What Would You Like to See in a Charge Controller? (long post)

Van Living Forum

Help Support Van Living Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Canine

Well-known member
Joined
Dec 9, 2013
Messages
2,688
Reaction score
1
Location
Great Falls, MT
If you had the opportunity to build a charge controller, what options/features would you want? What would you not want ?

This can get incredibly complicated as some of you out there will want to talk about the pros and cons of different types of diodes. I don't want to get that detailed as that would overwhelm most of us and the laymen would get no benefit. (Hopefully, I'm not a hypocrite and don't get too detailed.) If you want to talk about that level of electrical nuance, another thread toward that end would complement this one well. Also for simplicity, this will be for 12 volt battery banks only, off grid only, and for mobile RVs/vans/trailers only. Not for houses, grid tie systems, or 24 volt batteries.

1) Can handle minimum 450 watts of solar panels.
2) Can handle minimum 80 voc (volts) from solar panels
3) Can handle minimum 40 amps from solar panels

^^^ A more simple way to describe the first three specs above is that it will be able to handle at least one of the largest, most powerful panels made today with a little room for advances in solar panel technology.

4) Has an interactive, digital panel with limited or no DIP switches. Kind of like the 3000i or the Midnight Kid.

5) One distinct menu for each menu. Such as one menu for displays- displaying inputs, outputs, amps, volts, level of charge, days from last full charge, total of amps in, total of amps out, highest voltage recorded, etc. The next menu for setting battery specs. The next menu for setting charging specs. Set up options , display options, and reset options should not be in the same menu.

6) No buttons should have multiple uses. One button as the select button only; press and hold for 3 seconds to select. One button as the next button only. One button as the back button only. One button to increase, one button to decrease. You should not have to first press the next button then press the back button while holding the next button to decrease an amount when in the set up menus. This is NOT intuitive.

7) If the buttons have a shape, they should intuitively do what they are shaped as. A right arrow button should not be a select button- it should be a next button. A square button should not be a next button (not intuitive). A left arrow button should be a back button (intuitive). Up and down arrows could be to increase or decrease values. Round or maybe square could be select. Left and right arrows could be back and forward. Probably 6 buttons total.

8) Wires should not have to bend at a right angle in order to be attached.

9) Lights should not be bright. Avoid most blues and most greens and stick with reds. When I'm sleeping, I don't want bright lights disturbing my sleep. Red is much, much, much less disturbing and preserves night vision far better.

10) The ability to be networked with other controllers. If that is the case, make it easy to get twisted pair wire.

11) A control panel with a shunt for more accurate charging. I'm not sure if it should be an option or if it should be automatically included. Probably separate due to price. When a shunt is included, twisted pair wire should be an easy option to get.

12) No fans.

13) Maybe switches to shut off the solar panels before shutting off or disconnecting the charge controller.

14) A switch to shut off the charge controller. We shouldn't have to add a switch so that we can get a nice, crisp connection. Sometimes, it's hard to get a good, crisp connection when hooking up wires on the battery.

15) Maybe have a fuse block at the connection points.

16) I'm not sure how important it is to be able to connect to a lap top. If it does connect to a lap top, it should not be a necessity to be able to make detailed changes to accommodate specific batteries' needs. I have no desire to get into the programming software and change that. Having a data log that can be printed would be cool to see. Maybe be able to put info into pie charts or graphs to see how much electricity is used throughout the year. This is not something I would want, but it could be a popular option.

17) When it is in bulk, acceptance, float, and equalization modes, a flashing light that flashes at different speeds is terribly hard to try to decipher when you aren't used to it. You should be able to easily see which mode you are in at a glance instead of looking at a blinking light then figuring out how fast it is flashing to figure out which charging mode it is in.

18) A temp sensor to the battery should be optional. No temp sensor on the controller itself.

19) $300 with no display. $350 with minimal display/options. $480 or so with advanced display/full options. This is important because if more than one controller is networked, not every controller needs a display. All that is needed is one minimal display or an advanced display for all the controllers since they would all be connected together. A minimal display would be standard and a less expensive option than an advanced display and still be able to be networked. An advanced display with full options would be optional. No display would be optional to keep costs down when networking several controllers. I would like to be able to make the advanced display fit right on the controller instead of a separate unit, but that might be asking too much to put into too small of a space.

