batteries/generator for repeater

[DAL-COM]

A local PD has a Midland 100W repeater with a propane generator for secondary power. However they have no battery bank, I am wondering if they should, and how much runtime (ie AmpHrs) would be desirable. Is a 1 hour battery runtime considered adequate in the event of primary and secondary power failure?
Any suggestions on battery chargers/controllers?
Thanks

[Znarx]

a good rule of thumb is to provide enough power to last 1.25 X response time (16 hour response time = 20 hours of battery power)

you will have to calculate battery capacity based on time required and current draw of your equipment

...Z

[RKG]

I believe a battery back up power supply is a good idea, basically for two reasons. First, if the generator fails to start, or if it starts and takes load and then drops it (for instance because it was loaded too quickly while cold, stumbled, and tripped out on either low voltage of off-frequency), you can keep the radio going long enough to get the Highway Department up to the site with a portable generator. Second, even if the auto-start generator works, there is a long enough outage to cause the radio (and any associated powered equipment) to re-boot, which not only takes it out of service for a finite amount of time, but raises the potential of causing a boot-diagnostic error, which will keep it out of service until cleared.

There are basically two ways to back up a remote repeater: using a built-in function provided by the radio, or the way I'm going to describe. I don't like the former. Most radio backup provisions are hastily designed, and designed by people who don't know much about batteries. As a result, they use unregulated "trickle" chargers to float the batteries, and these chargers gas the batteries into premature failure (which may not be detected until the lights go out), while at the same time the charger functions will be incapable of truly charging the batteries in the event of a real drawdown, thus drastically shortening battery life. Some of these devices reduce transmitter power, or shut down some features, or both, to save battery power. And some have a transfer delay that is almost as long as the delay in a power transfer to an auto-start generator.

Here is the solution I use, which thus far has been without issue. The required equipment is a device called the "Heart Interface" (by Xantrex), usually in the 1 kW size (about $1,000), two Group 27 or Group 31 gel cell batteries (nominally 85-90 AH capacity per) (about $150-200 per), a good DC circuit breaker (such as Blue Seas 710x series), rated for 0.2 times the capacity of the Heart Interface (i.e., 200A for a 1 kW Interface), a device known as an "E Meter" (now sold as Heart Link 10; West Marine SKU 255200), two duplex 110V-15A outlets, outlet box, and suitable lengths of NMC conduit and some Romex.

System is wired as follows:

DC cables from Heart Interface are wired to batteries in parallel, through the 12V breaker.

Heart "AC In" is fed directly by a breaker on the Critical Circuits AC panel (the panel supported by the generator).

Heart "AC Out" is wired to the two outlets, mounted near the radio in a box that is painted red and legended: "Danger: These outlets are live even during loss of off-site power!"

Radio is plugged into the outlets.

System works as follows:

While there is line-power, it is transferred to the outlets by the Heart Interface.

In the event of loss of line-power (defined as voltage above 135 RMS or below 90 RMS, or frequency error greater than 4 Hz), the Heart Interface switches into inverter mode and supplies 110V power to the outlets from the batteries. Transfer time is about 6 msec, and so quick that the radio will not see it and will not have to reboot. Radio operates at full power and with all features (it doesn't know that off-site power has been lost.)

Once either line-power is restored or generator starts and accepts load, Heart Interface switches to charger mode and restores any nick taken out of the batteries. This is a computer-controlled "Smart" charger, so it will not overcharge the batteries or cause them to gas, yet it will truly recharge them in a hour or so.

The E-Meter gives the tech good info on battery state and condition, and records all cycles for download to a PC. This permits one to keep track of battery performance, and any weakness, long before the batteries fail.

Since the gel cells do not gas, they can safely be located in occupied spaces.

For a 110W Quantar, which draws about 60W AC when not transmitting and about 600W AC when transmitting, we estimate that the two gels cells will support the radio for between 2 and 10 hours without drawing the batteries below the 50% of capacity point (which should be the minimum drawdown of any deep cycle battery). In fact, during testing we have run the Quantar for up to 3 hours without seeing battery voltage drop below about 12.5VDC.

