Radio Repeater Solar System

David and Laura

It's been another eight months since we last posted, and over a year since our first post, so we figured that it was time to provide an update on how things have been progressing.

Main Home System

Our home system has been running quite well, however, we have yet to commission the monitoring equipment that we've procured, which includes:

* A pQube Power Quality Monitor connected to our main inverter output
* A Magnum MagWeb Inverter Monitor, connected to our main inverter
* An SC-20 to measure voltage levels at the home
* An eMonitor circuit-level current meter

The eMonitor is now installed and wired into our main panel, and we're still building the DIN panel where the remainder of the items will be mounted.

As a consequence, we don't yet have good visibility into our power production. We'll report back on each of these above items as we get them operational, and once we have data being logged, we be able to look deeper into how our system is operating.

Internet Repeater

As we mentioned in our first post, we will be putting up a communications tower on our property at some point in the next couple of years. In the meantime, as our home is located in a valley, in order to get Internet connectivity from our local island wireless internet service provider, we need to construct a radio repeater. Below is our current design for this stand-alone system.

We've located a site that is around 1 Km from our home, and has line-of-sight visibility to both the nearest ISP tower, and to our home. The site has ample sun, and some trees we can use to mount the radios. Below is a photo of the site:

Loads

We have four loads:

* Three Ubiquity wireless radios (Two NanoStation Loco NS2L radios, and one AirMax NanoStation Loco M9 uplink radio)
* A B&B MESR921 ModBus converter to allow us to remotely monitor the solar system

As the equipment (especially the radios) are very sensitive to line voltage, we've specified two DC-DC converters to regulate voltages and provide current limiting.

The manufacturer power specs are 4 watts at 12 volts for the NS2Ls, and 6.5 watts at 24 volts for the LocoM9. The ModBus converter draws 3 watts at 24 volts. This works out to a raw load of approximately 13.4 watts at 24 volts, and if you factor in the inefficiency of the voltage converters, let's round up to 15 watts. This represents an absolutely worst-case draw, and we expect this to be far lower in real life, but until we've measured the current draw, these numbers are a good starting point.

At 15 watts, we're looking at around 360 watt-hours per day. Since we have some fairly long dark and rainy periods in the winter and spring, and we'd like to have at least a few days of reserve, we can multiply this by 3, and double it to come up with a first estimate for how large we should size the battery banks, which works out to a 2.1 KWh bank.

Proposed System

The below diagram illustrates our current design for the system:

Attachment not found.

Black is DC power, Red is Ethernet, and Blue is Serial.

We've been able to salvage eight old Kyocera 50 watt panels, and the specs are approximately 21 volts open circuit, 17 volts at max power, and 3.3 amps at max power.

We propose to organize these into four strings to ensure we don't over load the SunSaver MPPT charge controller.

We're looking at getting four 6 volt East Penn AGMs (something like a 8GGC2), giving us around 200 amp-hours at 24 volts. This translates into 4.8 KW hours, which should let us ride out six days without recharge before we hit 50% discharge.

Assuming 4 hours at 100% solar exposure, and 2 hours at 50% solar exposure in the winter, we should be seeing around 2 KW hours per day from the panels. That should allow us to completely recharge the batteries from 50% discharge in a little over a day, which we think isn't too bad of a balance.

The SunSaver then supplies power to the voltage regulators, which in turn supply power to the radios and ModBus converter. The balance of the system is the string combiner, breakers, wiring, a NEMA enclosure for the electronics, a wooden box to keep the rain off the batteries, racks for the panels, and a bunch of serial and STP wiring to connect the radios.

We're also looking at some PoE injectors/surge arrestors from L-Com to get the required voltages onto the STP cable that runs to the radios.

So, given this description, any thoughts on how this system looks? Are our calculations in order, and look reasonable?

Peace,

David & Laura

Cariboocoot

Re: Radio Repeater Solar System

Looking over the numbers from the only way I know how.
360 Watt hours per day on 24 VDC = roughly 15 Amp hours. Minimum battery bank for one day: 30 Amp hours. Better be larger. Three days worth would be 90 Amp hours or more.

You've got eight 50 Watt panels, that's 400 Watts. You're going with the MS 15 MPPT, and 400 Watts is at the top of the panel capacity for it. Expect about 12 Amps. Maximum battery bank that could support: 240 Amp hours @ 24 VDC.

