Best small MPPT?

[Dsanduril]

Originally Posted by Jon Hacking

...
I wish he'd tested what happens when you shade a single cell of a panel. Many folks have said that this shuts the whole panel down (or similar) but I can't see that happening. I'd think it would drop the panel voltage (by 1.2v/shaded cell) but the current shouldn't be affected significantly, but I haven't tested it as my panels have always been wired in parallel (a situation I'd like to repair).

Unfortunately, it's not the voltage that suffers when a cell is shaded, it is the current. The cells are wired in series in a module, and so each cell has to pass the same current. Once a cell is shaded it can't move electrons, and so current stops. Think of it as a current choke. That reduces the current produced by other cells even if they are in full sun. So, in short, one shaded cell can cut the output of a string to essentially zero. [This is one reason manufacturers go to so much trouble to match cells, the lowest current cell drives the whole string/channel. It isn't too hard to find the same modules rated 300, 305, 310W - all identical, but the cell matches lead to a slightly different rating]

Most higher voltage modules are series or series/parallel with multiple strings and bypass diodes. So with only one shaded cell it is possible to get decent output from a module. It all depends on the module configuration. For instance, a three-string 40V module with one shaded cell should still produce full current at ~27V as the string with the shaded cell is bypassed. With lower voltage modules (21V and less) the loss of one-third of the voltage usually drops you below battery voltage and thus all output is effectively lost.

[alctel]

Originally Posted by Dsanduril

Posting mostly so I see what responses you get, as I think you’re after the same unicorn I’d like to find.

Since every thread here and elsewhere recommends a single MPPT controller per panel I think the market is ripe for a “dumb” MPPT module that handles the solar tracking and DC-DC conversion but knows nothing about the batteries. Multiple modules would be combined with a single battery monitor/charge controller that tells the MPPT modules what to do, tracks Amps/Volts in/out and provides a single place to program charging profiles, etc.

Any patent attorneys out there?

I was JUST thinking yesterday I'd kill for one of those.

I have 3 small MPPT modules (two victrons and a tracer (which yesterday exploded, so I'm replacing with a victron) and none of them talk to each other

[john61ct]

Originally Posted by Jon Hacking

One annoying thing I'm noticing is that NONE of these units measure the current into the battery. Therefore, they can't know the state of charge of the battery. They all seem to be voltage &/or time driven. This is OK (though not optimal) for our LiFePO4 batteries, but lead/acid batteries (all types) really need that input, so they can transition to Float at the correct time. Our ancient (c2001) Blue Sky SB50 has this input (I actually bought it when the company was still RV Power Products) but I think it's been ripped out of their newer offerings, as a friend has an SB30 & it doesn't have this feature.

There are a few that do, In another thread recently I listed them all,

but they all require an add-on battery monitor or "remote" making the setup very expensive.

Victron used to with their BMV, but removed it from the firmware.

By calibrating AHT against trailing amps easy enough to get close enough.

Not needed at all with LFP, since you don't want to get that full anyway, just set AHT to the minimum.

[john61ct]

Originally Posted by Jon Hacking

The Victron is small enough that we'd have to get 2, & not having an external display is knocking it out of our personal running, despite its features & heritage.

Both of those are not actually problems if you do some research beyond watching YouTubers.

Very large capacities and external displays are readily available from Victron.

> We're still looking, but at present the front runner for us is one of the 2 that Will calls "rebranded" (SRNE ML2430 or ML4860

Do not sacrifice quality. Besides Victron, stick to Morningstar, Blue Sky, Midnight and Outback.

People living in their van may feel they need to go for cheap chinese, but really the price difference amounts to a couple nice meals out.

[MikeFergie]

You get a plug in display screen now for the victron mppt controllers. I prefer just to use the Bluetooth app. I see you also get (victron) Bluetooth voltage and temperature sensors now to attach to your batteries . On the app you can now set up a victron smart ve network that feeds this info to your mppt controller to assist it in being more accurate. I have fitted these to my LiFePo4 batteries.