The only thing missing from this system is a form of notification that the radio is operating on the batteries; any "battery revert" function built into the radio (such as in Quantars) isn't available because the radio doesn't see the loss of AC power. Assuming a need, there are a host of simple ways to gin up such a notification device.

Total cost of this system (exclusive of labor) is under $2,000. Batteries should last at least 5 years (and, in fact, have lasted more than 10 years and counting in some of our installations).

Sorry for the long-winded response, but you've touched on one of my favorite subjects. (No comments, Paul.)

[DAL-COM]

I would think 2 hrs would be adequate for a response time, as the site is in a city. The Heart Interface plus gel cells sounds attractive, RKG.(I knew you would have to reply, based on earlier postings!)
I will check into it and post further if needed.
RKG, if you have any schematics, however rough, can you email to me?
Thanks.
ps 600W to feed a 110W repeater??? What is Moto doing?

[RKG]

Never bothered with a schematic, as the connections are so simple.

The Quantar starts by taking 110VAC in and jacking it up to 400VDC, then drops and distributes it. Plus a lot of other things are going on while the radio is transmitting. According to M, that adds up to close to 600W, and by inspection they are not far off.

Remember that the number of discharge/charge cycles you can get out of a well-made storage battery is an inverse function of how deeply you discharge them: <30%, probably 1,000 cycles; ~50%, about 500 cycles; >65%, maybe as few as 200 cycles. Given that and the fact that we want to support adequate voltage for the Heart, and we have 2 G-31 gel cells. Better to have a bit more than you need than a bit less.

[ricciticcitembo]

Yeah those RKG power supply responses are like Gold.
I'm going to print out the above and use it for a shopping list.
I've never heard or seen anything simpler, cheaper, or better all
at the same time.

I use 1.4 KW APC uninterruptibles with a pair of DieHard Golds,
and a seperate computer Charger, with the APC internal charge circuit disconnected. Very similar operating characteristics as what
is mentioned above.

But that was before I had the benifit of RKG's answers.
His is a couple notches better.
But in all fairness my setups have
been running around 5 years and the Golds are still good. I have
3 identical setups at different locations like this.

Dave

[RKG]

The only reservation I have is that Die Hards are "flooded cell" batteries; they will give off some hydrogen gas when charging, and that makes them dubious for installation in occupied spaces.

[ricciticcitembo]

Yup. I really should put 'em in the boat or car and get some
sealed Gels like you said. I chose them because the outgas the
least of all the wet cells I tested. But it is still NOT a good idea.
They handle a 200A Draw well though. They barely get warm even
at full throttle on the APC running a Laser Jet and a couple of big
screens. (only to test with. Actual load is a LOT less than that)

[Susan157]

We have been using ProPane In Canada for 30 Years
for remote radio Sites before the Solar Panels
Came Along.They Are ProPane In and 13.6 volts Out.
They Worked !00% With Very LittLe Maintenance.
They also had batteries that were wired to them.

The sites were not looked at in the Winter.
Only If You Had a Break-Down Then There
Was The Helicopter Trip.Great Old Units.

From The Great White North (Canada I Mean)

[KitN1MCC]

a quik and dirty setup that i have used is good UPS. for short drops out. and some what longer as well.

Plus UPS works great for DC remotes

[jbxx]

I've used L16 Trojan batteries on a couple of
my installations, they are beasts.
6V @ 400Ah . They average about 7 years before needing
replacement. About $200 a pop
J.B. AD6WX

[RKG]

Big 6V batteries are basically glorified golf cart batteries, and they are quite robust. You wire two of them in series (not parallel) to get 12V. The problem with the Trojans is that they, too, are flooded cell batteries (which means that electrolyte has to be monitored and periodically replenished and they should not be used in occupied spaces), and the problem (sometimes) with 6V batteries is that they are 2 inches or more taller than G-27 or G-31 12V batteries, and may not fit in space designed for 12V batteries. If you have the space, though, try the Dryfit (or Dryfit-licensed) 6V gel cells, such as West Marine SKU 129355 ($230). Since the load you will put on them is lower in amps than a golf cart motor, they should be virtually indestructible provided you charge them with a charger designed (or programmed) for gel cells.

