I'm trying to figure out what speed my devices are communicating with the hub. I have been using the Z-Wave Details page, and know there is the kbps value, but I'm wondering if that is an accurate reflection of how fast the switch and hub are talking to each other. Is it, or is that the speed of the last device in the chain? I've been recording the kbps speed and calculating the # of hops each device takes each day for weeks now to see if the mesh is improving on speed/hops now that I've replaced the C-7 with a C-8 and also to see if there is any correlation between the speed and # of hops. (BTW, I'm doing that by copying the Z-Wave details page each day and pasting it into an Excel spreadsheet where I let macros do all the massaging/calculations, so it only takes a minute or less each day).
I was thinking of using the avg. ms number instead of the kbps reported, but the problem with that is that a significant # of devices are not reporting that back at all. I see that none of my Inovelli Black switches are reporting any ms time. Likewise, 2 out of 3 of my Fibaro KeyFobs have no ms time (but one does), and 4 of my 24 Inovelli Red Dimmers do not show an ms time. Neither of my 2 Aeotec 7 range extenders have an ms time nor does my Aeotec NanoMote Quad Button (But my 2 Aeotec plugs do have an ms time).
Would the ms time be a more accurate measure? And if so, any idea why these devices aren't reporting an ms time and/or how to get them too?
Any other suggestions on how to accurately track a device's "speed"?
@Stu_The_K The speed it's talking to the hub is reflected in the last column in kbps. It will either be 9.6, 40, or 100k. It is very hard to account for varied latency though.
Thanks. That's what I've been using. While I've been recording the # of devices at 100, 40 and 9.6, as well as the # that were faster, slower or the same as my C-7 since April, I only started calculating my avg. speed and avg. hops since June 20. Plus, on July 1st I made 2 changes: 1) I moved one of my two Aeotec 7 Range Extenders. (No device was really going through either one before and now no devices are going through either one. So I don't think that really mattered); 2) I had the TP-Link satellite wi-fi device I had my C-7 and C-8 plugged into laying on its side to help hide it, but on July 1 st I stood it upright. Since July 1st I've seen a noticeable uptick in the # of devices at moving up from 9.6 to 40 or 100 and 40s moving to 100. Plus, the hop average, which fluctuates between the 1.6 and 1.4 range since I've been tracking it, is significantly down from the 2.2 average I had on the last day I was using the C-7. I also noticed that since standing the TP-Link upright the # of devices that are faster now with the C-8 is significantly higher than the # of devices that are now slower on the C-8. I'm not going to plug the C-7 back in, so I won't be able to say if the C-8 is better than the C-7 as I don't want to mess with how well it is running right now.
I've also started tracking the correlation between the avg. speed and avg. hops. Again, I started that on June 20 and will report back in at the end of July when I have some more data.
I'm not sure what all the measuring and testing is accomplishing. So many variables come into play. For myself any way I look at it this way. My hub is almost dead center of the house. 98% of my z-wave stuff directly connects to the hub, the rest due to their range may hop through a repeater. Zigbee mesh is pretty normal. Most mains devices directly connect to the hub and battery sensors connect mostly to the mains repeaters. All in all my setup is instantaneous (and that includes my 60 odd lutron devices). I guess what I'm saying is that it all works fine. I'm not gonna sweat the small things... It's not like it's an ethernet network. As long as it's generally stable is all I care about,.
This all started out trying to see if the C-8 was giving me any better connections than the C-7 because I wasn't able to update to the latest (last) firmware Inovelli did for their Red Dimmers as many failed or only 1/2 of the 2 parts would install and I wanted to know (in the words of the Marathon Man) "is it safe?" to try and update the firmware now.
Plus, some devices seem to react slowly (2-3 seconds) with a laborious mesh path.
I also wanted to see if the C-8 could help fight/end the dreaded 30+ second wait that suddenly happens with my devices.
And its kind of fun for me to write the code to do all of this.
Since it seems to be humming along now that I stood the TP-Link satellite up I'll probably stop at the end of the month as I'll then have six weeks of speed and hop data and I want to answer, at least for my setup, is there any correlation between hops and speed (at least for a (hopefully) stable mesh, particularly as many people report that were seeing more direct connections with the C-8 (vs. reporting they were seeing more 100 speeds, for example).
So I have now recorded data for myZ-Wave devices 69 times. I have data for the average speed (100 * # devices at 100) + (40 * # of devices at 40) + (9.6 * # of devices at 9.6)/# of devices (which in my case happens to be 49).
I also calculated the average # of hops per device using a similar type of calculation. I then calculated the average speed per hop.
I also calculated the correlation between average speed and average hops. While it naturally varies a bit from day to day the final # is -0.51426077. The range for a correlation is -1 to +1 where -1 is absolutely no correlation while +1 is a 100% correlation. My -.51 means there is a moderate non-correlation (i.e. there is moderate evidence of no correlation).
There was one thing I did that made a tremendous difference. My C-8 is plugged into a TP-Link satellite router. As they are on top of a wall unit I had laid my TP-Link on its side so it wouldn't be visible. On the 9th day of my data point I stood it upright. That instantly brought a big jump up in my data speeds and reduced my # of hops dramatically. I know Hubitat suggests having the hub about a foot away from the router, but I never saw anything about the orientation of the router (and I had assumed it was spherical to get the best all-around coverage). One of the reasons I went with Z-Wave vs. Zigbee way back when was I had read that Zigbee was at 2.4, the same as one of the WiFi frequencies. So I'm very surprised that the orientation of the TP-Link satellite had such a profound effect on it.
In any event, here are a couple of charts showing my results. Because the # of hops is between 1-4 and the speeds are 9.6-100 the # of hops was so small that the differences were very hard to make out. So I made another chart with just the # of hops and overlaid it on the other chart and made it transparent so you could see those fluctuations better. (BTW, there was 1 day where 1 device was at 4 hops, but I didn't include 4 hops in the 3rd chart, but it was used in calculating the average # of hops per device and the speed/hops)
One final note: The #'s for the C-8 in both average speed and average hops were worse than the C-7 when the TP-Link was on its side. But I did not hook up C-7 back up once I stood the TP-Link upright so I can't say how it would have preformed with the TP-Link standing upright. As such, I can't say if the C-8 is "better" than the C-7 or not.
That's very strange considering z-wave doesn't use wifi channels. I wonder if simply moving the hub father away from your router would have effected the same change (less rf interference) Zigbee would certainly be affected.
I only have a couple of Zigbee devices and I can't really say I noticed any difference on how fast they respond. Am I correct that there is no place to see Zigbee speeds in Hubitat?
I'm not sure where the "lqi" is. Is that on the Zigbee page? Also, I'm not sure what the RSSI is and how it relates to the speed a device is communicating with the hub. Here is a screenshot of 4 devices that all say that are at 100 kbs, but have different RSSI numbers. Sorry to be such a noob, but could you please explain to me what this means.
BTW, I find it strange that the 1st one, 17, goes through 24, and 24 goes through 17. I would have thought that one of those 2 would go straight to the hub and the other through it. Not both using the other. But, is that just the Alice in Wonderland nature of Z-Wave (curiouser and curiouser)?
RSSI, or “Received Signal Strength Indicator,” is a measurement of how well your device can hear a signal from an access point or router. It's a value that is useful for determining if you have enough signal to get a good wireless connection.
LQI is a metric of the link quality of the received signal
An LQI value above 80 is well optimized. If it is between 60 – 80, it is optimized but needs more improvement. If it is below 60 then it needs a lot of improvement. Each section in the LQI has a clear description of the recommended standard value and the errors in the listing that requires to be fixed.
I know about the Z-Wave details page. That is where I was getting all the data from. But I still don't know where the lqi is listed. On my Zigbee details page it only shows this:
You're not actually typing "yourhubip" are you? That's meant for the ip address of your hub. Put it in your web browser. For instance, this is what it looks like for me... http://192.168.2.122/hub/zigbee/getChildAndRouteInfo , you simply use the ip address of your own hub.