[Talbot]

I have the Outback Flex 60. Fully customisable.
BUT too small for your application. I run 3x185w panels.
However the Flex 60 does have a bigger brother, which has the same customisation ability. Don't know if it is big enough for your application

[Jon Hacking]

Originally Posted by Dsanduril

Unfortunately, it's not the voltage that suffers when a cell is shaded, it is the current. The cells are wired in series in a module, and so each cell has to pass the same current. Once a cell is shaded it can't move electrons, and so current stops. Think of it as a current choke. That reduces the current produced by other cells even if they are in full sun. So, in short, one shaded cell can cut the output of a string to essentially zero. [This is one reason manufacturers go to so much trouble to match cells, the lowest current cell drives the whole string/channel. It isn't too hard to find the same modules rated 300, 305, 310W - all identical, but the cell matches lead to a slightly different rating]

Most higher voltage modules are series or series/parallel with multiple strings and bypass diodes. So with only one shaded cell it is possible to get decent output from a module. It all depends on the module configuration. For instance, a three-string 40V module with one shaded cell should still produce full current at ~27V as the string with the shaded cell is bypassed. With lower voltage modules (21V and less) the loss of one-third of the voltage usually drops you below battery voltage and thus all output is effectively lost.

Dsanduril, I'm looking forward to testing this. My current panels are all wired in parallel, so I can't yet. I would have thought that a shaded cell turned into a (somewhat leaky) diode, but continued to let current through, since it's really just a p/n junction. If it really chokes off all current, that would make shading a huge issue. The panels I've seen seem to have all their cells wired in series (except for some very strange ones using partial cells) with each cell producing about 0.5v, so a 36-cell panel has it's MPPV at about 18v. But maybe there are some internal connections I haven't seen.

[carlwk3c]

Originally Posted by john61ct

The shunt based measurement of bank acceptance does not need to be performed by the SC itself.

A Battery Monitor that is more accurate at guesstimating bank SoC is IMO the better tool.

The latter can cut off the former's connection if there isn't a more elegant signaling scheme in common.

Or you just calibrate AHT on the SC so that "most cycles" your endAmps goal is reached before dropping to Float,

getting there every cycle is not required for great longevity.

I would start out looking at Victron's systems, SmartSolar 100/30 per panel, BMV-712 per bank, perhaps making use of its relay, as a failsafe to prevent too long AHT

what chemistry is the bank?

And Color Control GX, Venus or your own MCU based system using their (basically) open protocols as a central system logger / controller.

The Victron user forums are well supported, sometimes their founding engineer stepping in to participate when needed.

Like he said. The Victrons can be networked together with the Battery monitor (and optionally their inverter/charger) to make a system that plays well together and can be monitored and controlled from your smartphone, tablet, and/or notebook via Bluetooth. The above Venus GX adds further integration and remote monitoring and control over the internet. (Plus the Venus accepts up to 3 tank level senders and 2 temperature sensors and puts that data on your network as well.)

[Dsanduril]

Originally Posted by Jon Hacking

Dsanduril, I'm looking forward to testing this. My current panels are all wired in parallel, so I can't yet. I would have thought that a shaded cell turned into a (somewhat leaky) diode, but continued to let current through, since it's really just a p/n junction. If it really chokes off all current, that would make shading a huge issue. The panels I've seen seem to have all their cells wired in series (except for some very strange ones using partial cells) with each cell producing about 0.5v, so a 36-cell panel has it's MPPV at about 18v. But maybe there are some internal connections I haven't seen.

Most modules are a single series string (just as you describe), but if you look in the connection box you will frequently find bypass diodes that break the module into two (36-cell modules), three (60-cell and 72-cell), or four (96-cell) smaller strings. A 36-cell module that loses half its capacity from shading no longer puts out enough voltage to charge a battery so the bypass diode is not truly useful (except possibly from a safety perspective) unless the module is connected in series with other modules.



One curiosity that comes out of this is that you can shade one cell, or two, or three in the same string and the effect is the same - you lose the output from that string. But you can also shade only one cell in each string and lose the whole module output. It all depends on whether or not the shaded cells share a string, or are in different strings.