[kf4sqb]

RKG, that's a nice setup you detailed there. Now if only I could get the wife to let me buy one to put the house on......

[srefurd]

Does anybody know if the Optima series of batteries outgas?? I think they are a "paste" cell and I know that there are deep cycle version available. Just a thought.

[RKG]

I believe that Optima batteries are of a type called AGM (absorbent glass mat), and they are designed to be "recombinant," like gel cells, which means that any hydrogen generated during charging is to be recombined with oxygen to re-form water within the cell. However, I'm not an expert in AGM batteries.

However, AGM batteries use charge parameters that are different from both flooded cell parameters and gel cell parameters, so if you use AGM batteries, but sure to use a charger that is designed to charge AGM batteries. There aren't many of those around.

One other point about the Heart Interface, which will impress those who are NEC-conversant:

When line-power is available, the Heart Interface is smart enough to realize that it is functioning as a "sub-panel," and it does NOT bond the neutral and ground wires. This is important, because if the neutral and ground wires are bonded in a sub-panel, the ground between the sub-panel and main panel will be "hot" during routine operations, and this could be dangerous - potentially fatal - to someone working in the main panel. When line-power is available, the Heart Interface passes the hot, neutral and ground wires separately to the line hot, neutral and ground wires through a three-pole switch.

When line-power is not available, however, and the Heart Interface is working as an inverter, it is smart enough to realize that it is now functioning as a "main panel," and it DOES bond the neutral and ground wires. This is important to avoid a voltage differential between the neutral and ground wires, which could induce stray currents in attached devices. When line-power is not available, the connections between the Heart Interface and the line hot, neutral and ground wires is lifted by the three-pole switch.

Those who use other devices to construct similar systems should consult the technical literature that comes with their devices to be sure that they perform the same way.

[ricciticcitembo]

Those who use other devices to construct similar systems should consult the technical literature that comes with their devices to be sure that they perform the same way.

Yup Again. And it is a pet peeve of mine as well. The APC's do
exactly what the Heart does, and have Ground Fault Indicators on
the front panel. And your correct in assuming that MOST of the UPS's DON'T do it right.

It really bugs me when the Ground and Neutral are tied together
in a sub panel.

And another thing that bugs me is when there is a three prong
outlet(s) hooked up to 100 Years old Knob and tube wiring with NO
ground at all. And then they just attatch some romex to it, and add
a couple more outlets with the bare ground wire just floating.

Damm. I just can't believe it. But it seems that a lot of people do it.

I hate that, and have no qualms whatsoever ripping big holes in
the wire lath construction to replace the feed lines completely.
No matter how many walls I have to knock down to do it.

[CuriousGeorge]

For a 110W Quantar, which draws about 60W AC when not transmitting and about 600W AC when transmitting

Anyone have a good rule-of-thumb for sizing power supplies? Yes, every radio is going to have its own set of exact specs but there is probably some good guesstimations one can make as well. I'm particularly interested in mobiles used in (temporary) fixed base operations but, obviously, a good rule of thumb should apply in many other situations as well. Separate RX and TX values would be even more helpful since some cases these radios are being used a glorified scanenrs -- albeit with better sensitivity, etc.

Or, better yet, any good reference for the draw on various radios? (Besides their literature/label which is either lost, worn, not available until the radio is in your hands, etc.) A 25W Radius mobile? A 40W CDM series? (etc.)

[DAL-COM]

RKG, why would the neutral-ground bond cause the neutral between main and sub to be "hot"during routine operations? Maybe I need a schematic to see it.....

[RKG]

If the neutral and ground are bonded in a sub-panel, and if a device (motor or light) powered by the sub-panel is "on," then the current flowing through the device that would otherwise go back to the street through the neutral will now split and flow down both the neutral and so much of the ground as is south of the sub-panel. If, in that situation, one were to touch the ground in the main panel while himself grounded, he'd get a shock.

