Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Tuicemen]

Here is a weird experience I've encountered while playing with the cm19 and the PXH.

Originally the cm19a while worked with the PXH, it was slow to control things.
I added a USB addon board reported here: http://forums.x10.com/index.php?topic=30753.msg179923#msg179923 manly to allow a Zwave dongle.
This caused my cm19a speed at executing things to improved dramatically.
Even Alexa control is substantially improved though part of that may be my move to the HomeGenieEchoBridge in HG.
My Alexa OK response is 99% and x10 control is 100% even with out the OK response from Alexa

You would think the ZeroW would have had a hard time running this add on board but it doesn't.

[petera]

Here is a weird experience I've encountered while playing with the cm19 and the PXH.

Originally the cm19a while worked with the PXH, it was slow to control things.
I added a USB addon board reported here: http://forums.x10.com/index.php?topic=30753.msg179923#msg179923 manly to allow a Zwave dongle.
This caused my cm19a speed at executing things to improved dramatically.
Even Alexa control is substantially improved though part of that may be my move to the HomeGenieEchoBridge in HG.
My Alexa OK response is 99% and x10 control is 100% even with out the OK response from Alexa

You would think the ZeroW would have had a hard time running this add on board but it doesn't.

The assumption here is that the Z Wave antenna may somehow be propagating the signal to and from your CM19. Clearly the bands don't collide 310 v 908 MHz but you've two antennae in close proximity attached to the same unit which may be producing some form of dipole effect. I'll leave that one to the anoraks to work out

[Tuicemen]

The assumption here is that the Z Wave antenna may somehow be propagating the signal to and from your CM19. Clearly the bands don't collide 310 v 908 MHz but you've two antennae in close proximity attached to the same unit which may be producing some form of dipole effect. I'll leave that one to the anoraks to work out

Although that is a possibility, I highly doubt it as they are not very close to each other.
I'll remove the Zwave Dongle and run a few tests to see if things change as I can live with out Zwave control for a bit.

[Tuicemen]

I got thinking more about RF signals being affected by things.
All signals used here are RF and RF signals can be beamed with passive antennas which increases their strength in specific directions.

The wi-fi antenna on the ZeroW is mounted on the board and experiments with external ones show little improvement in performance. However these experiments I've read are based using simple small Wi-fi antennas and not beaming ones.
This addon board sits over the onboard Zero W Wi-Fi antenna which I suspect acts to reflect the Wi-fi signal a bit more in one direction then the others. This would increase its strength and range in that direction.

Removing the Zwave Dongle had no affect in performance however flipping my Zero board over did
  it may be worth experimenting a bit with Wi-Fi signal beaming for users having difficulty with the Pi Zero W.
 

 

[Brian H]

I saw a USB add on board for the Pi Zero and Zero W.
Mounted on the back side. With nylon spacers and screws. Pressure contacts connected to some of the test pads on the back.
They updated it to add a ferrite ring. You put over one of the pressure contacts. Seems the Wi-Fi RF was getting into the board through one of the signals connection contact.

[Tuicemen]

I saw a USB add on board for the Pi Zero and Zero W.
Mounted on the back side. With nylon spacers and screws. Pressure contacts connected to some of the test pads on the back.
They updated it to add a ferrite ring. You put over one of the pressure contacts. Seems the Wi-Fi RF was getting into the board through one of the signals connection contact.

My addon board is designed specifically for the Zero W though I imagine it would work with the Zero as well.
It has metal spacers and screws, only 3 full sized USB ports + 1 UART connection, and no ferrite ring though it does use the same test pads.

[Tuicemen]

Genes latest build 1.2-stable.35 loads so much faster then previous builds (for me) that my zero W with cm19 is a pleasure to use. The update also improved the cm15 operation on the ZeroW. So I suspect all Pis will benifit.  

[petera]

Genes latest build 1.2-stable.35 loads so much faster then previous builds (for me) that my zero W with cm19 is a pleasure to use. The update also improved the cm15 operation on the ZeroW. So I suspect all Pis will benifit.

Such a shame I'm not seeing release notes or change logs on Genes releases of late. Last release notes were v33

Just noticed a change in the MQTT library by sifting through the code. You now report an improvement in CM19 operation and load times for the API. Curious to know how that came about

[alzy]

Is the best way to update HG in image 1.1 to HG 1.2 stable.35 with the HG update capability?

[bkenobi]

I am skeptical that the upgrade system will pick up the new version.  In the past when Gene updated major version number it didn't detect.  In either case, I personally would recommend creating a backup of your setup and then install the new version from command line prior to restoring your settings.  The only issue with that approach may be that HGEB may not work unless you redo it (whatever that entails).

[alzy]

Thanks!

[Tuicemen]

Gene fixed the update from inside HG. However petera and some others noticed a issue with the security interface which require disabling and reconfiguring I believe re-starting fixed it.

[Tuicemen]

Genes latest build 1.2-stable.35 loads so much faster then previous builds (for me) that my zero W with cm19 is a pleasure to use. The update also improved the cm15 operation on the ZeroW. So I suspect all Pis will benifit.

Such a shame I'm not seeing release notes or change logs on Genes releases of late. Last release notes were v33

Just noticed a change in the MQTT library by sifting through the code. You now report an improvement in CM19 operation and load times for the API. Curious to know how that came about

Maybe some disabling of some auto loading things as some have reported the security widget not working after updating and now you see a change in the MQTT. In any case I'm enjoying the page load speed improvements and now enjoy loading it and configuring from my phone.

[petera]

Genes latest build 1.2-stable.35 loads so much faster then previous builds (for me) that my zero W with cm19 is a pleasure to use. The update also improved the cm15 operation on the ZeroW. So I suspect all Pis will benifit.

Such a shame I'm not seeing release notes or change logs on Genes releases of late. Last release notes were v33

Just noticed a change in the MQTT library by sifting through the code. You now report an improvement in CM19 operation and load times for the API. Curious to know how that came about

Maybe some disabling of some auto loading things as some have reported the security widget not working after updating and now you see a change in the MQTT. In any case I'm enjoying the page load speed improvements and now enjoy loading it and configuring from my phone.

You do realise if the inbuilt HG updater fails the only way you can achieve an update is to do it the manual way. Of course should you do that and restore you HG config backup your setup is minus the HGEB.

For anyone who was struggling to install it in the first place you will have to start all over again including your Alexa discovery process.

There may be a bit of politics at play here (the HGEB was developed in the forked veresion of HG) but I would have assumed Gene would have included what appears to be a fairly popular and much demanded MIG Interface in the stock HG install so the HG restore process would pick it up.

Might be worth posting a GitHub request to have this included or for the backup/restore support for user defined MIGS. I've other requests already awaiting approval and I don't want to be hogging the GitHub.

The problem with GitHub forks is fragmentation. One version has support for CM19, the other doesn't and vice versa with the HGEB. Now you can see what problems lie ahead.

[alzy]

I update to 1.2 stable .35 using the update capability in HG. Update seemed to execute ok. I still see the cm19 double click issue after the update.

BTW, I also installed the windows version of HG on my win10 laptop and it performs flawlessly with the same cm19.