Dongle is no longer recognized in GUI or Device Manager

jdk99jdk99
edited April 2017 in Cyton
Recently, the GUI and Device manager has ceased to recognize the Bluetooth dongle.  Prior to this, I had very few problems, occasionally I would have to restart the GUI because it failed to initialize, but other than that there were not any problems. All of a sudden, the COM Port for the dongle, or any COM port for that matter, was no longer there when trying to start the GUI.  Within Device Manager, it says that no COM Port is active, while the dongle is plugged in with the blue light steady and the switch on GPIO_6.  The board(32bit) is switched to PC, the blue light on it is steady, and the batteries are fresh.  I do not remember making any changes, other than just normal windows updates, but nothing that I would expect could cause something like this.  I have restated my computer multiple times, downloaded the Processing GUI again, and uninstalled/reinstalled the FTDI VCP drives, but the port still is not recognized by the GUI or in Device Manager, and I am having a very difficult time trying to figure out why.  The computer is running on windows 8 and I downloaded the 2.12.10 executable file from the FTDI chip website, if that information can help at all.  I tried to look through the forum, but was not able to find someone else with the same issue or anything that could solve mine, I am sorry if I missed one that did discuss a similar issue.  I would greatly appreciate any help anyone could provide.            
«1

Comments

  • wjcroftwjcroft Mount Shasta, CA
    @jdk99, hi. Thanks for your summary, it does sound like you've checked most of the common areas. The fact that you were working for some time is helpful. A couple other checks you can try:

    (1) does the dongle show up when plugged into another laptop? Either as a COM port on Windows or a /dev on Mac / Linux?

    (2) when first plugged in, the red light on the dongle flashes briefly, then the blue light comes on solid. You did say you had the blue light.

    As you probably found with the search box, there are some similar cases, but in some of these the blue LED is not coming on at all. If your COM port is not showing up, then reprogramming the dongle flash also seems out of reach. Joel @biomurph will likely comment here.

    http://openbci.com/forum/index.php?p=/discussion/554/usb-dongle-stopped-working-suddenly

    http://openbci.com/forum/index.php?p=/discussion/429/broken-dongle

    http://openbci.com/forum/index.php?p=/discussion/204/getting-started-dongle-drivers

    William
  • jdk99jdk99
    @wjcroft Thank you for the reply.  When the dongle is first plugged in, there is no red flash, it just goes straight to a solid blue light.  I downloaded the drivers on another laptop (which was running Windows 10) and plugged in the dongle, but I had the same results.  No red flash, just the solid blue led, and the COM port did not show up in the Device Manager.
  • biomurphbiomurph Brooklyn, NY
    @jdk99
    It does sound like the Dongle is getting power, because the blue LED is lit. However, the FTDI chip seems to be not communicating with the PC.
    Please email me joel-at-openbci-dot-com to discuss how to get the situation remedied.
  • brolyss360brolyss360 boston
    Hello,
    My dongle seems to have a similar issue, in that it was working yesterday, then tonight I went in to just double check the connections for tomorrow's testing and boom the dongle and the BCI board will not establish a connection. :(
    BCI chip is solid blue when turned on.

    On my mac I will get the blue LED it will blink three times then twice long(like 3 seconds) then it will go dim(barely visible, verrrry dim). (also no red LED activity at all) I cannot find anything on the forum describing similar behavior. 

    On my PC as I plug it in, the blue LED goes solid blue however the device manager says UNKNOWN USB Device (Device Descriptor Request Failed).

    Things I have tried:
    Mac-reinstalling ftdi drivers, attempted to connect on Open Bci gui and Neuromore
    PC-reinstalling ftdi drivers, attempted to connect on Open Bci gui and Neuromore
    Headset-new batteries
    headset-charged lipo battery

    Any suggestions?!
    PLEASE HELP!

  • brolyss360brolyss360 boston
    edited March 2016
    the dongle is not listed in my list of available serial ports :(

    $ ls /dev/tty.*

    /dev/tty.Bluetooth-Incoming-Port

  • wjcroftwjcroft Mount Shasta, CA
    @brolyss360, it does look like you have the same situation as @jdk99. And you've reloaded your drivers, checked your Device Manager and /dev. There is no way you could reflash the dongle if it is not showing up in your dev list. So... Joel @biomurph should be getting these notifications.

