[Openeeg-list] Trimming is changing frequency! Help!

[Anton K. <ant...@gm...>]

Hi there,

I've recently started up working on my v1.1.0 ModularEEG again, and have hit
the same stumbling block. I have built both boards myself, and am looking at
the output on Electric Guru.
The problem is that when I turn the screws on the trimpot, the frequency
changes instead of the amplitude. This occurs when a channel is connected to
U_cal and Cal_GND, but also  when it is unconnected it allows me to change
the frequency of an irregular square wave (or at least it looks square
because it hits the top and bottom of the graph on Electric Guru). The
amplitude does not change whatever I do.

I've spent many days now trying to find what's wrong, it's very frustrating
considering I am so close to completion.
The digital board seems fine, when the analog board is not connected
electric guru shows a flat line at 0. I am running the P2 firmware.

Now on the analog board:
When the trimpot has a low resistance (measured between leg 1 of P202/3 and
VGND) the frequency is very short, when I turn the screw and the resistance
increases, the frequency rises. The amplitude is always at maximum (between
0 and 1023) on electric guru.

In the testing tips it recommended measuring the voltages between AGND and
some IC pins on the analog board. There were problems here.

Checkpoint Voltage
VGND               +2V correct
IC201 pin 1        +2V correct
IC202/3 pin 6     +2V correct
IC205/6 pin 1     +2V correct
IC205/6 pin 7     WRONG! was 4V instead of 2V. (This could explain it maxing
out electric guru...)

The resistors seem to check out fine. But measuring the resistances between
various pins of the IC didn't always produce the results that were described
on previous emails (although that was for v0.4 whereas I have v1.1.0 so I
don't know if it's relevant). Would any other readings be helpful?

Now, my suspicion is that it is some of the ICs that are faulty, but really,
I don't know and it's highly frustrating.

Does anyone else think this is the case? Has anyone had this problem before?

Any help would be appreciated, thank you.

Anton Kan