Now, in a perfect world, any time you are working inside a live panel, you put your left hand in your pocket and work with only one hand. (That way, if your right hand touches a live conductor, the likelihood of current traveling across your chest and zapping your heart is reduced.) However, in the real world this isn't always possible, and having current flow over the ground is considered a severe safety issue under the Code.

RTT: one of my practices, for which I have on occasion received a bagful, is to run a crude test of the ground on every outlet I work with. I use a neon tester shorted across the hot and ground, and if it lights, I consider the ground acceptable. (Yes, I know there are better ways of doing this.) If the tester doesn't light, I not only won't proceed, I'll red-tag the outlet. Needless to say, if you tend to work in 100 year old public buildings that are still using the original wiring -- not at all unusual in New England -- this can cause some displeasure. Oh, well.

[kf4sqb]

If the neutral and ground are bonded in a sub-panel, and if a device (motor or light) powered by the sub-panel is "on," then the current flowing through the device that would otherwise go back to the street through the neutral will now split and flow down both the neutral and so much of the ground as is south of the sub-panel. If, in that situation, one were to touch the ground in the main panel while himself grounded, he'd get a shock.

Exactly why NEC no longer allows 3-wire cordsets on clothes dryers. Heating element: 240vac. Motor: 120vac. They have been using the ground as a current-carrying conductor for years!

[DAL-COM]

I would argue that you won't get a shock from the subpanel unless two things happen: the neutral conductor opens (it does happen) AND the ground conductor opens (which it could under full load since it is generally smaller than the neutral). But simply running one light or motor probably is not enough to open the ground.
Regardless, it would be a dangerous situation so the code rightly forbids it.
Ground conductors must not carry current under normal operation.

Which reminds me of a common cause of structure fires- an open neutral forces current to flow in metal pipework that connects two buildings, causing a fire in the building next to the one with the fault.

[Tom in D.C.]

My two cents worth about a UPS goes like this. About fifteen years ago I needed a reliable UPS when my wife was working mostly at home as a reporter and e-mailed most of her copy in to the paper from the house. I found the Best Power (tm) units to be far better, at least on paper, than the standard "computer" units so we bought an ME500 with extended runtime. The unit weighed a ton, but worked for us for years without a bit of trouble. Later on, when we didn't need it any longer, I sold the ME500 to my company where it's still working today after two battery changes. If you look for the UPS at your Home Depot or local supermarket, where the system must be online at all times, you'll most often find that they're using a big Best Power unit. I'm sure that after all these years there are today other companies who make really good units similar to the Best Power rigs, but if you investigate how various UPSs are wired and set up you'll find out that the most reliable ones cost a lot more because they're just plain better.

Me? Right now I'm using an older APC Back-UPS on my P4 that I paid all of $15 for at a hamfest. So far it's suffered through one power dip when I was online and the whole setup never blinked.

Tom in D.C.
W2NJS

[DAL-COM]

CuriousGeorge, re your question about sizing supplies: most modern 50W mobiles seem to take about 100W from a 13.8VDC source during TX, ie about 7A. Most vendors recommend doubling that for reliability ie around 15A. Usually, I will install a 20A unit for longer life and quiet operation, unless price is the dominant spec. I generally use linear supplies because weight is not a factor in a base station.

[ASTROMODAT]

I run big Ratelco inverters on my Quantars that should provide 96 hours of normal service in the event of commercial power failure.

Should the inverters fail, I also have 48 hours of DC battery backup, based on the stock Quantar power supply with automatic DC back-up (and suficient sealed lead acid batteries).

I also have a diesel generator at the site, with enough diesel fuel and engine oil to run for 45 days.

It's the old Boeing's phenomenon of "wearing a belt and suspeners!"

Larry

[CuriousGeorge]

CuriousGeorge, re your question about sizing supplies: most modern 50W mobiles seem to take about 100W from a 13.8VDC source during TX, ie about 7A. Most vendors recommend doubling that for reliability ie around 15A. Usually, I will install a 20A unit for longer life and quiet operation, unless price is the dominant spec. I generally use linear supplies because weight is not a factor in a base station.

Thanks. I do have the datasheets for the CDM series radios from Motorola and have been relying on them as a guide. But your comments are definately appreciated.