    But the lab is also very busy right now with the various engineering tasks from the last Kickstarter. I'd suggest following Joel's advice on the Jan 4 post above. Write to him directly at his email address and include contact at openbci.com. I'll also mention @Conor here. Most likely you'll need to do a swap of your dongle with the fullfillment center.

    Regards,

    William

  • brolyss360brolyss360 boston
    @wjcroft ;
    I called the brooklyn office this morning and Joel being the LEGEND he is, helped me sort something out. Thank you!
  • wjcroftwjcroft Mount Shasta, CA
    Wow, I'd be interested to hear if your dongle is working now, and what Joel suggested. Posting it here may help others.

    Thanks,

  • brolyss360brolyss360 boston
    oh well he's is sending me a new one, and I am sending in mine. He will be trouble shooting it to see if it is and user error fail or manufacture defect.

    I will follow up with him and ask what it was after all down the road. Or he might just post it up here if it is something fixable.
  • brolyss360brolyss360 boston
    Ok so i received the dongle today in the mail it arrived as a channel 4 and I'm not sure how to program it to channel 5 to match my headset
    But Im and not clear on the following:
    -what do i use to interact with the dongle through serial connetion on mac
    it seems I could use terminal or the arduino serial monitor which do you recommend?
    -what am i supposed to send to the dongle. is it this:
    // place this above the setup()
    #include <RFduinoGZLL.h> // using the Gazelle Stack
    device_t role = HOST; // This is the HOST

    void setup(){
    RFduinoGZLL.channel = 6; // use channels 2-25. 1 is same as 0 and 0-8 in normal GZLL library
    RFduinoGZLL.begin(role); // start the GZLL stack
    // more stuff here
    }
    And simply change the channel number to the channel of my board device? (channel 5 in my case)

    Please help

  • wjcroftwjcroft Mount Shasta, CA
    edited March 2016
    The documentation link you mentions says specifically to use Arduino IDE v1.5.8 (BETA). It will not work with the Terminal program. After loading the 32 bit host code into the IDE, change the RFduinoGZLL.channel assignment to your channel number. That's all you need to modify in the source. Then upload to the dongle with the IDE.

  • brolyss360brolyss360 boston
    I got it Thank you YAY everything is back to normaL!
  • brolyss360brolyss360 boston
    edited March 2016
    https://drive.google.com/file/d/0B7epH6R9TdXWbFRtdWp2THg4ZXM/view?usp=drive_web

    so now i have a new issue. 
    The BCI chip is giveing me crazy reading, for example the movement readings from the accelerometer  are going nuts and im sitting perefectly still. On top of that C6 siganl directly reacts when I move my head as if it was reading accelerometer. The top thre (Fp1, Fp2,C5) are all ralied fo rno reason. lastly when I blink O2 and O1 are the ones to react however Fp1 and Fp2 should be giving me the blink artifacts.

  • wjcroftwjcroft Mount Shasta, CA
    It's normal for the EEG signals to react strongly to head movement. They are not being mixed up with the accelerometer.

    To test your EEG wires and electrodes, use a blob of paste or alligator clips to connect together any or all of the 8 input pins, along with SRB2 (reference) and Bias. This should result in a signal close to 0 microvolts.

    What electrodes are you using?

  • brolyss360brolyss360 boston
    edited March 2016
    I am using 8 channels in the formation of the picture in the Neuromore screenshot. 

    So i dont alligator clips so I connected the cup electrodes, stacked them together and taped them stacked. I got the following screen shots.


    https://drive.google.com/open?id=0B7epH6R9TdXWbXJYWVlwMV9uN2c
    https://drive.google.com/open?id=0B7epH6R9TdXWNDk2UWlvcTVWNjg

    So i t did not go to 0 microvolts does this mean my board is bad?

    Also, I tried to look into re-programming the board however I kept running into the compiling errors related to plib.h in arduino. :(
    so i couldn't do it. I have been at this since 9 am, with my friend, we have tried so many things on both mac/pc.

