Backup power for surveillance systems generally rely on batteries, especially since UPSes for computers are common and easily available.
However, uncertainty in picking the right backup power supply sized with the right batteries is a common problem, and the pitfalls of poor selection stretch beyond just having a weak system. In the note, we look at battery backup, the most common method for surveillance power backup.
Inside we examine:
UPSes run time delivered
Understanding UPS power units of measure
How to calculate surveillance system wattage
Using runtime graphs to determine supply duration
How much backup runtime is needed
Common factors affecting runtime
Why consumer UPSes often are too weak
Why battery equipped power supplies may not be enough
Thanks for putting out this article. Surveillance without battery backup is a problem waiting to happen. Too often the integrator forgets to add or let's their customer supply a UPS. Cost is the big reason that they don't automatically add them into their design. The correct UPS should eliminate a number of problems that may be the cause of call backs to a job site.
I think further information into the 3 main types of UPS products would be very helpful to your members. Knowing the differences between a Standby, Line Interactive, or Online/Double Conversion should produce produce better results. Also, the importance of power factors when comparing different brands can help reduce costs. A unit with a greater power factor can allow a 2000VA unit to produce 1760 watts of backup versus a 2200VA unit that yields only 1640 watts. Most would assume the 2200VA would automatically be superior in performance due to its higher VA rating.
Finally, how do you calculate PoE switches into your example? I have always used the full power potential of the switch (i.e. 16 ports X 15.4 watts = 246 watts) as my number to calculate even though I may only be using 12 cameras at that time. I always design with the future in mind in case more cameras are needed down the road. 246 watts is more than 72 watts (12 x 6 watts) you considered in your example. Any thoughts?
Re: "Finally, how do you calculate PoE switches into your example?"
There are two basic options:
1) I think your approach to use the max output of the switch is smart for planning, especially newer systems that can expand. The total power needed is whatever the switch can output combined with the overhead would be the approach.
Take for example this unit with 10 W overhead and up to 50 W PoE over 4 ports. You'd just strike the values of the cameras and roll them into the switch demand.
The second approach is more granular:
2) For mature systems, PoE power is still issued from the switch, but you take the actual wattage values used for PoE and add those to switch overhead. Plans are based on the max demands of the cameras, rather than the max potential of the switch.
The risk with this second option is lacking flexibility for changes or adds, so it only makes sense for systems that will not change and have no plans of expansion. The benefit though is that UPSes are not overspec'd using this method, potentially avoiding expensive overcapacity and maintenance.
Re: "Knowing the differences between a Standby, Line Interactive, or Online/Double Conversion should produce produce better results. Also, the importance of power factors when comparing different brands can help reduce costs."
Great feedback. I'll add sections on PF/power factors and the three UPS function methods to this post.
I am running into a problem when generator testing creates issues even with UPS installed.NVR gets spiked and locks up , cameras need rebooting. Apparently using a more expensive UPS that utilizes transformer is the only way to protect from if. Any comments?
The fact that the NVR is getting "spikes" while running the generator indicates that the UPS is either passing dirty power from the gen, or maybe the spikes are really drops from the UPS switching on and off if the gen output isn't stable. If possible take a scope and see what the AC looks like at the input and output of the UPS, that will help you identify the problem.
The type of UPS you're using can make a difference. A lot of UPSs just pass AC through without any real filtering as long as an acceptable input voltage is present; these are the offline/standby type. So if your gen's AC is dirty that can get through and cause problems. Or if the UPS senses an AC lose it has to switch internally from the AC source to the batteries, and spikes can occur. Depending on your equipment this can be a problem.
If you don't have a scope, you could try running the NVR directly off the gen and seeing if it runs OK, and then try running it off the UPS with no AC input, and finally plugging/unplugging the UPS from a wall outlet to see what happens when power comes and goes. This should help you identify what's causing the problem.
A line-interactive UPS is a better choice than a standby type, but I'd only recommend an online/double-conversion one for mission critical applications. These never pass AC through the UPS, the output is always fed directly from the battery so no switching occurs and the input AC is always filtered by the rectifier, though you'll see a slightly higher power use because of this.
Depending on the cost of your UPS, an alternative to replacing may be to use an automatic voltage regulator/line power conditioner. You should be able to find one for a lot less than a good UPS, and you could put it between the gen and the UPS or between the UPS and NVR depending on what your testing indicates. Unfortunately even with the testing this is a bit of a 'try and see' solution.
I would definitely use an online UPS when connected to a generator for the reasons stated above. More expensive but necessary to handle the "dirty" power output. You will probably need only 5 minutes of runtime in your calculations to cover the startup of the generator. Good luck.
Perhaps one of you guys with the appropriate background could define "dirty" power from a generator. I'm pretty sure it has to do with the sine wave (heard about this during the hurricanes of 04-05) but have not had access to anyone with knowledge to discuss in depth. I've heard this can wreck electronic components, sometimes not right away. It would be good to know to pass along to owners.
When you think about adding to this article may I suggest including reference to the use of LiFePO4 batteries as an alternative to standard lead acid/AGM batteries? Because they last so much longer, use the same charging technology, and are made in sizes that match many of the existing UPS battery size, they can be a drop-in replacement, if the application can justify the higher cost.
To me this is generally a no-brainer, since standard batteries lose capacity so quickly and we generally don't find out until it's too late, but they are not cheap.
I have little knowledge and ideas of the differencies but i know there are differencies.As far as i know some wave form might be better than other (or maybe it's all about the applications so one wave form is better for one application as another wave is better for another application?).That's why i think this,very well done article,should include also a part about wave form since when you buy an ups if you read the datasheet the wave form is reported
Tapping into my rusty memory, sine waves are always the 'best', i.e. the most compatible with all equipment and the least likely to cause problems.
Square waves by their nature have a lot of harmonics, which means they are really composed of a lot of sine wave signals at different frequencies mixed together. Equipment can interpret this as noise, and depending on the equipment this can cause problems. PCs are a classic example of a device which generally doesn't do well on a square wave power supply.
The noise of a square wave can also radiate to nearby equipment, showing up as RFI (radio frequency interference). A sine wave supply doesn't do that.
Of course sine wave supplies come at increased cost, so that's the trade-off.
A decent article that sums up some of the issues is at: