Re: [buildcheapeeg] NaffEEG

From: Joerg Hansmann (info_at_jhansmann.de)
Date: 2002-02-01 12:21:18


Hi Jim,

----- Original Message -----
From: Jim Peters <jim_at_uazu.net>
To: <buildcheapeeg_at_yahoogroups.com>
Sent: Friday, February 01, 2002 2:13 AM
Subject: [buildcheapeeg] NaffEEG

> I've now built the BrainMaster input stage (using AD620 + OP90), and
> it seems to be working. I had to buy 5 of the polypropylene
> capacitors, so I picked a good matching pair anyway, just in case it
> made a difference.

10 nF capacitors ?
(see other thread about BM-input stage...)

> I'm now getting 50Hz hum, no problem at all! I'm sampling at 5000Hz
> (5kHz), and piping this through to my filterbank display that I posted
> PNGs of earlier.
>
> I've improvised some electrodes from 1 pence pieces tinned with solder
> on one side. I also experimented with little wrappers for them dipped
> in saline. (However, I'm guessing now that probably it's not a good
> idea to use electrodes with lead (Pb) in them).

I would like to know what electrode impedances you get with this low-cost-
solution. My guess is that without proper electrolyte (e.g. 10/20 paste wax)
the impedance is far too high. (1k to 50 kOhms is best matched with
AD620 IIRR)
Has there been a tendency, that the results were somewhat better when you
used saline as electrolyte ?

> I'm not seeing any brain-waves yet, though.

What total gain has your circuit ?

Have you tried to calibrate it with e.g. a simple 555 oscillating
at 10Hz and a voltage divider that gives an output of e.g 50uV ?

> With electrodes on my forehead (ground in the middle, other two either
> side), I get a clean trace with 50Hz hum in the middle. If I tap the
> left or right electrode, I get a pulse, but the middle one doesn't
> give one, or it only gives a little one. This seems to be more or
> less as expected.

Yes. I agree.

> If I frown or clench my teeth, I get patterns appearing, and they peak
> around two to four times the amplitude of the 50Hz hum.

It should be EMG voltage that looks like white (or whatever color ..) noise
when 50Hz is absent.

> So, what do I need to do to improve this situation ? Do I need to buy
> some proper electrodes and paste, or should I look at reducing this
> hum first ?

You can do both.

Try to disconnect the electrode cables and tie the electrode inputs
together to your analog ground. The hum should be gone now and you should
see the amplifier noise. (post picture, I would like to see it)

If you still have hum, place the circuit in a metal box connected to
analog ground. That has helped when I had 50Hz hum in RS232EEG.

> At the moment the electrode connection is passing through about 3m of
> microphone cable, with about 8 inches unshielded at the electrode end.
> Would shortening this cable make a significant improvement ?

Probably not.

> Can I improvise better electrodes ? I'm still looking for a UK
> supplier of Ag/AgCl electrodes.

Buy thin silver metal sheet, cut disks with 1 cm diameter and use this
as electrodes with 10/20 paste wax.

BTW: What isolation do you use ? Optocoupler ?
Isolation is not only good for your health, it also reduces common mode
hum voltage to about 1/10th ...

Regards,

Joerg



This archive was generated by hypermail 2.1.4 : 2002-07-27 12:28:37 BST