I've been talking full-cell shading, if you partially shade (hard shadow) you generally lose current in relation to the percentage of the cell that is shaded, but this can depend in orientation of the shadow compared with direction of current flow in the cell. Soft shadows become even more complicated.

The reason that bypass diodes showed up in the first place is that a fully shaded cell actually consumes current from its unshaded brethren. If that goes on for long enough the heat generated can cause a fire. So bypass diodes pull an underperforming cell (string) from the circuit, eliminating the current feeding the shaded cell. In theory we would bypass each cell, but that adds complexity and cost, the industry has settled on ~18-24 cells as a bypass unit.

https://www.electronics-tutorials.ws...ss-diodes.html

[Dsanduril]

John (and everyone),

there's another thread running here on CF about multiple MPPT controllers. On that thread Justin Cook, a Victron supplier just posted about a firmware update (not yet released, but soon) that will allow exactly what was discussed up-thread - using a BMV to control/sync multiple small MPPTs. To me that puts them in the leader spot - if you're willing to spend the money.

Currently, 1.47 is only available through VRM update or direct load-in, but will be available through VictronConnect within the next week or so.
Synchronized charging will require that the controllers be on a VE.Smart network together (over bluetooth), so to make it work you will either need a BMV-712 or a Smart BatterySense to create the network, then put all the controllers onto that network and you're basically good to go; I was part of the initial round of pre-release beta testing and it appeared to work quite automatically - any controllers on the network together automatically synchronize charging now

[Wolfgang.Schau]

I am using TS-60 and TSM from Morningstar. I have 3 350W panels in series and it works just fine. Since they are mounted on my catamaran on the davits, shading is not an issue. I can adjust the inclination of the panels to get the best sun coverage. However, they either all point right or all wrong when sailing south. They produce so much power, that I can even run a small A/C system with it.
If you have mounted your panels on the ceiling, shading from your sails will be an issue. However, panels have diodes and if one panel is shaded, the others still work. A separate MMPT for each panel will not help to take the shade away. If the panels have a different inclination, a separate controller per panel will definitely help to squeeze the most out of your panels by using the optimum operating point, besides the redundancy another advantage and the TSM can monitor multiple controllers with the use of a hub.

A word to "in series vs in parallel", the major difference is the higher current for the parallel configuration and thus requiring thicker cables from the panel to the controller, adding weight and power loss along the wires if not thick enough. Second, each MPPT has its best performance input voltage. Usually the higher input voltage makes the MPPT more efficient, too. A parallel configuration has the same effect than the diode, if one panel is shaded, it will not effect the other panels. The only advantage of a paralle connection is - if you have a corroded connection, you may lose only one panel and not all three - which happened to me.
Bottom line, if all panels are mounted flat like on a roof and shading is not an issue, a serious configuration and a single MPPT will do the job. If they are mounted differently, individual MPPTs may get up 0 - 15% more energy.

[foggysail]

Originally Posted by noelex 77

The Victron range includes some quite large controllers. Even the Smartsolar series has models which will handle up to at least 100A.

So although many prefer multiple controllers, a single controller can be chosen for even a large array if you prefer.

There are also many display options if you do not like the bluetooth connection. The SmartSolar Control Display or the Victron MPPT control display are only a couple. Personally, I use a Pico battery monitor that shows the output of my three panels on the main display. There are other cheaper third party products that can monitor solar output.

Yes to a single controller used with panels configured for the highest acceptable voltage. Higher voltages allow easier wiring. Single controllers minimize both wiring and hardware cluster gutchas especially in limited available space.

[foggysail]

Originally Posted by Wolfgang.Schau

I am using TS-60 and TSM from Morningstar. I have 3 350W panels in series and it works just fine. Since they are mounted on my catamaran on the davits, shading is not an issue. I can adjust the inclination of the panels to get the best sun coverage. However, they either all point right or all wrong when sailing south. They produce so much power, that I can even run a small A/C system with it.
If you have mounted your panels on the ceiling, shading from your sails will be an issue. However, panels have diodes and if one panel is shaded, the others still work. A separate MMPT for each panel will not help to take the shade away. If the panels have a different inclination, a separate controller per panel will definitely help to squeeze the most out of your panels by using the optimum operating point, besides the redundancy another advantage and the TSM can monitor multiple controllers with the use of a hub.

