Z-Wave Locks & Battery Issues

[Person]

Few things I noticed
1) I have yale door lock w/ lever (not deadbolt). Installed mid Jan this year. Battery @60% today. You don't have to unlock to open from inside so fewer (3-4) z-wave/keypad locks/unlocks per day. I guess w/ deadbolt you always need to unlock to leave so this adds more unlocks. I bought it in Jan this year as well. So I seem to get much better battery life. I do poll the lock every time I arm alarm system (leaving, staying, going to bed) so about 2-3 times a day.
2) Thanks for the info on power transfer hinge. Didn't know such thing existed!
3) For powering lock, I would not install battery at all. Bring 5V from a small UPS supply and you have both 5Vand backup. Here are examples of UPS I found (I don't own any of them so this is not recommendation but they should work):
http://www.somroli.com/store/product_info.php?products_id=56
http://www.acedepot.com/5v-mini-ups.html
http://www.amazon.com/Fingertec-Mini-Uniterrupted-Power-Supply/dp/B005UO9YJO
http://www.parts-express.com/pe/showdetl.cfm?partnumber=320-140
4) Here is nice way of putting in fake batteries to connect power. You can probaly just wrap wires/push them behind contacts but this will be more secure
http://www.instructables.com/id/Make-a-battery-substitute-power-connector/#intro

[gearyt]

power transfer hinges...

http://www.commandaccess.com/product...fer%20Products

several company s make them, select hinge makes full length ones that I use

[NeverDie]

3) For powering lock, I would not install battery at all. Bring 5V from a small UPS supply and you have both 5Vand backup. Here are examples of UPS I found (I don't own any of them so this is not recommendation but they should work):
http://www.somroli.com/store/product_info.php?products_id=56
http://www.acedepot.com/5v-mini-ups.html
http://www.amazon.com/Fingertec-Mini-Uniterrupted-Power-Supply/dp/B005UO9YJO

Thanks for posting these suggestions. I like their relatively small form factor. Not inexpensive, but these DC output UPS's should perform better than commonly available UPS's that output 110vac, which would then have to be down-converted to the low DC voltage (http://board.homeseer.com/showthread.php?t=162367).

[VirtualPanther]

Found it. Names like "electric hinge" or "power transfer hinge" will bring it up.

Now just need to get the power from over at the hinges to where the deadbolt is.... Just drill a bore hole using a really long auger bit (so as to keep it straight)?

Anyhow, for someone who is determined (virtualpanther?), it sounds like a feasible project.

From what I've read, float charging only works on lead acid batteries. I'd be more than happy if someone shows I'm wrong about that, because float charging is obviously simpler. I'm pretty sure other types of rechargeables like Lithium Ion or Nmh require a charger circuit that cuts out, but those should be readily available.

I called Assa Abloy (a.k.a. Yale) and asked their tech support about precise power requirements for this lock. They refused to answer, saying they are not allowed to provide advice on unauthorized modifications. I didn't even tell him why I was asking that!

Running a wire would not be an issue for me. I'd probably use a typical commercial armored door cord, designed to bridge the gap:
http://www.smarthome.com/77441/SECO-...or-Cord/p.aspx
http://www.amazon.com/Schlage-788-12.../dp/B004HN48NQ

On the door itlef, to the lock, I'd use a surface-mounted wire chase and then caulk and repaint the whole door. Not fancy, I know. But I had to do this for a different reason once (sensor mounting) and it looked surprisingly OK. Definitely easier that trying to drill and chase wire through a typical residential hollow metal door.

I am so fed up with batteries, I don't really care how much it would cost. I just haven't found a very reliable way of ensuring that the power supply used will not under- or over- power the electronics in the lock. Also, I have not come up with an esthetically-pleasing way of plugging said power supply into lock's battery compartment in a do-it-once-and-forget-it way.