Your proposed battery bank at 200 Amp hours covers need #1 nicely. In fact it's almost double the minimum size. And it's under the maximum the controller can handle. Charge rate could be around 6%.

Problem: after days of no sun when the bank is the most depleted you have a minimal amount of recharging capacity. You might not have enough daylight (especially in Winter) to fully recharge the bank before the next day. The load will be running at the same time.

This is going to sound weird, but I'd recommend a smaller battery bank. Go for 120 Amp hours +/-. At 50% DOD maximum that gives you 60 Amp hours to work with, or 4 days @ 15 Amp hours max. It will allow a 10% recharge rate which should be able to bring the battery bank back up to full given one good sunny day.

mike95490

Re: Radio Repeater Solar System

the 15a breakers in the combiner, are too large. for a 12V, 50w panel, you could expect 3 amps, so a 5A breaker would be a better match.

David and Laura

Re: Radio Repeater Solar System

Re #1 — Good point. The load draw combined with minimal expected charging during dark wet periods may be a problem. One possible way to fix this would be to use the SunSaver Lighting Control feature so that after several days of minimal charging, the repeater is automatically shut down during low use periods (such as overnight). This would allow the system to automatically extend runtime during these conditions.

Re #2 — Each string would have two panels, so we expect to see 6 - 7 amps peak per string. Would a 10 amp breaker work better then?

---

We just received the ModBus converter, and our order from Northern Arizona Wind & Sun arrives in the mail tomorrow.

Cariboocoot

Re: Radio Repeater Solar System

String panels in series and the Voltage goes up; current stays the same. Like Mike said, probably 3 Amps (50 Watts / 17.5 Vmp = 2.85 +/-). 5 Amps protection per string should do it. Nothing wrong with fuses for this function either.

I think you'd either have to use the lighting control function or not; no automatic way of turning it on if conditions warrant. How accessible will this repeater be? Possible to have a small gen up there to give the batteries a boost if needed?

David and Laura

Re: Radio Repeater Solar System

Re #5 — You're absolutely right, I was thinking watts, not amps. I'll get some 5 amp DC breakers ordered up, and use those instead.

With respect to turning the lighting control feature on and off, the good news is that MSView uses ModBus to configure this feature, so even though they haven't yet fully updated their ModBus documentation to describe the registers used for this feature, it's fairly straightforward to capture the information used to enable and disable it. This would let us set up a little background task that uses information about the load, state of charge, and daily charge current to enable/disable it.

The downside of this approach is a) that it takes a bit of work, and b) that once you enable the lighting control feature, and it turns off the load, we'll loose connectivity, meaning that we will just have to wait until it turns the load back on (or lug the laptop up the hill).

If we used a ModBus relay to turn off everything but ModBus connectivity, we could selectively shed two thirds of the load, but that would require some extra equipment (for example, Control by Web's WebRelay or a CLIC PLC). Of course, this would also burn up another 2.8 watts.

The repeater isn't too remote, an easy half-hour walk during the day. In the night, and in the rain, it's it little more dicey, as there are some steep climbs.

With respect to hauling up a little generator to top up the batteries, that's a good suggestion, except that we're trying to go fossil fuel free. It's too bad that we can't use our chainsaw batteries for that purpose —*They each store 89 watt-hours, almost enough to half-charge the bank.

mike95490

Re: Radio Repeater Solar System

David & Laura wrote: »

..... It's too bad that we can't use our chainsaw batteries for that purpose —*They each store 89 watt-hours, almost enough to half-charge the bank.

Expand on that a bit more, please

David and Laura

Re: Radio Repeater Solar System

That should read 890 watt-hours.

We have a pair of 75 amp hour 12v batteries mounted on a dolly with a 1200W inverter and charger that we used to power our corded electric chainsaw and hammer drill. They're fairly old, and not in the best of shape, and since we've purchased an Oregon PowerNow chainsaw, we haven't been using it much.

The thought was that worst case, we could lug the dolly up the hill, and possibly use it to recharge the battery bank, but that wouldn't work because the SunSaver will only charge from a supply voltage larger than the battery bank voltage.

Cariboocoot

Re: Radio Repeater Solar System

If needs must, the battery/inverter combination could power a charger to boost the repeater batteries up. Not very efficient at all, and heavier than a Honda 1000. But it would work.