  • wjcroftwjcroft Mount Shasta, CA
    I cannot see any of your images. You can only use the Rich Text icon for image (tv screen) with jpg or maybe png. I removed the icon enclosure to expose your URLs, but now cannot see even those because they are protected. You need to set the sharing permissions on your Google Drive files to all access for Everyone.

    This thread discusses workarounds for the plib.h,

    http://openbci.com/forum/index.php?p=/discussion/598/openbci-32bit-compile-error-plib-h-no-such-file-or-directory

    By "0 microvolts" I meant "close to 0", a few microvolts is ok. Is it possible you are in a high noise EMF area, such as near a wall wart or laptop power supply. Or directly over AC power cords or conduits?

  • brolyss360brolyss360 boston
    sorry about the images, I fixed the sharing permissions.
    I am in my room where I have always collected montages :( so no laptop chargers, no AC conduit and such.

    The solutions in that thread did not work for us, we tried them. 

  • wjcroftwjcroft Mount Shasta, CA
    It looks to me like you have a bad reference lead or electrode. Can you try swaping around your leads with the SRB2 to see if you can get a working one.

    If the reference lead is open (no connection), then the input channels will just pickup noise.

    Similarly it could be a bad Bias lead, but the SRB2 is more critical.

    Because all 8 channels are picking up exactly the same noise signals, suspect a common lead issue.
  • brolyss360brolyss360 boston
    you mean like use different electrode on srb2 and bais, or just reverse them?
    (have been using earlobe clips for both)
  • wjcroftwjcroft Mount Shasta, CA
    Do you have a DVM / ohmmeter there? Check the continuity of your suspect leads. Should be close to 0 ohms.

    Try to substitute known good leads for your SRB2 and Bias lines. Then do the "everything glommed together" test again.

  • brolyss360brolyss360 boston
    https://drive.google.com/file/d/0B7epH6R9TdXWdlZUaGZqZzM5UG8/view?usp=sharing

    So here I used electrode 1 as SRB2 and 6 as bias and turned off pins 2 and 7 since nothing was connected. 
    itried to get resistance readig ewteen each pin indvidually but I could not. I think it is because my multimeter sucks. 
    when I globbed the cup electrodes together and took a reading I got 0.01-3 ohms.
  • wjcroftwjcroft Mount Shasta, CA
    Wasnt your mainboard working just fine before the dongle went out? What has changed since then?

    How come this picture is so much different than the last ones? If the electrode leads were not suspect, then you'd assume the GUI graphs would look the same? Everything is railed here.

    Have you tried just going through the steps in the intro tutorial again? That did work for you previously, correct? In other words, see if you can get channels 2 or 7 to pickup alpha with SRB2 and Bias. Don't adjust any of the settings in the Chan Set tab, leave everything at default power up values. Turn off the unused channels. Swap around the SRB2 and Bias leads to see if a bad lead may be causing these rails.


    When you say you use "ear clips", what are you referring to. OpenBCI ships with the cup electrodes. Did you purchase ear clips elsewhere? Are they also gold to match the cups? You can't mix electrode metals. Frequently, mixing metal types can cause a battery effect or noise. 

    When you say, "So here I used electrode 1 as SRB2 and 6 as bias" I'm not sure we're speaking the same reference terms. Cup electrodes and leads do not have any numbers, they are just wires. When plugged into a channel pin, they become that channel. I'm asking you to swap out different electrode leads into the SRB2 pin position. (And also Bias.) To confirm that the railings are not caused by an open electrode lead. Use your ohmmeter to check EACH lead separately, while it is disconnected from the OpenBCI. You want to confirm each lead is close to 0 ohms. Sometimes wiggling or putting tension on the lead connector or cup can cause it to open the circuit, breaking the connection. In other words, a bad lead. Pitch it.


  • brolyss360brolyss360 boston
    https://drive.google.com/file/d/0B7epH6R9TdXWdzF6eEJYZjhmLXc/view?usp=sharing

    In the picture above I tested each of the electrodes off the headset and they all check out with ohms close to zero. 
    On the left is the (2x) Ear-clip electrode (for reference & ground) that came in my box with my Ultracortex (Mark III) EEG Headset Kit.

