Is my Cyton broken? -> incomplete mainboard RFduino firmware upgrade
Dear all,
I would like to begin with the appropriate newbie disclaimer: I very much depend on step-by-step instructions (like the excellent "Getting Started" section) and feel a bit lost when I encounter a problem on my own. Expect stupid questions from me! Thank you!
So, I bought a Cyton two years ago for an EEG-project, did a basic set up back then and, as far as I remember, everything went fine. Only that the project wasn't our highest priority, was post-poned several times and we finally rebooted some weeks ago; meanwhile, Cyton was not in use.
Now, I downloaded the newest version of the OpenBCI software and updated firmwares of the dongle and the Cyton board to the current versions.
Again, I walked through the Getting Started section, but this time, I have an oddly strong constant noise on all electrodes. The following image is with no electrodes attached:
So channel 1 shows a constant 1 Hz-signal and all the other channels are just caotic. Note that scaling is set to 'auto' because signal strength goes far beyond the available scales. Impedances are also in the 10s of thousands.
I nevertheless proceeded to set up an arrangement for polysomnography:
(SRB2: right mastoid, N1P: outer corner of the right eye, N2P: outer corner of the left eye, N3P: left frontal, N4P left parietal, N5P: left occipital, N6P: left chin, N6P2: clavicule, N7P: empty, N8P: chest; BIAS: empty, AGND: left mastoid).
But besides getting a decent EKG on chanel 8, there was little change. And the EKG might also disappear with any handling of software or hardware. But overall, detaching or reattaching electrodes made no notable difference, be it ground, reference or any of the others. I think I used a sufficient amount of contact gel, but that didn't improve impedances. Channel 1 looks a bit more organic, but that's just by filtering, and it is not synchronous to heart beat, even though it might appear so on first glance.
One last curiosity I want to point out is that when I silence a channel, it will actually be the subsequent channel that goes flat.
Here, I turned off channel 5 (yellow), as you can see by the greyed number 5; it is however channel 6 that goes silent; it might eventually come back.
As for trouble-shooting, I didn't know what to do beyond checking the electrodes themselves (which seem fine) and reinstalling the software and re-uploading the firmware. So I'm stuck - and would be happy about any suggestions!
Thanks in advance,
Derek
I would like to begin with the appropriate newbie disclaimer: I very much depend on step-by-step instructions (like the excellent "Getting Started" section) and feel a bit lost when I encounter a problem on my own. Expect stupid questions from me! Thank you!
So, I bought a Cyton two years ago for an EEG-project, did a basic set up back then and, as far as I remember, everything went fine. Only that the project wasn't our highest priority, was post-poned several times and we finally rebooted some weeks ago; meanwhile, Cyton was not in use.
Now, I downloaded the newest version of the OpenBCI software and updated firmwares of the dongle and the Cyton board to the current versions.
Again, I walked through the Getting Started section, but this time, I have an oddly strong constant noise on all electrodes. The following image is with no electrodes attached:
So channel 1 shows a constant 1 Hz-signal and all the other channels are just caotic. Note that scaling is set to 'auto' because signal strength goes far beyond the available scales. Impedances are also in the 10s of thousands.
I nevertheless proceeded to set up an arrangement for polysomnography:
(SRB2: right mastoid, N1P: outer corner of the right eye, N2P: outer corner of the left eye, N3P: left frontal, N4P left parietal, N5P: left occipital, N6P: left chin, N6P2: clavicule, N7P: empty, N8P: chest; BIAS: empty, AGND: left mastoid).
But besides getting a decent EKG on chanel 8, there was little change. And the EKG might also disappear with any handling of software or hardware. But overall, detaching or reattaching electrodes made no notable difference, be it ground, reference or any of the others. I think I used a sufficient amount of contact gel, but that didn't improve impedances. Channel 1 looks a bit more organic, but that's just by filtering, and it is not synchronous to heart beat, even though it might appear so on first glance.
One last curiosity I want to point out is that when I silence a channel, it will actually be the subsequent channel that goes flat.
Here, I turned off channel 5 (yellow), as you can see by the greyed number 5; it is however channel 6 that goes silent; it might eventually come back.
As for trouble-shooting, I didn't know what to do beyond checking the electrodes themselves (which seem fine) and reinstalling the software and re-uploading the firmware. So I'm stuck - and would be happy about any suggestions!
Thanks in advance,
Derek
Comments
Thanks for the rapid reply!