[NeverDie]

Hopefully others will chime in on this, but once you determine the voltage and current requirements, I expect your best option would be to use a *regulated* power supply that can deliver more than the required current. In that case, the voltage will be fixed and your device will pull only as much current as it needs. A ballpark estimate would be (# batteries)*1.5 for volts, and a 1 to 2 amp power supply would probably cover maximum current surge when the deadbolt turns. If you can't find the power input leads into the mechanism to test delivered voltage directly, you could use voltmeter probes to check for maximum voltage found in the battery pack. That would be a precaution in case not all the batteries are connected in series.

There's a thread on this forum somewhere that shows how some people here are replacing batteries with one or two dowel rods with power supply leads screwed on the ends. Using that trick it would then also be easy to verify peak current required.

[etc6849]

You need a AC/DC current clamp such as this one:
http://www.fluke.com/fluke/usen/acce....htm?PID=56297

Along with a digital oscilloscope used in single shot (aka single mode) to capture the current transient. Once you have that I'd up size the peak current by 20% to be safe.

So far, all I did was disable the features mentioned in my previous post. The good news is the lithium batteries I put in on 10/20 from samsclub (12 pack Energizer Ultimate Lithium batteries for $19.98) are still at 100% today when I polled the lock (I'm using the VRC0P +3 with a custom driver and not the Homeseer module)!

>N2SS128,2
<E000
<N002:152,128,238,218,076,121,222,138,153,171
<E000
<X000
<N002:152,064
<n002:000,128,003,100

I'll update this thread in a month or so. Feel free to PM me if I forget. Today I added a lock count property to count the number of actuations. I wish I'd have added this before when I put the batteries in, but oh well...

[gearyt]

my major battery drain = polling freq

[etc6849]

You shouldn't need to use polling if the lock is associated with your controller properly (I don't use Homeseer, but in general this is true if the z-wave device supports the "Association" class and the controller support secure commands).

I don't use polling and get near instant updates (in less than 1 second) when the lock is manually actuated. You are right, frequent polling would suck the battery life away.

my major battery drain = polling freq

[VirtualPanther]

You need a AC/DC current clamp such as this one:
http://www.fluke.com/fluke/usen/acce....htm?PID=56297

Along with a digital oscilloscope used in single shot (aka single mode) to capture the current transient. Once you have that I'd up size the peak current by 20% to be safe.

So far, all I did was disable the features mentioned in my previous post. The good news is the lithium batteries I put in on 10/20 from samsclub (12 pack Energizer Ultimate Lithium batteries for $19.98) are still at 100% today when I polled the lock (I'm using the VRC0P +3 with a custom driver and not the Homeseer module)!

>N2SS128,2
<E000
<N002:152,128,238,218,076,121,222,138,153,171
<E000
<X000
<N002:152,064
<n002:000,128,003,100

I'll update this thread in a month or so. Feel free to PM me if I forget. Today I added a lock count property to count the number of actuations. I wish I'd have added this before when I put the batteries in, but oh well...

I am glad it's working for you. I am not a big fan of energy-saving solutions offered here. To me, using a device to its fullest potential is the whole point of getting it in the first place. It's like getting an iPhone and then disabling WiFi, Bluetooth, e-mail push, and background sync -- all to save battery. The point is, I want a solution, using which I can set up polling, auto-lock, tons of notifications, etc. -- all without having to think about batteries being depleted overnight. It is quite apparent that these locks are not well suited not to be hardwired. Between touch screen, ability to auto-lock, security protocols in wireless communication -- all amount to large battery drain, even without taking into account frequency of use. Disabling features is not a solution. Constant power is. As I mentioned above, if there was a reliable way to connect a wire to it, I would not hesitate spending time and money to do it.

On a separate note, notifications and battery polling in these locks are anything but reliable. Even with brand new batteries put it, I can not reliably count on the HS notifying me when any of the standard malfunctions occur and the related notification is supposed to happen (i.e. jammed lock, critically low battery, lock unlocked for more than X number of minutes, wrong code entered X number of times, etc.).

