From: Joerg Hansmann (info_at_jhansmann.de)
Date: 2002-01-25 02:43:58
Hi Andreas,
----- Original Message -----
From: sleeper75se <sleeper75se_at_yahoo.se>
To: <buildcheapeeg_at_yahoogroups.com>
Sent: Thursday, January 24, 2002 2:06 PM
Subject: [buildcheapeeg] Re: virtual ground problem
> --- In buildcheapeeg_at_yahoogroups.com, "Joerg Hansmann" <info_at_jhansmann.de> wrote:
>
> Hi Joerg,
>
> you appear to have full control over the math
No, not really. It is only a bit feedback theory and
a handful of SPICE.
> and the comments I make
> are made without any real calculations to back it up, so this might
> be irrelevant. :-)
>
> Ok, the point I was trying to raise is:
>
> Wouldn't it be possible to treat the virtual ground purely as a
> reference signal without any capacitances loading it down at all?
It could be done, but without capacitive bypassing
the impedance of the virtual ground will become 100Ohms and
more for HF.
> If
> the amplifiers were treated as single-supply amps, then they would
> only be decoupled across V+/V- without involving virtual ground.
>
> Consequently, the capacitive loading of the virtual ground amplifier
> could be reduced to strays only and perhaps the ringing you
> experience would cease?
No capacitors would mean no ringing but bad HF properties.
On the other hand bypassing can be easily done with a tantalum
capacitor with the right ESR range (about 500mOhms to 1000mOhms)
and additional 100nF ceramics.
This is essential if the virtual ground is used for HF rejection
and ESD protection.
Input HF rejection could be alternatively done with Cs to digital
GND (or -2V). However my ESD/user protection circuit would not
work when connected to -2V.
Another advantage of bypassing the output of the rail splitter is
reducing the output noise level of the rail splitter (in this case a TLC272 op
amp).
BTW: Ringing can also appear with LDOs used with wrong bypassing.
How is this solved in TinyEEG ?
Regards,
Joerg
This archive was generated by hypermail 2.1.4 : 2002-07-27 12:28:37 BST