A word to "in serious vs in parallel", they major difference is the higher current for the parallel configuration and thus requiring thicker cables from the panel to the controller, adding weight and power loss along the wires if not thick enough. Second, each MPPT has its best performance input voltage. Usually the higher input voltage makes the MPPT more efficient, too. A parallel configuration has the same effect than the diode, if one panel is shaded, it will not effect the other panels. The only advantage of a paralle connection is - if you have a corroded connection, you may lose only one panel and not all three - which happened to me.
Bottom line, if all panels are mounted flat like on a roof and shading is not an issue, a serious configuration and a single MPPT will do the job. If they are mounted differently, individual MPPTs may get up 0 - 15% more energy.

ARE YOU SURE ABOUT THAT???

I have not had interest to measure paralleled panel’s output voltages under mixed shaded conditions. Solar panels are current sources, not voltage sources so if paralleled panels have different output voltage caused by shade, I believe the combined output voltage will be lower resulting in a lower Vmp to the controller lowering the delivered power.

[Dsanduril]

Originally Posted by Wolfgang.Schau

...Usually the higher input voltage makes the MPPT more efficient, too...

This has been stated here several times, with zero evidence. Every technical report I have seen shows that a DC-DC conversion is most efficient when the input voltage is closest to the output voltage. Think of it this way; if you have 20 electrons at 48V you will need to distribute that energy to 80 electrons at 12V. If your input is at 24V instead then you’ll have 40 electrons to provide the same energy transfer, and thus there is less work in splitting up their energy. There’s a pretty good paper here. If you don’t want to wade through the deep stuff there are graphs on the last couple of pages showing the changes in efficiency with changes to Vin/Vout. At 4:1 efficiency drops to about 90% while at 1:1 it is about 98% (showing that there are losses in the circuit even when it does nothing).

However, the DC-DC conversion isn’t everything, the input wiring needs to change (as you noted) to accommodate a higher input current. As a system a higher input voltage may lead to lower overall losses, but this is a question of system design, not the MPPT itself.

[Jon Hacking]

Originally Posted by Dsanduril

... Every technical report I have seen shows that a DC-DC conversion is most efficient when the input voltage is closest to the output voltage. ... At 4:1 efficiency drops to about 90% while at 1:1 it is about 98% ...

However, the DC-DC conversion isn’t everything, the input wiring needs to change ... to accommodate a higher input current. As a system a higher input voltage may lead to lower overall losses, but this is a question of system design, not the MPPT itself.

Dsanduril, Thank you very much for your several contributions above. Good stuff.

Conversion efficiency drops a bit at higher voltages (which surprised me as well) but not very much. It's a 2nd or 3rd order issue compared to shading.

If shading really does throttle the whole series string (a concept I'm having to come to terms with) then multiple bypass diodes, even 1/cell, sound like a really good idea. But that depends on the panel having enough voltage for the DC/DC even after it looses a string (or 2 or 3). So a series connection of panels still makes a lot of sense, if your panels have multiple bypass diodes. (Now I see why my 12v 36-cell panels have bypass diodes - it's for when the panels are connected in series, as those diodes don't make much sense in parallel, or for stand-alone panels.)

It's becoming apparent that the answer to my original question is probably Victron (depending on your definition of "best"), but you have to buy into their entire line, which gets very expensive. They still don't have all the features I'd like, & their communications between modules (over Bluetooth?!?) seems sub-optimal as well. Maybe they'll improve that.

The designer of the VSR & WS500 alternator controllers (best on the market, IMHO) is thinking about designing small MPPTs for 1-2 panels with distributed BMS, voltage sensing, battery current sensing, & display, all communicating over CAN-bus, so perhaps this is the wave of the future.