly high input-impedance amplifiers are used.
..."
10MOhms differential input impedance EEGs with a worst case signal loss
of 0.5% seem to be OK for the authors of this article.
Regards,
Joerg
This archive was generated by hypermail 2.1.4 : 2002-07-27 12:28:33 BST