    On the bottom are the gold cup electrodes, and the ONLY time I used them was to plug in and 'glob' them together to test for zero microvolts from your suggestion above. I have never used these for montages.

    I have not mixed different metal electrodes whatsoever at any point.

    I will run some tests, first thing in the morning. Using different leads as SRB2 & bias, and one lead at a time for reading eeg to see if i get anything interesting.


  • wjcroftwjcroft Mount Shasta, CA
    Mixing the gold and silver chloride is usually not as bad as the other metals such as silver or tin. It's possible that your ear clip contact points (inside the clip) are somehow marginal. Try a test as shown in the tutorial with 3 leads of the gold cups and paste: O2, then SRB2 and Bias at the ears. Use tape to secure to ears. See if you get some alpha from that. All the pins you use on the board are the inner ones, closest to the board surface. Also try your glom test with the three gold cup leads and a ball of paste.

    So you had the Ultracortex working ok at one point before the dongle broke.


  • brolyss360brolyss360 boston
    I don't have paste for the gold cup electrodes, as i mentioned previously in the last comment i have never used the gold cup electrodes.
  • wjcroftwjcroft Mount Shasta, CA
    Sorry, I misunderstood. Your previous test with stacking the gold cup electrodes, I thought you used paste with that. I see now you just taped them together. Probably not optimal.

    Can you just try this. (I'm running out of suggestions. And I'm sure you're finding this frustrating as well.) In a small cup or container place a bit of water, maybe a couple tablespoons. Then add a half teaspoon of salt. Dissolve. Then place 3 gold cup electrodes into this solution, connect to channel 1, SRB2, and Bias. All pins closest to board. Then see if you get close to 0 microvolts. If you DON'T get 0, try swapping around the various leads in case there is a bad one. Be sure you are using SRB2, closest to board and not SRB1, farthest from board. Use OpenBCI_GUI to check your signals.

  • brolyss360brolyss360 boston
    https://drive.google.com/folderview?id=0B7epH6R9TdXWM0F1bG9jVFg2NWc&usp=sharing

    I had to move all my pics, my drive was getting messy.
    The latest pic was the readings i got with the gold cup electrodes in the solution. 
    I want to clarify what exactly are you refering to about when you use the word "Leads".
    also I have only ever used the pins closest to the board.
     
  • wjcroftwjcroft Mount Shasta, CA
    I recommend you contact Joel @biomurph and tell him all the tests we've tried. It looks like your mainboard is not working. Yet very suspicious that it worked before the dongle failed, what could have bonked your mainboard?

  • brolyss360brolyss360 boston
    edited March 2016
    What do you mean with the word leads???
    Im not ready to give up.

    is it possible to have bad pins on the board?

    because I just did two tests:
    1: I taped one of the smooth spring electrodes to my forehead and attached it to each pin on the board and tested for eye blinking artifacts and for alpha signals. while having all other inactive channels off. All of the pins did fine(showing blinks and alphs). 
    Then I did it again this time I left all channels on. It seems that Pins N1P and N3P (closest to the board) interfere a lot with all the other pins.

    2: I connected all the electrodes But N1P(the suspected bad pin) and shut off pins/channels N1P and N3P(also suspected however I left it connected and just shut it off). With this configuration the eeg signals seem normal/accurate.
    I repeated the following tests on both neuromore and OpenBCI GUI both share the same results.



    The images of the following tests
  • wjcroftwjcroft Mount Shasta, CA
    Well, certainly strange. I would mention this to Joel on your call or email.

    Leads = electrode leads consisting of a wire with a connector on one end and cup or other electrode at the other end. Lead is a common term used when referring to electronics and test equipment.

    You might check with a magnifying glass the surface of your board very carefully, especially around the ADS1299 chip. See if any metal flake particles might have lodged in between the solder connections there. Also look around other parts of the board for any conductive material.

    The header pin edge connectors rarely go bad since they are so large and have such large amount of solder connecting to the board.
Sign In or Register to comment.

© Vanilla Keystone Theme 2026