Announcement

Collapse
No announcement yet.

Does anyone have a solid working G3 system at this time?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    #31
    Originally posted by Vit View Post
    Not sure it is directly related but:

    attempted to update firmware on a couple of Zooz (500-series) devices.

    1. HS FW updater + Zooz 700-series stick => 3 attempts reached ~40% and failed with time-out error. Attempt #4 completed successfully but took more than 3 hours.
    2. HS FW updater + Zooz 500-series stick => completed in 10-15 mins.
    3. PC Controller 5 + Zooz 700-series stick => completed in ~10 mins.
    Yeah, this thread is about 700 series USB controllers using Silicon Labs chipsets and HS4. May want to start a new thread etc.

    Comment


      #32
      Everything is now working exactly as it should after reverting back to the HS 500 chip SmartStick. I'll be happy and patient to stick with it (pun intended) until this issue is finally resolved. The WAF has improved dramatically and I'm not looking to do anything to change that.

      Rich, thanks for your continued work on this issue.

      Originally posted by avpman View Post

      I rolled everything back to the 500 chip hardware and restored the devices to the controller stick. Individual commands are working. I'll be able to tell tonight and again in the morning when my multicast routines run if there is an improvement from the rollback. I'll report back tomorrow.

      Comment


        #33
        I've tried to roll back to my 500 series stick (Aeotec Stick with custom firmware), but restoring the backup to the stick seems to make all the nodes go missing. I'll see if I can use Silabs tools to roll back instead.

        Comment


          #34
          Yesterday I restored successfully to HS4/G2 and the interface freezing issues experienced with HS4/G3 have ceased. Question, when adding a new 700 series device with a SmartStick+ G2 (500 series) is there any functionality/feature loss during the inclusion versus using the SmartStick+ G3 (700 series)? For example, is there a benefit to rescanning a 700 series device with the G3 at a later date if it was originally included using a G2?

          Comment


            #35
            Originally posted by avpman View Post
            Everything is now working exactly as it should after reverting back to the HS 500 chip SmartStick. I'll be happy and patient to stick with it (pun intended) until this issue is finally resolved. The WAF has improved dramatically and I'm not looking to do anything to change that.

            Rich, thanks for your continued work on this issue.
            How, exactly, did you do the "roll-back" to the older SmartStick? I backed up my SS+ G3 last night, then tried to restore it to a older 500-series SS+, and it gave me an error that the destination interface was too old.

            Comment


              #36
              Originally posted by SteveW View Post

              How, exactly, did you do the "roll-back" to the older SmartStick? I backed up my SS+ G3 last night, then tried to restore it to a older 500-series SS+, and it gave me an error that the destination interface was too old.
              I'm using Windows 10 - HS4 Pro/G3 and Z-Wave plugin 3.0.7.0 with Z-Wave UI. Plugins -> Controller Management -> Z-Wave Interfaces -> Action -> Backup this Interface -> Save File Name -> replace G3 with G2 -> Use same procedure and just restore File Name to G2 after set new COM port etc. (make sure your not trying to restore an older backup with a different version of Z-Wave plugin)

              Comment


                #37
                Originally posted by STB400 View Post

                I'm using Windows 10 - HS4 Pro/G3 and Z-Wave plugin 3.0.7.0 with Z-Wave UI. Plugins -> Controller Management -> Z-Wave Interfaces -> Action -> Backup this Interface -> Save File Name -> replace G3 with G2 -> Use same procedure and just restore File Name to G2 after set new COM port etc. (make sure your not trying to restore an older backup with a different version of Z-Wave plugin)
                Thanks. Yes, that's what I did. I backed up the interface and the entire installation, replaced the dongle, then tried to restore it. Of course, it first creates a hurdle because it needs to assign a different "home" ID to the dongle, but it just fails with the "too old" error message.

                Comment


                  #38
                  Originally posted by rjh View Post
                  Note that the issue we are seeing is reproducible using Silicon Labs tools. They provide a "PC Controller" application that allows you to send and receive Z-Wave commands. I can make that software fail also and HS is not involved. So I am convinced that the problem is not due to our software. Now it's possible that our software may aggravate the issue. I have looked for dropped frames over the serial connection and have adjusted some timing, but nothing seems to help. It's possible Ezlo is using an older version of the serial API code which may not have the issue. We are using 7.15 and I have tried 7.14 and 7.16.

                  Also, some systems seem to work ok. One of our employees is using the G3 at home without any issues. So it's not clear exactly what is triggering it for some. I think its related to Z-Wave devices transmitting to the controller and its locking up the interface at times. This causes delays and sometimes a total lockup.

                  If you need to roll back to the G2, just do a backup with the G3 and you should be able to restore to the G2. If you did not add any devices since you switched to the G3, then just replace the G3 with the G2 and set the correct COM port in the Z-Wave settings. You don't need to do a backup/restore. After switching, restart the plugin so it syncs the nodes with HS.
                  Just wondering if Silicon Labs has gotten back with you and of there were any updates you can share.

                  Thanks in advance.

                  Comment


                    #39
                    The developers are in Copenhagen and they have been on vacation, so that is holding things up. I am working on a simpler test so they don't need to find any third party hardware. I am also experimenting with our driver code, maybe I can make a change to at least keep the issue from happening. Will post when I have any info.

                    Originally posted by jjason View Post

                    Just wondering if Silicon Labs has gotten back with you and of there were any updates you can share.

                    Thanks in advance.
                    website | buy now | support | youtube

                    Comment


                      #40
                      Any update. Still using my G2 stick until the G3 get straitened out.

                      Comment


                        #41
                        The main problems I had were related to attempts to migrate back and forth between the two generations of controllers. I now have a stable system running with the G3 SmartStick+ on a Raspberry Pi 4 (with Raspberry Pi OS and HS4 Standard).

                        Comment


                          #42
                          Still working the Silicon Labs. They want me to give them a test case using only their tools and I cannot make it fail with just their tools, they need some real devices. So still battling. It really looks like the reports from energy reporting devices is causing an issue. As soon as I have any information I will post.
                          website | buy now | support | youtube

                          Comment


                            #43
                            Thanks for the update. let me know if there's a better thread to track this issue. also let me know if I can help (try to reproduce using their tool?). i bought a bunch of 700 series stuff i haven't installed yet. i'm just waiting for the stick upgrade before adding a bunch of stuff.

                            Comment


                              #44
                              I know this is taking a while, but we are still working on it. Silicon Labs wants a procedure to make this fail and I just don't have one. The issue is totally random but seems to have something to do with receiving a command (like a central scene) and then immediately sending a Z-Wave command (like on/off) to another device. Adding a delay in this process makes the issue less likely to happen, but it still happens. We are looking at some changes to our packet handling.
                              website | buy now | support | youtube

                              Comment

                              Working...
                              X