Bias register selection
I looked at the current code and have a question that I hope someone can answer.
For Rev 2 of the board negative inputs are used with SRB2 as the reference for channels 1-7 and possibly 8.
When enabling a channel the SRB2 bit is set and it seems you are turning on the bit in BIAS_SENSN. Shouldn't you be turning on the appropriate bit in BIAS_SENSP? It seems like the reference drive would attenuate which ever single signal or add noise to multiple signals that are active. It should be better to not set any bits in BIAS_SENSN, right?
Also, with the current firmware, how do you enable channel 8 without connecting SRB2 to INP? Wasn't the intent that channel 8 could be a bipolar input?
I am working with the ADS1299 development board as a proof of concept before making my circuit board. I will be building a Nintendo Wii controller for my dad who has ALS. Along with EOG/fEMG I will have a 9 axis IMU and pressure sensor for sip-puff.
I hope that I can either learn why you have configured it like this, or possibly even improve your design. It sure has been helpful for me to have your schematic for debugging the development kit. : )
Grant
For Rev 2 of the board negative inputs are used with SRB2 as the reference for channels 1-7 and possibly 8.
When enabling a channel the SRB2 bit is set and it seems you are turning on the bit in BIAS_SENSN. Shouldn't you be turning on the appropriate bit in BIAS_SENSP? It seems like the reference drive would attenuate which ever single signal or add noise to multiple signals that are active. It should be better to not set any bits in BIAS_SENSN, right?
Also, with the current firmware, how do you enable channel 8 without connecting SRB2 to INP? Wasn't the intent that channel 8 could be a bipolar input?
I am working with the ADS1299 development board as a proof of concept before making my circuit board. I will be building a Nintendo Wii controller for my dad who has ALS. Along with EOG/fEMG I will have a 9 axis IMU and pressure sensor for sip-puff.
I hope that I can either learn why you have configured it like this, or possibly even improve your design. It sure has been helpful for me to have your schematic for debugging the development kit. : )
Grant
Comments
Thanks for the outstanding questions! I'll address the easier one first...
Regarding 8N and 8P...man, you are very observant! While it is true that for the V2 board Joel brought out the N inputs instead of the P (at my request...possibly it was not my best idea), and while I also asked him to bring out the 8P input to pair to the 8N input, it is not required that one run them as a bipolar pair. The ADS1299 has an amazingly flexible (and complicated) input MUX that allows you to choose bipolar/ganged for every possible combination of inputs. So, the default software build available on OpenBCI's sourceforge/Github, actually ignored the 8P input and treats 1N-8N equally....all are referenced to SRB2.
If you're a hacker, you can choose to go in and modify the ADS1299Manager code and enable the bipolar use of channel 8. By default, though, it is ganged with SRB2 just like channels 1-7.
As for the harder question of Bias sense P vs Bians sens N, my interpretation of the Datasheet was that, to generate the bias signal, one should listen to the inputs that are actually connected to the human body, which would be 1N-8N. On the V2 board, the P connections are not connected to anything, so I thought that they should be ignored. In theory, that seemed to me to be the right call.
Subsequently, Joel has done testing to suggest that the P connections should be part of the bias generation, as you also suggest. Personally, I have not tested this to confirm Joel's findings either way. It is very possible that I missed something in my reading of the ADS1299 datasheet. If so, it would please me to find that y'all had found the better answer.
Chip