20) Mounting should be easy. Some controllers are stupidly hard to fasten to a wall.


Anyway, you can see how detailed and lengthy this can be. Hopefully, my expectations are somewhat realistic. Try to be not like me and keep it short.
 
With regard to #13, shut off the panel switches, I would rather a design that just doesn't fry itself.  Self frying charge controllers are the result of incompetence.  No excuse.  

In my opinion actually charging the battery is important.  Getting to 100% before switching to float mode, shutting off the charge, is more important than light colors and button shapes.  You want shunts, #11, and temperature sensors, #18, to be optional.  With the controller only able to see the voltage at its end of the wire that goes to the battery do you expect the controller to know the battery state of charge?  Do you expect the controller to shut off, float mode, at the right moment, when the battery is full?
 
I'm not clear on the OP mandate here.

There is a question (what do *you* want?) then follows a list of specs (what OP wants).

Are the latter the boundaries within which we are meant to operate or just "one example" set of specs?

Also I think it's relevant to ask "why are you asking?"

Are you working with someone looking for market research, to design and build a new type of SC?

If not, maybe it would be productive to ask "which SC on the market (comes closest to) fits these specs?"

Also, asking for all the ideal bells & whistles to come under a certain price point is not IMO very realistic.
 
I've been thinking about an open source solar charge controller / display project run from raspberry pi or similar, shared on git. Abstract away the hardware details so you can tie it to as $$$ hardware or cheap as one wants.

Folks could write/edit/share MPPT algos or whatever they like. We find problems and address them.
epic configurability
data storage in something tiny like sqlite, held for x days
analysis and configuration via tiny webserver like lighttp
retrieve common default settings for particular battery types, etc
export and save settings.
wifi/BT interface for the price of a $10 generic dongle.

I was in an R/C forum where we banded together to have a custom board built. We had to wait a couple months for enough orders to complete a whole sheet (or whatever it's called) at once, but they ended up costing $3/ea or something silly. Hardware revisions were easily tracked. It was supah cool.
 
More stuff for the fantasy open source project:

Powerpoint identification could be accelerated (fewer PP tested to find the right one) in changing conditions if we saved some PPT info in another microDB:

Our panel's lab Vmp is X.
For the last few days around this time the power output we needed was around X voltage.
When power output is widely varied but changes slowly we probably are experiencing clouds. Try this.
When power output oscillates quickly but shallowly (?) we might be under waving branches. Try this other thing.

Batteries are below freezing. Don't charge Lithium.
Keep my lithium in the middle 80%

Run multiple charging circuits from the same pi "head"; effectively one charger per panel if desired. SHould be cheaper than buying separate retail units each with their own chips.

Aggregating info might help shed light on how well brands/chemistries of batteries are failing. This much load causes this much voltage drop on this kind of bank.

If someone needed help they could cut/paste troubleshooting info from their web interface into the forum (or pastebin) so we could tell what we are looking at.
 
Sorry for all the one liners.  Will move my thoughts on this to a blog post so I can update as I think of stuff.

The most recent thing I thought of was this:  if one pi controller is running X charging modules seperately, and one panel shows an increase in power at a different setting, the others should try it too and revert if no improvement.
 
Two piece unit, one (epoxy encapsulated) with the controller 'guts', located near the battery, with temp sensor there. The remote display mounted where you need it. Wired (common modular plugs) or wireless options.

Set and forget simplicity. Smart algorithms to figure things out when unattended by humans.

Under $150.

Display that dims a few hours after sunset, brighter again in daylight. Or timed dimming after key-presses.

Display would be lit green when all is good, yellow for low, red for very low. Large easily understood digits and symbols.

Internal protection against shorts, reverse polarity, no battery connected, etc.

Robust construction. Minimal RFI footprint.

Brushed aluminum panel.

American sounding name, even if made elsewhere.

Five year 'no questions asked' warranty.

All this and a bag of chips. Do-able!
 
Yes, I agree the arduino wins in terms of popularity and leanness.

