Reviewed under their "Things that Work" banner, Perez and Schwartz practically gush over this unit:
Every so often a product comes along that significantly changes the RE industry. The Solar Boost 50 is one of those products...They conclude that everyone can expect a dramatic, almost magical increase in output power from their PV arrays, and that getting one of these units is a no-brainer. Unfortunately, this just isn't true.
As I described in detail in my Notes on Peak Power Tracking, how much one can gain from a peak power tracker depends on several factors. The most important are ambient temperature, and whether you float your batteries (e.g., in a grid-tied system) or cycle them daily (e.g., in off-grid use).
MPPTs give their most impressive gains in cold weather, because that raises the PV array's peak power voltage point well above the usual battery operating voltage. But in warm weather, the peak power voltage and the maximum available array power both decrease substantially. This brings the optimum array operating voltage closer to the battery voltage, decreasing the gains to be had with a MPPT.
MPPTs are also more advantageous in an off-grid system where the batteries are cycled every night than in a grid-tied system where the batteries are floated fully charged, waiting for a grid failure. When charging begins each morning in an off-grid system, the battery voltage will be well below the float voltage, dragging the array operating voltage (without a MPPT) further below the peak power point and decreasing the array output power. A MPPT allows the array to operate at its optimum voltage, independent of the battery voltage.
By testing my own array of 12 Astropower AP1206s, I found that a MPPT would yield perhaps an additional 30W in my grid-tied system here in mild San Diego. This is a far cry from the impressive gains reported by Home Power, and not worth the extra cost of the unit.
So my first critique of the HP review is the lack of any testing under ambient temperatures higher than 61.7F. They did most of their testing at temperatures in the 40's and 50s. Much of the southern and southwestern US is warmer than this even on winter days, and there certainly aren't many parts of the country that stay this cool in the summer!
My second and larger critique has to do with their test method. The Solar Boost 50 unit has a built-in digital ammeter that can show (with a selector switch) array (input) current and output current. Perez and Schwartz take the difference between output and input current as a direct measurement of the extra power yielded by the use of MPPT.
For example, the first entry in their data sheet shows a PV voltage of 34.4V, a battery voltage of 25.65V, an array input current of 12.2A and an output current of 15.8A. From this they compute an "amps boost" of 3.6A and a "watts boost" of 92.3W. The latter figure is apparently computed by multiplying the "amps boost" of 3.6A by the battery voltage of 25.65V.
Unfortunately, this computation is completely wrong, so the results are meaningless!
The basic problem is that the reviewers assumed their PV array was a perfect constant-current source, i.e., it would deliver exactly 12.2A whether it operated into a 34.4V load (as it did here) or into a 25.65V load. While real PV arrays are good approximations of constant-current sources when operated well below their peak-power points (e.g., short-circuited), this is not true at higher voltages, especially near the peak-power point. Had they operated their array with a conventional charge controller, the array would have been forced to operate at the battery voltage of 25.65V, and it would have produced more than 12.2A at this lower operating voltage.
How much more current we cannot say, because it depends on the intrinsic properties of their particular PV panels, their manufacturing variations, and their precise operating conditions. All we can say is that the numbers in the table consistently overstate the MPPT boost, although we don't know by how much. The only way to find out is to actually operate the array at both voltages and measure the change in output current. This difference must be subtracted from the "amps boost" given by the ammeters in the Solar Boost 50 to give the true increase in usable power yielded by peak power tracking.
One way would be to switch the PV array between the Solar Boost 50 and a conventional PV controller, both feeding the same battery bank, and measure the change in power delivered to the battery. There are undoubtedly conditions where the MPPT would produce more power than the conventional controller, i.e., cold temperatures and/or low battery voltages. And there may well be conditions when it would produce less power than a conventional controller due to the losses in the MPPT itself when the battery voltage is already close to the array peak power point, e.g., when the battery is full and the panels are warm.
Unfortunately, this review does not correctly quantify these issues.
Phil Karn
24 July 2000
Updated 11 Aug 2000