Re: [buildcheapeeg] Re: Sound Card-Mixer

From: Joerg Hansmann (info_at_jhansmann.de)
Date: 2001-05-16 10:54:29


Hi,

----- Original Message -----
From: Praveen Kumar <pkk7_at_y...>
To: <buildcheapeeg_at_yahoogroups.com>
Sent: Tuesday, May 15, 2001 6:49 PM
Subject: [buildcheapeeg] Re: Sound Card-Mixer

...
> However, if analog multiplexer is tried the system can
> become very cheap, indeed. But the problem of low
> frequency response remains as it is.

That is not quite right. The time-domain-multiplexing method
I described a time ago is capable of accurately measuring
DC voltages even though the soundcard has a input
high pass that blocks DC.

The trick is, that each input-channel voltage (between 0..4V)
is framed alternating with a low and a high reference voltage (e.g 0V and 4V).
It is true that the multiplexed signal still can have a DC-component, and that
this component will not get through the input hp of the soundcard.

However it does not matter because after sampling and digitizing
the DC-component of each input-voltage can be easily reconstructed
by subtracting the nearest (in time domain) 0V reference voltage.

Also gain errors (or even unknown gain) of the soundcard can easily
be compensated utilizing the reference frame voltages.

Linearized opto-Isolation could be done as Anand John has posted some time ago:

From: Anand John <deep_river79_at_y...>
To: <buildcheapeeg_at_yahoogroups.com>
Sent: Tuesday, May 01, 2001 2:35 PM
Subject: [buildcheapeeg] a different input circuit

However I share the opinion of Moritz, that the cost reduction
compared to a uC solution (e.g. AT90S4433) is only marginal.

A further disadvantage is, that decoding the multiplexed signal will
lead to more complex software and demands more CPU-power
from the PC.

Regards,

Joerg



This archive was generated by hypermail 2.1.4 : 2002-07-27 12:28:29 BST