I seem to remember that device is restricted to C for progrmming. The Pi would allow scripted, compiled, and interpreted languages and a huge collection of appliations. Would be easier for more folks to contribute to a joint project.
 
Trebor, I would want shunts as an option because if I buy two charge controllers, I would need only one shunt. One shunt is needed to better calculate the best charge for the battery. Without it it's not as accurate. The float should stay on as long as there is sun and the battery is full.

John61CT, I did want input then got the thread rolling with what I thought were good options in a charge controller. They aren't boundaries, just ideas I had. I'm not looking to build one, but want to learn more about what makes a good charge controller. I had considered asking what current charge controller is the best, but thought the results wouldn't be as good compared to the way I presented the thread. Once I learn what options are the ones I want, then I can see which controller best fits that. The goal isn't to find a CC so I can buy one, but more out of curiosity and for fun. Perhaps in the future I will upgrade, but that isn't my present plan. I don't see a problem with expecting a certain price point. I have a very general idea what a charge controller costs with most of those features and some of them come under the price point I stated. I could be off by a mile, though, since I'm not an engineer.

frater, having an open source has been proven to be a powerful tool. Many minds are more powerful than the world's most powerful computer. Retrieving battery specs would be awesome; type in the model number and the CC automatically knows the a/h, voltage, discharge rate, charge efficiency, etc. This would help people who don't want to/can't become erudite about charging batteries, but still want their expensive investment to last as long as possible. Running multiple panels individually to one controller is an interesting idea. That would save space and money compared to installing additional CCs. There is going to be more heat with more panels; enough panels will eventually create too much heat to be able to dissipate within one CC. Arduino has an amazing display. Very easy to read and understand with the icons. I didn't quite understand every point you made, but that's OK; I'm learning.

tx2sturgis, having the controller always as two separate pieces could be extremely beneficial for installing. If you have the display, you don't need all the guts in the living space. Tuck those guts down below. Keep the basic and advanced displays separate instead of making them so they have to be fastened onto the CC itself. The colors you mentioned are intuitive. Having it epoxy encapsulated would make it super tough. You see it a lot in chainsaws and other extreme conditions. Might be a fairly cheap thing to do. The internal protections you mentioned should be absolutely mandatory- no exceptions. Totally agree with the dim display at night- it's very important to me. The more I think about the brushed aluminum panel, the more I like the looks of it. Not having it powder coated would also decrease price. Perhaps not having the aluminum body coated could increase the odds of a short, though.
 
a lifetime, satisfaction guaranteed, warranty.

how hard is that?

;-)
 
I know this is an old thread.
Id like to see a solar output to run a small vent fan durring the day. If there is enough power, run the fan. Don't need the fan at night.
This is for a trailer that will be parked most of the time, but it needs an active fan durring the day to prevent an oven.
 
I looked up on my marine forums, the all-in-one recommended unit is Marinco, some say same as Nicro.

Uses a standard C cell rechargeable, need replacing every once in a while.

Not for high-CFM usage needs, just a steady rate for airing, while the space is otherwise sealed up.
 
BobW55 said:
Id like to see a solar output to run a small vent fan durring the day.  If there is enough power, run the fan.  Don't need the fan at night.

If you hook up the leads from a small panel (about 3 to 5 watts) directly to a small fan like a computer fan, it will run when sunlight hits the panel and stop running as it gets dark.

You can scale this up if you need more airflow. For a fantastic fan, you will need a much larger panel.
 
For a high CFM fan like that, might as well integrate into the House bank setup
 
Take something like a Midnight Kid or Classic, and add Remote capability with a SIM card so I can access and control it remotely via cell Data. I'm trying to do this now with my Classic, but my Verizon Jetpack appears to have gone off line and I cannot "call" it to reset it.

In addition, such an access would have the ability to control 4-5 Digital outputs to turn things on/off remotely, manually, natively and some analog and digital inputs to read other items. The Classic can do this with one output, but takes third party software to do.

A built in ethernet switch for other items would be nice.

Native control for an inverter built in.

When does it ship? :p
 
I have a 12 volt 4 watt panel running a 3" fan directly now.
Need about a 10 watt so I can run a 6" fan.
Would rather not have to add the 3rd panel just for a vent fan.
 
Top