[time-nuts] UPS for my time rack

Tom Van Baak tvb at LeapSecond.com
Tue Oct 13 14:07:39 EDT 2015

I found this UPS article fascinating, because it echoed what I eventually found in my own lab:


I used to rely on one massive UPS (along with natural gas generator) for my entire lab. Eventually I found it more reliable and convenient to have localized power backup. By local I mean backup for a single shelf, or even a single instrument.

Consider that many Rb/Cs standards and even some Qz standards have internal batteries. Even if one chooses not to use their internal batteries, most of these instruments still feature dual power inputs. In addition to power redundancy it also makes it easy to move equipment or cables around without power loss. Most importantly, local backup like this avoids the possibility of single-point lab-wide power failures.

Recently, as some of my gear works from 5 VDC, those ~2600 mAh mobile phone USB backup power bricks make an excellent mini-UPS. The ideal models are those without LEDs or on/off buttons so they discharge and charge/float seamlessly without manual intervention, even if fully drained.

Multiple units can be placed in series for additional, if slightly inefficient, capacity. A good self-test is:



----- Original Message ----- 
From: "Poul-Henning Kamp" <phk at phk.freebsd.dk>
To: "Discussion of precise time and frequency measurement" <time-nuts at febo.com>; "Bill Byrom" <time at radio.sent.com>
Sent: Monday, October 12, 2015 11:44 PM
Subject: Re: [time-nuts] UPS for my time rack

> --------
> In message <1444701906.379459.408467601.5676DFA5 at webmail.messagingengine.com>, 
> Bill Byrom writes:
>>Anything can (and will) fail, [...]
> The interesting thing is that several sources in that business have
> reported to me that about 30-40% of all power related downtime is
> caused by Battery, generator and UPS failure, in that order.
> Many sites simply have lower uptime after they install UPS systems.
> -- 
> Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
> phk at FreeBSD.ORG         | TCP/IP since RFC 956
> FreeBSD committer       | BSD since 4.3-tahoe    
> Never attribute to malice what can adequately be explained by incompetence.

More information about the time-nuts mailing list