Roof Lift Test
Two 100W Renogy solar panels taped to the Intrepid’s roof.
After wrestling with semi-flexible panels vs framed conventional ones, plus a myriad of ways to mount semi-flexible panels to avoid heat and vibration-related failure, it began to sink in that it was really going to boil down to a quasi-lifestyle choice.
I’ve been leaning heavily toward heaps of reliable electrical power with minimal inconvenience, which translates to using some some 600 watts-worth of lightweight solar panels on the Grandby’s roof. Find a campsite, raise the roof, and voila, you’re camping! What could be better? Then I found that semi-flexible panel technology is currently in a bit of a crisis because they do more than just lose efficiency in heat: they distort and fail. So they must be allowed access to cooler air, top and bottom. Vibration or regular flexing kills them as well. The common method of gluing them to the roof turns out to be the worst approach, and they can stretch and sag if suspended without support. Low points caused by distortion can hold rain run-off debris that creates hot spots or local overheating and damage, which is already much more likely in affordable-class panels using low-quality cells. They are not as durable in exposure as conventional panels, because they are faced in plastic instead of glass. Still, they are more efficient and compact, lighter, and can offer 600 watts on the Grandby. But it comes at a price: $1,200 for all six. The unsettling part, besides the need to juryrig a custom mounting system that these panels are not designed for, is that they have largely been withdrawn from the market. WindyNation is the only reputable low-end source I’m considering, and even then, they only offer them direct instead of via Amazon. With so much money at stake, that prompts one to think carefully.
This side view shows how much roof length the framed panel takes up.
Or, mount four rigid, framed panels for 400 watts of rooftop power, supplementing it later with an additional 200 watts of ground panels to be deployed when necessary. The attraction of this approach is that framed panels just work. There’s very little to go wrong with them, and they shrug off things like branch scrapes. In a basic 400-watt setup, I already have two monocrystalline panels, so two more would cost $270. Mounts add to that, plus any extensions of wiring. The downside of framed panels in this application is their weight at 17 pounds each, or 68 total. That’s a concern over time and miles, since the sometimes violent up and down movements created by dirt roads can warp a poorly-loaded 12-foot roof. Placing the panels near the corners of the Grandby’s roof should cut the stress considerably, since the lift struts attach there. The only remaining risk of damage is to me! Would I be able to reliably lift a roof with four panels permanently attached?
Once an appropriate time arrived to test the feasibility of this approach, I taped up the face side of each framed panel to protect the roof’s baked enamel paint, and heaved them both up onto the Grandby’s roof as shown in the first photo above. A few pieces of Gorilla tape to prevent sliding off during the tilt, and that was that. The recommended lift sequence of the Grandby is to push up the rear and then the front, which has less leverage. The rear was its usual no problem, but the front showed a decided new reluctance to rise. Since I’ve been slacking off on bench-pressing small European cars as a hobby, I found that putting my shoulder to the roof’s forward push-board got the job done in a repeatable way. That was a relief, to find out that a limited number of framed panels would be do-able rather than require a heroic effort. Adding the remaining two panels to the rear should make the front lift still tougher, but only a little bit. There will likely be room left at the roof’s center for a fifth framed panel, but I’d prefer to avoid that load without some field experience.
Raising the roof with two solar panels at the front end is potentially significant.
Adding two ground panels would up the ante, not just for the additional $270 panel cost, but for an additional solar controller. That’s the “Oh” part. I found that wiring roof and ground panels in one big series loop would be easy, since my existing MPPT controller can handle the combined 114 volts and 600 watts. But shading the roof while laying the ground panels in the sun collapses output from the entire panel grid. Why use series connections at all, then? In the case of Renogy 100W panels, the wires heading for the solar charge controller carry just over 5 amps at most, which allows the use of inexpensive 10AWG wire for most installations. (amps x volts = watts.) High voltage is no issue for wires, but high amperage is, requiring much thicker wire gauges and shorter runs of it. High voltage is a no-no for PWM (Pulse Width Modulated) controllers however, so a more expensive MPPT (Maximum Power Point Tracking) controller is required in order to convert the “excessive” array voltage into charging amps to the battery.