[VirtualPanther]

You need a AC/DC current clamp such as this one:
http://www.fluke.com/fluke/usen/acce....htm?PID=56297

Very good taste in tools!

[gearyt]

You shouldn't need to use polling if the lock is associated with your controller properly (I don't use Homeseer, but in general this is true if the z-wave device supports the "Association" class and the controller support secure commands).

I don't use polling and get near instant updates (in less than 1 second) when the lock is manually actuated. You are right, frequent polling would suck the battery life away.

yep and maybe someday it will work in HS3.. just not yet

[etc6849]

I'm not sure if HS supports it, but the controller I use is Leviton's VRC0P +3. It has it's own Zensys like Z-Wave protocol that is open and easier to integrate, so I and others built our own Z-Wave driver for Motorola Premise (free user supported HA software). Not as neat as what HS uses, but it is very versatile as users can modify the module in vbscript as needed to support new devices.

I just did a battery check today. I'm still at 100% battery level over a month later! I now wonder if the black Friday batteries I bought at Home Depot were defective or not. I still have every option possible disabled, and agree it sucks not hearing the pretty sound the lock used to make.

If the lock makes it more than 4 months on these batteries, I may replace another Kwikset that is going bad. The Kwikset z-wave locks have no rubber seal like the Yale locks, so if you live in a humid area, they do get condensation inside and go bad.

I think you can remove the z-wave module from the Kwikset locks and buy a replacement Kwikset lock with a keypad from Home Depot, but in my mind the design is flawed and they shouldn't have ever sold it. At least the Yale lock looks very well designed mechanically.

>N2SS128,2
<E000
<N002:152,128,101,001,022,090,176,219,079,177
<E000
<X000
<N002:152,064
<n002:000,128,003,100

[etc6849]

So I guess battery reporting doesn't work on the Yale. I should have suspected this when it was 100% after a month of use.

No matter how low the batteries get, it still reports 100%?!?

Anyways, found the lock completely dead today. The li-on batteries measure .36v. So, they made it a little longer than the alkaline, but no where near the 9x promised on the package.

Next I'm going to try some Ni-MH rechargeable batteries. I'll be selling all my z-wave locks on ebay before the next move. Not worth constantly changing the batteries. The kwikset's were better, but they were a POS mechanically compared to the Yale.

>N2SS128,2
<E000
<N002:152,128,150,109,109,045,199,239,024,238
<E000
<X000
<N002:152,064
<n002:000,128,003,100

[VirtualPanther]

So I guess battery reporting doesn't work on the Yale. I should have suspected this when it was 100% after a month of use.

No matter how low the batteries get, it still reports 100%?!?

Has been same for me ever since i got my two locks. Battery status doesn't work, jamming alarm doesn't work, critical low battery alarm doesn't work, wrong code entered doesn't work. The only notifications I get, out of all I configured, is "door unlocked with code X" Oh, and in lock configuration, you can't name assigned lock codes (i.e. babysitter, etc.). It's "Code 1", "Code 2" and so on. So I have to keep a separate list of what the heck each code means.

[etc6849]

The user naming thing is an issue with Homeseer. The Zensys Z-Wave protocol (specifically the User Code command class) by itself does not support user naming. If I remember right, the zensys protocol document states 255 users could be allowed, but these are numbered, not named.

I use the Leviton VRC0P +3 RS232 controller (what I posted is from spying on the RS232 port). It is at the z-wave protocol level that the device does not give the correct battery status. This appears to have slipped through QA and whatever certification Zensys requires. It's my understanding that battery status is required to work since the battery command class is used. I wouldn't blame Homeseer for this, but I would blame Yale.

Has been same for me ever since i got my two locks. Battery status doesn't work, jamming alarm doesn't work, critical low battery alarm doesn't work, wrong code entered doesn't work. The only notifications I get, out of all I configured, is "door unlocked with code X" Oh, and in lock configuration, you can't name assigned lock codes (i.e. babysitter, etc.). It's "Code 1", "Code 2" and so on. So I have to keep a separate list of what the heck each code means.