That's weird, I can see the images in my thread. But I'll post links here:
image one: no electrode connected
https://www.dropbox.com/s/hmod1bpofzyi8t6/No_electrodes.png?dl=0
image two: electrodes connected, decent ECG-signal
https://www.dropbox.com/s/wx79nwel5hbqldg/connected.png?dl=0
image three: shutting down channel 5 will silence channel 6
https://www.dropbox.com/s/9t85fi7zqkdl01b/Silent_6.png?dl=0
I failed to mention in my original post that I indeed adjusted gain to account for the differences in ECG, EMG and EEG signal strength. It didn't make much of a difference, though. The noise on all the channels is in the 100,000 µV range (as were the impedances), which doesn't make a lot of sense. And that couldn't be fixed by loads of contact gel.
Thanks for pointing out that I have to connect BIAS instead of AGND for ground, that was definitely a mistake I made - I did however switch back and forth to BIAS at some point when I was just randomly trying out different options. And it was all the same, just high amplitude noise, no mattter if I connected or disconnected the ground, reference or any other electrode. Even the moment of (dis)connecting was indistinguishable in the time series of the respective electrode.
As you can see in image two, I did get an ECG signal, but it wasn't reliably repreducible. Any thoughts in particular on the 1-Hz-signal that I always have on channel 1? Note how it phase shifts when turning off channel 5 in image three.
Thanks again for any feedback!
Cheers,
Derek
Your intuition seems to be correct. Even though I was pretty sure that my firmware update had also included the dongle, I was apparently mistaken - I found no trace of previously having downloaded the pertinent files for the dongle.
So now I updated the HOST-firmware but I am missing the tools to update the DEVICE-firmware to Cyton. Correct me if I'm wrong, but this is the firmware for the bluetooth radio and not the chipset of the Cyton board itself, which I had updated using the ChipKit Core - right?
Unfortunately, now I have no means of connecting to the Cyton (neither Arduino nor the OpenBCI GUI will find it) and I guess this won't be possible until I have the tools necessary for wiring to Cyton and updating the DEVICE-firmware - right? There is some small success here that the OpenBCI GUI communicates correctly with the dongle ("Success: System up") while previously it would post an error message (something along the line of "Error:check connection/port"), even though it would eventually connect to Cyton and now doesn't.
Sorry about not being precise in my description, but the first version is correct: the image "connected.png" is with all channels in use except for CH7.
Cyton did work before the upgrade to V3, even though we only did brief test-runs.
BTW: do I understand correctly that while flashing the dongle, channel definition must read
radio.begin(OPENBCI_MODE_DEVICE,2);
if I am using channel 2, which is the channel my Cyton is tuned to? It says '20' in the original code and I take that to be just a placeholder.
Thank you very much for your time and effort!
Best,
Derek
I managed to downgrade back to HOST 1.x.x and now the dongle and Cyton reestablished communication (with all the artifacts motivating my original post). But the connection is fine for Arduino and programming the board itself.
If I were to buy a Wifi Shield and my artifacts do arise from an incompatibility of the chipset firmware and the bluetooth firmware - is it likely that the signal will be just fine once I connect via Wifi?
Thanks!
Derek
So I'll start all over and let you know how it went.
Thank you so much!
Derek
Thanks again!
Well, I could get the RFduino USB shield for some 30 bucks on the internet (2-3 weeks delivery) but unlike in the tutorial, it comes without capacitor, so there's not much of an advantage here. Alternatively, I can walk over to the nearby electrical retail, buy a breadboard, jumper cables, a capacitor, and female header pins for virtually no money at all, dig out the soldering iron that I inherited, take a youtube course on soldering and then go for it.
What could possibly go wrong..? (I will elaborate on that question in a later post) ;-)
Regards,
Derek
I am happy to report that it worked! I got the breadboard, cables, etc., soldered the header pins to the dongle, built the circuitry according to the instructions and was able to flash the RFduino firmware. And now i have a very beautiful EEG signal.
This is so elating: I feel like having evolved from being a rookie without the foggiest notion of what he's doing to being a rookie without the foggiest notion of what he's doing but with a stroke of luck.
I thank you sincerely for your patient and thoughtful reflections on my problems that eventually set everything on the right track! I would have been stuck!
Best,
Derek
(Oh, BTW, this post hasn't really been about polysomnography; if anyone happens to have a question about clinical polysomnography, contact me and maybe I can provide some help)