This diagram from Solbian shows panels wired in series to a single controller. The panels are connected in much the same way as the D-cell batteries are stacked in an old-style flashlight. In the case of Renogy panels, the controller here would be receiving up to 38 volts and 200 watts from two panels, which works out to 5.26 amps – not a big deal for 10AWG wires to carry.
Sticking with purely parallel connections would work much better if I add ground panels to a roof system (so I could park in the cool shade), but also require tearing out the roof’s solar wiring to replace it with heavy cable and an upsized weatherproof connector. Not the best. With six panels total, the combined 600 watts comes through the solar wires at just 19 volts at best, which calculates to nearly 32 amps. That much amperage requires thick and expensive cable in very short runs. Otherwise, much of the panel output is lost to heat and big voltage drops before it ever sees the controller. Panel voltages and currents need to match, too.
In this parallel connection, the panels are both connected to the controller in the same way, positive to positive and negative to negative. The controller receives up to 19 volts and 200 watts, which equates to 10.5 amps here. (Only panels with blocking diodes should be wired parallel like this.)
At the moment, the wiring for the roof and ground ports join inside the Grandby’s cabinetry before leading to the battery. The smart way to go with ground panels here seems to be to alter the camper’s solar wiring to separate the roof from ground panel wiring, making two distinct wire pairs that each lead to their own controller, and then finally join at the battery. Thus if one set or other of the panels is taking a shade hit (or suffers a component failure), the system that’s still having sun doesn’t know and doesn’t care. It just marches resolutely on by itself. Since each subsystem is independent of the other, each can be wired with its own panels in series, in order to keep the wiring affordable and efficient, or in parallel, when an array is relatively weak.
Two solar panel arrays that may be in two different exposure conditions are best provided with their own controllers. Each of those arrays can be wired in a different way, if needed.
The use of framed panels greatly simplifies screwless mounting. Here, the options range from extruded aluminum stands with Mary A’s suggestion of 3M VHB tape, to prefab aluminum or ABS mounts.
This Instapark strip sticks to the roof with VHB tape, and the solar panel can be attached with tape or screws. One of three types, this one is 20.5″ long!
I found the Instapark mount above interesting, since it is almost as long as the panel is wide, and allows very easy external fastener access to remove the panel. Short and corner versions are also available. An extruded aluminum L-shape could do much the same thing, functionally. As with all adhesive mounts, the adhesion on your typical sheet aluminum roof may be compromised by the fact that the sheet rarely has full support underneath, or is generally not fastened in any way to what is underneath it. That may potentially allow the aluminum sheet to not allow full mounting pressure, or to point-stress the adhesive strip, particularly when the roof surface is very hot. I don’t see this as a likely practical issue in general use, particularly when the VHB tape contact area is generous.
Here’s a view from the driver’s side rear, showing that the panel must not be allowed to shroud the solar panel roof connector, which is the bump toward the right side. This panel really needs to move toward the roof vent, and forward to get away from the connector.
And this shot from the passenger side shows that the panel can’t afford to be moved forward very much. Taper near the roof edges and aerodynamic pressure at speed will test the mettle of any panel mounted with adhesive. These are monocrystalline panels, which are the only type narrow enough to fit in this position.
The tan crossbars you can see front and back are the only surfaces usable for shoving the roof up. Absentmindedly pushing on the ceiling itself can puncture it.
Strolling Amok 
When you’re ready to move on to the next toy, I want to be first in line for the Granby.
Uh! Too late! It shall likely be a treasured heirloom of the vast Begley holdings upon my passing, a fact not ignored by my esteemed son after he found the standing interior headroom to his liking months before I even ordered it. I suspect he already has his first campsite in mind for a family camp-out. 😉
Hmmm.. totally new approach. Looks like it’ll work.
Merry Christmas!
Well, sometimes when you explore the best option, it turns into a convoluted snakepit of issues. I decided to forgo the rather expensive gamble and back up to live within the simple limitations of Old School. Enjoy your Christmas, Mary!
Hey Doug, first Merry Xmas!
Second I am in Q now with my slide in camper. I have a 245 watt 36 volt panel with a MPPT controller and a 1500 watt inverter. I microwave my oatmeal in the AM and use my drip coffee maker plus lights and water pump. Plus TV and signal booster with no ptoblem. Bake potato too.
Holler if u have questions.
Bill n Mic plus Chips
Thank you Bill, and enjoy some reasonably warm holidays!
Yeah, I’ve got a question, Bill – what vendor did you get your solar panel from?
We used 3M VHB tape for all our solar panels. My conversion guys were concerned about doing that on my most recent van until I assured them that’s what AM Solar did on our previous rigs. If a place with the reputation AM Solar has does it, you can trust it.
True enough. My own voucher is that it’s used to hold semi trailer boxes together, and autos have used it for decades on spoilers and such. Thanks for seconding Mary.
Maybe you mentioned this and I missed it (I did get slightly confused about the Grandby existing wiring), but another option might be to run two series/parallel pairs (one series pair on the roof, and one series pair on the ground). This way you could run one controller, but shade on the roof pair wouldn’t “kill” the ground pair since the two pairs would be in parallel. Of course maybe for some other reason you prefer two smaller controllers.
Your confusion is understandable, Luna. I sometimes blanch at the prospect of expanding things to make them clear, versus just relating a mention of fact. It’s called being lazy. If it helps, the Grandby is prewired such that one pair of wires drop from the roof, while another pair come in from a connector in the rear wall. They meet in a cabinet and are connected together in parallel, pos to pos and neg to neg, before a single set then drops down to the battery compartment in the hope of finding a controller there. So, the wiring is expecting to see one 12V panel on the roof and one 12V panel on the ground. Your idea is to simply replace each single panel with an array of them, as/if needed to get more power.
I got your idea and wrote to Morningstar a week ago with that very approach. It was plain that the responder was an electronics engineer when the first thing he typed was a recommendation that I find a local solar designer. Hint, hint. Then, “…the controller will see parallel arrays as one array if they are connected as such. There are lots of implications here that are most easily summed up as you would want the separate arrays to be as closely matched in voltage and current as possible due to performance issues relating to a possible imbalance. If that is not practical then perhaps using a second controller makes better sense.” Solbian mentions similar to FWC’s factory dual-panel setup as an alternative “where installing one controller per panel is not possible”. There are apparently some performance compromises with connecting mismatched panels to a controller in parallel, though it is commonly done by RVers. Solbian claims that blocking diodes come heavily into play in these setups, and that there will be up to a 0.7V drop across the diodes. Not being an EE, I’m not sure what that’s about, but I do know from past experience that diodes are not especially reliable.
So, as a result, I’m reluctant to connect up two arrays in parallel based on the factory wiring, since one array would output twice of whatever than the other. It certainly can be done, but I’ll already be compromising output from sticking with horizontal roof panels, and I tend to take Morningstar’s engineering advice seriously. That’s probably my former career leaking out – I’ve been conditioned to take to heart the advice of the people who designed whatever gizmo, when I’m planning to build it into whatever I’m creating. Tends to ward off unmet expectations and little mystery issues once it’s done. Just a personal choice based on a lack of interest in learning EE and a preference to do this just once. Others might choose differently.
I see what you are saying now, I think, which is that you can only have the roof pair in parallel, since the roof is pre-wired that way (did I get that right). In which case you’d have one set of two individual panels (roof) and one series pair (ground, so you could have a nice long cable with reduced voltage drop). And yeah, not so good as you’d have one “panel” (roof two) at around 12 amps and 18 volts, and then one “panel” (ground two) at around 6 amps and 36 volts. From what I understand that would drag the higher voltage pair down to the lower voltage, so what’s the point.
I have a similar situation in that I’m running a pair of 100 watt ground panels in series (to a Morningstar Sunsaver MPPT 15) and will be adding two roof panels. But, I can wire those in series, so I’ll either add another Sunsaver 15, or run both pairs (as series pairs coming together in parallel) on one larger controller.