Photowhit

Re: Radio Repeater Solar System

I challange a winter exposure of basicly 5 hours of full sun for winter in British Columbia, It looks like 2 hours or less average exposure. Unless this is somewhere further south.

I disagree with a smaller battery bank, you have the same loads regaurdless, so the batteries would remain in an even lower state of charge. I think your battery is fine and you'll want a larger array based on <2 hours of solar issolation(?).

Cariboocoot

Re: Radio Repeater Solar System

You certainly don't get 5 hours of sun in Winter at my Lat. in B.C.; the days dwindle down to about 6 hours total, and you'd be lucky to have half of that in usable sun on a good day.

I stand by my suggestion of a slightly smaller battery capacity to improve recharge time after days of no sun.

niel

Re: Radio Repeater Solar System

i will only speak in generalities on this as i am not as familiar with internet repeaters. i will say there seems like somethings amiss here when you have to spend that kind of money just to repeat the internet and to me it seems like a big waste as satellite service should be able to provide the same level of service at a cheaper cost to you. mind you i'm not real familiar with satellite services either, but i'd be darn if i'd spend that much just to have internet service.

Cariboocoot

Re: Radio Repeater Solar System

I've got not one but two different satellite Internet connections. They are not necessarily cheap to install or operate either. Sometimes the location prevents you getting a line-of-site with the satellite so they won't work.

Even so, it might be worth a look at for your install. The new Xplornet system we have works better than the old one, and is less expensive per month (around $50). Could save a lot of trouble if all you have to do is up the solar at one location to handle the extra power demand.

Photowhit

Re: Radio Repeater Solar System

I kinda recalled their earlier post, and they are providing internet for a few people in a community on Gulf Island(from an earlier post), They are in BC so my concerns about winter sunlight will need to be adressed.

Looks like a WISP (wireless Internet Service Provider) using directional antennas, likely you can have a much greater band width, I think Hughes and WildBlue restrict throughput after 2GB per day and have restrictions per month as well. This would be of even greater concern with multiple users.

icarus

Re: Radio Repeater Solar System

Niel,

My SAT ISP was ~$100/month. It had a download speed of .6mbs, up load of .1, with ping time near 1 second! And that was with ideal conditions, few people on the node, ideal weather. My current HSPA wireless system down loads ~5mbs, up loads nearly as fast, ans has tiny ping times,, all at a cost of $30-75 month depending on how much I use it (20 gigs is $75). Not to mention it only uses ~5 watts, for modem a nd router combined, vs 25for the sat modem and router.

Tony

niel

Re: Radio Repeater Solar System

photowhit,
i stand corrected as i overlooked what they had posted in another thread that it is for multiple broadband hookups throughout the valley there. a sat link won't due here so the hard road of powering the repeater is necessary.

d and l,
sorry for my interruption on your inquiry.

David and Laura

Re: Radio Repeater Solar System

Re #10 — This is based on the sun that we got this winter for our main solar system at the house. We're pretty close to the U.S. border, so it's about as good as it's going to get for Western Canada.

Looking at some nearby weather station data, on a clear day, we start getting sun at around 8 A.M. reach peak solar insolation at around 11 A.M., with levels starting to drop by 2 P.M. down to near zero by 3:30 P.M. Daily peaks range between 300 and 450 Watts/m^2.

We've set up a Davis weather station at our home, but we aren't data logging yet, so we don't have site-specific data yet.

Adding additional panels is something we'll consider based on how things run for the first year. The panels that we're using are ones that we had on hand from an older system, and we're trying to minimize costs.

I think that the key takeaway is to plan for situations where the batteries get depleted, and make sure that they don't stay at 50% discharge for long periods of time.

Re #11 through 16 —*We already have satellite Internet installed at our home from Xplornet. The service has extremely high latency (1 second RTT), and often will have up to 50% packet loss and 30 minute dropout periods, especially when the weather is bad. It's good for a last-resort backup, but nothing compared to the 20 MBit/s wireless service that we can get if we get line of sight to their tower.

I haven't measured the current draw of the satellite box, I'll make a point of doing so and get back with some real-world current draw numbers.

