The 800-series radio uses between 10-30 mA (50-150 mw) during transmission depending on the signal strength. This is minuscule, especially considering that most of the time, the radio is not transmitting.
I havenât followed this thread to see what issues youâve encountered, so I canât really comment on that.
Your experience is not common, and based on your hub's engineering log you don't have a systematic problem. Can you try swapping the antennas around and see if Z-Wave performs better? If that doesn't help, I would be screening the logs (past logs) for hyperactive Z-Wave devices. Also running a full Z-Wave repair and resolving any issues with devices that fail may help your mesh too.
I'm not disputing that the requirements for the chips are low ... but if we assume they need 150mw, and at that moment there are only 130 available for whatever reason ... how do the chips react? If the chipset resets after system startup, does it re-acquire the masquerade ID for the zwave mesh from the database?, is that ID flashed into non-volatile memory on the chip when a restore is performed? These are things I don't know ... I can only troubleshoot it as the tiny black box that it is; power goes in, signals come out - I have to guess what happens in the middle.
To sum up the thread ... my zwave mesh has repeated issues that only exist on the C8, and so far there is no explanation or viable theory as to what the problem is. The engineering logs indicate there is nothing wrong despite nothing working ... random mesh failure (from the user perspective - despite associations still working), random mesh recoveries without intervention (sometimes it needs a reboot), and now mesh slowdowns are the latest shenanigans.
The mesh was migrated unchanged from a C5 and never had so much as a hiccup, and when I fail back to the C5, everything is rock solid again. This leads me to believe one of two things:
There is something inherently wrong with all C8's that the engineering logs can't report, or there is just something wrong with my C8. I hope the problem is unique to my setup/hardware somehow ... but until its identified, this C8 is only slightly more useful than training my cat for home automation tasks.
Sorry I also havenât read the whole thread, but what leads you to believe your hub doesnât have sufficient power available?
Why not just clear the z-wave radio and start from scratch (leave rules in place and run device swap on your existing z-wave devices to swap back) ? I mean in the time you've been running this thread you could have redone your mesh several times over.
As to the z-wave chips needing more power, then that would indicate that everyone running z-wave would have the exact same issues. That is not the case. Most times z-wave issues are caused by a rogue device. Pin pointing that is often difficult. It could also be some kind of radio corruption (which resetting the radio would correct). Most of us don't have issues. Hell I have some Ring outdoor contact sensors (v2) that are a 100 feet away through 2 interior walls, 2 courses of brick, and the yard...(on back gates). They are directly connected to the hub. The only time I have a z-wave issue is if I'm not paying attention to a battery warning lol.
Grasping at straws for a tangible explanation that actually works - because nothing else has panned out so far.
Two reasons ...
First - I'm in the middle of a datacenter migration that's requiring countless hours - so I don't have the spare time (or energy) to start this from scratch right now.
Second - IF I do that, and encounter the same problems once I'm finished, I've sacrificed the option to roll back to a stable environment without repeating the entire process. --- Edit If I had confidence in this solution, I'd sit on the C5 and once I actually had time again I'd just do it this way. But this thread was started to try understand and fix the problem (granted, I was hoping someone had an answer on-tap) ... if an answer is found it probably benefits more people than just me .. so I'm still looking for one as I have time.
Or you'll be farther ahead than you are now..
True ... and if I'd had any inkling that this was going to drag the way it has with no answers, different choices would have been made at the outset. hindsight is 20/20
Well like I said, most people don't have this problem with a migration. So I doubt it's an issue with the hub itself. I will say too that not every device was probably retested on the c8. There have been some problematic ones.
As a reader of as many topics here as I can, I would say the number of people experiencing a low internal power scenario is very close to zero, including yourself. It's a theoretical possibility, but even then, the power from the USB brick they supply (1 amp) would be converted to heat and you'd certainly notice that, right? If hub needs .777 amps (wag) and the usb block is delivering 1.00 then in order for power to be unavailable to the ZRadios, the extra .223 amps + enough extra to sag the power, would have to be consumed, aka heat. That's 25%-30% more power consumed than ordinary.
My two C-8s are reporting internal temps (in C) of:
Indeed, I know that I'm a fringe case. So, what makes me different?
If I get time this weekend, I'm going to try snoop the activity on the mesh (borrowed a z-stick yesterday) and see if the issues are internal to the hub or if they are reflected in mesh.
All I have right now is theories. And while don't actually believe a power problem would be pervasive, there is no reason that I couldn't simply have a bad piece of silicon in one of the SMDs on the board. In which case, a resolution here doesn't actually benefit anyone besides myself - beyond recognizing the symptoms.
Which doesn't mean you don't have one going bad. Trying to suss that out is as I said somewhat difficult.
Unless you have an inordinate amount of s0 devices this is unlikely. I mean the z00z 4-in-1's have the ability to do that but turning off s0 fixes it. Though to be fair a bad device can spam... "usually" that shows in the logs though.
As I recall, the C5 only supported s0, and s2 wasn't available until the 700 series chipset in the C7 ... so ALL of my devices would be joined with s0 regardless of if they supported s2. I'm pretty sure they'd still be using s0 after the migration to the C8. What are the troubles with lots of s0 on the C8? Perhaps I'm misunderstanding your comment...
S0 is an option with the C-5, as it is with the C-7 and C-8, Choosing to NOT use it with the C7 and C8 is challenging under certain circumstances. The "normal" join on the C-5 would use "S-none" (no security key at all - plain text, if you will.) And that's what's recommended around here.. use S0 only when forced. For the C-5 that would be for Locks and Garage Door Openers (barrier devices) which require a Security key.
No the majority of your stuff according to your screenshots are set to none. (which in my opinion is a good thing)
That's what I would have done ... it's been a while since I joined anything, so memory of the options is a bit fuzzy. In that case, nothing would be s0 because I would have never selected it - disregard my last comment.
The only reason I pay for Hub Protect on my C7 (besides actually wanting to give Hubitat some money) is for radio/devices migration. I voided the warranty by doing the antenna mod -- and I'm fine with that.
If the only way to fix weird zwave issues after migration even while having Hub Protect is to reset the radio, there's really no reason to purchase Hub Protect. I'm kind of worried about that now.
If the zwave network works fine w/ C5, the C8 migrated zwave network doesn't, and then the C5 works fine again, i feel like there's either something wrong with the migration tool or the new C8 hardware/software/drivers.
@kitt001 has this been fixed? or has the hubitat team offered a C7 or another C8 for you to test migration to ?
Download the Hubitat app