I used to have a pop-top camper so I feel your pain (I could barely lift it empty, so never put solar on it). On my current (fixed top) RV, I had a grand plan for around 500 watts on the roof. But for various reasons, wanted to set out last spring before I was able to make certain decisions about them, hence had nothing on the roof. Lived by candlelight, just about. So I decided to set up a temporary system with two of the 100 watt “lightweight” (so-called flexible) panels in series on the ground (25′ lead) and the little Sunsaver. I was really surprised at this, but they basically took care of all my needs (mainly laptop, other gadgets, and a 50-liter compressor type cooler (Waeco, but basically like an Engel). Plus I was parking in the shade while my camping buddy roasted in the sun (roof panels).
Well, this caused me to totally re-evaluate. First question: What would happen in winter? Well, I’m parked in a spot where the sun goes down (behind a ridge) at 2:30 p.m. and I’m still doing fine. Not as “fat” as summer, but fine. Of course it is sunny 5 days out of 7, but short days. I think it must be because I can angle them perfectly, and I made an effort to stay below 2% voltage drop. I tend to use around 35-40 amp hours per day, but have taken in as much as 60 (but of course it won’t “come in” if I don’t use it – I do try to charge “all the things” when float rolls around and power is “free.”
Still, I want roof panels. Reason is that if I only stop for a night or two (or one night that turns into two), or I’m parked someplace where it is awkward or undesirable to put out the panels, then I have nothing. Then I can get into a “hole” that isn’t as easy to climb out of with only two panels. (And one broke, so for a while I only had one!). The lowest I’ve gotten is around 75 amp hours “down” (when I had only the one panel), and it didn’t take long to climb back out once I had two panels again. But anyway, so yeah, roof panels would make it so that oftentimes I wouldn’t even have to put the ground panels out, but I could. I am somewhat addicted to shade now in hot weather.
BTW, I would agree that the “lightweight” panels are fairly fragile. I baby mine (always leave air under them when on the ground, travel with them in their original boxes, upright, behind the driver’s seat), and I have had one failure, and one “damage thing.” The failure was that on one, the diode box melted (replaced under warranty). On the other one, I think it might have flopped over once (was already on ground, so just lightly tipped over) and there are four or five holes in the top plastic now. Still works, but for how long? They also do cup slightly (although not enough to capture dirt yet), and if you bed them at all when they are hot they “crackle,” which really can’t be doing anything good to them, I don’t think. I handle them VERY carefully. I also found out (when my one died I looked around) that they are nearly impossible to buy now. Grape apparently can’t get them due to something going in China; and Renogy is supposed to be re-vamping them. I still like them because I couldn’t stow glass panels where I carry these, but I will be surprised if I get two years out of them (They are around 8 months old now.) I will be putting glass/aluminum panels on the roof (now thinking 200-270 watts on roof, plus the 200 in ground panels.) I’ll do them as two series pairs, coming together in parallel, so either another 15 amp controller or one larger one and move the 15 to another project.
Okay, that was long! Anyway, I now understand why you weren’t going to just do two series pairs (at least not without some painful re-wiring, if that’s even possible).
I enjoy your posts – especially the “long-winded’ ones. Packed full of info and well told stories 🙂
Oops, make that “if I BEND them.” Not bedding my panels anytime soon!
Hey, I try not to judge… 😉
Interesting, Luna. I appreciate finding out the panel wattage you’re using, and how you’re deploying them. Also to find out first-hand confirmation of the fragility of semi-flex panels. FWC has made access to the 3-way joint between roof solar, back wall solar, and controller leads very open and easy, even though it’s inside a cabinet. So modifying it is easy, although crimp connectors aren’t as reliable as the new weird ones they’re using. Glad you get something out of by bloated and wandering writing style!
Interesting and educational. Thank you
Great to hear that you figured it out. No screws, no holes and you should do well with 400 watts. Even if you do end up with ground panels at times it will be nice to use the rig without having to do all the stuff you had to deal with on the intrepid ALL the time.
The whole rig is looking good. I like the fit to the truck box. Very clean looking.
Actually Ron, I started reconsidering the inconvenience of ground panels once I raised the Grandby’s roof on a sunny 70-degree day, windows closed. It instantly went from comfy to a little too warm because of canvas heat. So I “experienced” why the upper window openings are so large, why the Fantastic Fan is so popular, and why parking in available shade might be an attractive option when available.
It looks good to me too, but I’m biased. Maybe I won’t hang that big washtub out back after all. 😉