We're aiming for paying under $2,000 for the repeater, using lots of existing parts we already had on hand. The largest individual component cost is the batteries. Given that Satellite costs us $100 per month, and local service is less than half that, the system will pay for itself in around three years, ignoring the higher grade of service.

westbranch

Re: Radio Repeater Solar System

David & Laura wrote: »

Re #10 — This is based on the sun that we got this winter for our main solar system at the house. We're pretty close to the U.S. border, so it's about as good as it's going to get for Western Canada.

Sounds like you are on Saturna Is.

David & Laura wrote: »

Re #11 through 16 —*We already have satellite Internet installed at our home from Xplornet.

Friends just switched to Galaxy ISP from Xplornet, very impressed and a $35 drop per mo. on fees


HTH

David and Laura

Re: Radio Repeater Solar System

Here's another question:

Given that we're using a charger/load controller that is limited to 15 amps charging/load current (the SunSaver MPPT), and we'll be installing a 20 amp DC breaker between the battery and the controller, what wire gauge should we use to connect the batteries together into a string, and to the breaker?

Wire rated for 30 amps (10 AWG) feels like it is too small, and I know that oversizing helps reduce voltage drop. To what extent should we over-size the wire between the batteries and to the breaker and the charge controller?

The charge controller will only be two feet away from the batteries, so we're talking about a pretty short wire length.

Cariboocoot

Re: Radio Repeater Solar System

I'll hazard a guess that the Xplornet system you have is the one on the Hughes satellite. :roll:

We've had four of their systems; the first being their predecessor Lyncsat. That was the only choice then, and it was marvelous except for not being anything remotely like high speed and no large data capacity either. It would not operate with VOIP. The second, which is still in place here at the cabin and what I'm using now, is remarkably reliable and fairly speedy although streaming video is not really viable. The VOIP works, with some satellite delay. The third was the Hughes system put in at the place in town. Far, far worse than any other Internet service I've ever seen of any kind. No chance of VOIP or anything of the sort. Undependable, expensive, just horrible. Then they let us be Beta testers for #4 there, and it is far better than any of the others. $50 a month with 30 gigabytes allowance, no "FAPing" (satellite users will know what that means) and the speed is remarkable. Thinking about switching the cabin system over, if only to save $50 a month. It would soon pay for the install (expensive due to location).

We have no cell service available at either locale, even with yagi antennae and boosters. Not enough customer base to justify the install of towers able to service the area.

Cariboocoot

Re: Radio Repeater Solar System

David & Laura wrote: »

Here's another question:

Given that we're using a charger/load controller that is limited to 15 amps charging/load current (the SunSaver MPPT), and we'll be installing a 20 amp DC breaker between the battery and the controller, what wire gauge should we use to connect the batteries together into a string, and to the breaker?

Wire rated for 30 amps (10 AWG) feels like it is too small, and I know that oversizing helps reduce voltage drop. To what extent should we over-size the wire between the batteries and to the breaker and the charge controller?

The charge controller will only be two feet away from the batteries, so we're talking about a pretty short wire length.

Two feet is not going to be a big concern for V-drop. Even on 12 Volts @ 15 Amps. 12 AWG is sufficient, and will handle over 20 Amps which the circuit should never see.

David and Laura

Re: Radio Repeater Solar System

We've made a fair bit of progress on assembling the system. Below is the controller box temporarily mounted in the power shed for testing (and to keep the batteries charged while we finalize the location where the repeater will be installed).

Attachment not found.

At the top left, we have the MPPT controller, with the battery, panel and load breakers on the lower left. Beside the breakers are two DC voltage converters/regulators that ensure that the radios get stable power at 12 and 24 volts. Each load has individual breakers, which feed into PoE injectors. Finally, at the top right is the Modbus to Ethernet converter.

So far, everything seems to be running well.

Still to be done:

1. Finish building the enclosure for the batteries. The electronics enclosure will be mounted on the back.
2. Finish wiring in the PoE injectors
3. Ethernet cabling
4. Modbus cabling
5. After final mounting, install the correct cables to the panels and batteries.
6. Drag the parts and re-assemble everything up the hill
7. Install the radios up a tree, and commission everything

TheBackRoads

Re: Radio Repeater Solar System

Keep us updated! I'd like to see how this ends up!

David and Laura

Re: Radio Repeater Solar System

We finished building the shelter for the batteries, and hauled it and the panels up the hill.

Attachment not found. Attachment not found.

Next up: Moving the batteries and electronics.