Friday, November 28, 2014

Day 1: A Mini UPS

So I realize that some of you may think that I'm cheating on this one, but this is an idea that I had on my own.  First let me describe the problem that led to me thinking about the solution.

I use Vonage, which requires the use of an ethernet to telephone line interface.  I have an old Motorola Voice Terminal box (VT1005V) that I've had since I started with Vonage (around 2002).  It all works well in normal circumstances, but when I lose power in my house, it causes problems.  I'm sure that this isn't supposed to happen, but the VT1005V interferes with one of the FiOS TV box's reacquiring it's IP address via the FiOS router.  As a result, every time the power trips, even for a brief moment, I have to reset the router with the VT1005V unplugged, and let everything get set and then plug the VT1005V back in.

It's a hassle that could be solved by ensuring that the VT1005V didn't turn off.  The rest of my network gear (the FiOS router and the ethernet switch that the VT1005V is plugged into) are already on a UPS.  Unfortunately, the VT1005V is located in my garage all by itself so I can connect it to my house's telephone wiring.  I could just buy a normal small-ish UPS for the garage, but the least expensive one that I can find (having searched Google Shopping, MicroCenter and Amazon) is ~$40.  It also feels reasonably wrong that I'm plugging in a DC device into an AC output of a DC device (the UPS).

You may ask the reasonable question, "Why don't you just set up your VT1005V with a static IP address so that it doesn't interfere with the assignment of the other TV box?"  The answer is a bad one: the VT1005V doesn't respond to web requests, like it manual suggests that it would.  It doesn't even respond to ping requests.  So there is no setting this up with a custom configuration.

That leaves me with the option of making sure that the VT1005V never turns off.  $40 and I'm done.  But this device has very minimal power requirements and is DC, so why can't I buy a device better suited to that need.  I can't find any devices that meet my needs - at least that are readily available in retail.  This is where the internet comes in.

So, I knew what I wanted and kept looking.  Eventually, I did find a device that appears to get it right.  This would be perfect and if the Alibaba price of $16/unit was the retail price, I would be sold.  However, I'm assuming that the actual retail price would have to be roughly a 2x multiple of this.  At $30/unit, it feels less compelling, but still of interest.



There are a range of options, as it turns out.  The simplest would get the job done.  It's basically a battery, but a battery fit for it's intended purpose.  A more complex option (shown above), could get the job done in a cabinet where you have a number of DC-driven devices next to each other.  Even USB powered devices (which I think that we'll see more of).

So, it appears that somebody figured out there is a need.  But I can't get access to the gadget that is the solution.  Unless I want to become an importer.  Which I don't.  Unless I have six other really good ideas.

One Step Beyond

The only reason that I have interest in this product is that product designers have missed the boat.  What boat am I talking about?  It turns out that 75% of all electricity outages (at least reported to the APPA for 2013) last about 72 minutes or less.  See the below chart and look under the SAIDI (System Average Interruption Duration Index) column.

The second chart just suggests that these statistics don't vary too much over time.  So if manufacturers put batteries into their products that could power the product for 75 minutes, customers could avoid 75% of problems with running their devices in the first place without relying on a UPS.  For high power usage devices, this probably doesn't make sense.  My (old) 65" TV can draw almost up to 1 kW.  I have no expectation that Sharp would build in a battery.  That said, when its a clock, Voice Terminal, router, switch, Roku, or the like, I think that this is quite reasonable.  What would this look like for product designers to build this into their products?

The most obvious problems with building batteries into devices are the following:

  • Additional cost
  • Increased weight
  • Battery could fail before unit is done
  • Increased energy usage
That said, each of them could potentially be offset:
  • Install a standardized slot with a name-brand battery (Energizer).  Have Energizer subsidize the slot on the expectation of replacements down the road.
  • Advertise the benefits to consumers.  We all know that our power goes out a bit and especially no days with most of our computers being handhelds or laptops (that continue to have power during an outage), losing access to the internet is disruptive and inconvenient.  I'd pay extra.
  • Addressed in the first bullet point.
  • For energy usage, the manufacturers could compare their units usage to that of a combined UPS and their non-backed up device.  Their devices would probably compare favorably.

No comments: