I know the easiest & "practical" solution would be to upsize my system with more batteries & stronger charge controller. But I'm trying to think alternative methods. My shed load needs only a sustained 10 amps during daytime. I don't want to purchase a larger system which dumps a lot of energy into battery storage during sunny hours, then the load chews away at the batteries during cloudy periods, awaiting a recharge when the sun shines brightly again.
I'm exploring alternatives...
Even if my attempts are futile, it's a learning experience how PV panels behave under different loads. And yes, I understand the concept of hitting that sweet-spot where the PV output performance curves of amperage vs voltage create the highest power output. (The basic concept of MPPT). I can't just hook up light bulbs at random & expect the PV to put out max power....

 

You could try extending the usable day light by series connection of the panels at dawn/dusk.

 

Too bad. I like doing pirate grid tie Inverter design. I did write the "Basic GTI design for dummies' book on it after all.



I see. What I would be looking at is setting the systems up so that once the panel voltage goes above whatever it normally does when the MPPT unit is running at full draw it starts dumping the excess power above that into your secondary load.

Simple voltage limit based load dump transfer of sorts. Without knowing your level of skill that dumping systems could be a simple analog voltage clamping circuit (considerable amount of heat to be lost to the switching device doing the load transfer) or a more complex PWM based one. (little losses in the load transfer switching device.)

You have lots of options and ways to go about it depending on what you main goal is.

Thanks for your suggestions. My level of skill: Novice/Low-Intermediate. I understand basic electricity concepts. I do lots of vehicle electrical diagnostics and read wiring diagrams as an automotive technician & I have a good understanding of current flow, voltage drops, divider circuits, parallel vs series voltages & currents..... kind of basic stuff. I've done oscilloscope measurements & captured/analyzed circuit waveforms. Give me a datasheet about a simple IC and I can figure out what the pins do & how to hook it up to fit my needs. I'm proficient in Algebra + most Geometry & Trig. I finished Calculus Level 2 in high school with a "B" average. ...not bragging, just trying to give you an idea of my education history. However, I haven't used Calculus in 30 years. I'd have to re-learn Calc from scratch. Hand me a complex circuit diagram and I could offer educated guesses on what's going on.....it'll just take me a while.

PWM Scenario-Question: Say the sun is shining bright, I've got the two PV panels hooked in parallel offering the potential of 20 amps, the branch dump circuit is attached between the panels & MPPT, and finally the batteries need charging because they are low. For simplicity, say the PWM is at 50% duty cycle. The MPPT would recognize the batteries are low on charge. Wouldn't the MPPT "see" and try to process 20 amps half the time? Would be bad because it's rated for only 10 amps. I don't want to burn it up; the Genasun GV-10 cost me $110. According to the specs and my own measurements, it's a true high-efficiency MPPT. Not a cheapy knock-off No-Name brand.

Main goal is to maximize the length of time (during day) & provide a steady 9 amp power into my shed load. When it's heavy overcast (noontime), my MPPT provides about 4 amps to the battery bank. If I had 2 panels in parallel, that number could rise to 8 amps. But when the sun shines bright again, I don't want 20 amps burning up the MPPT. Send that potential energy to a branch circuit to heat some water. Furthermore, by having two PV panels, power production could start earlier and end later in daytime vs if I had only one PV panel connected. RONV offered a wiring schematic; I'll study that for a while. Thanks for your input !!

 

Not as bad as I thought.
View attachment 123076

Thank you very much; pls allow me some time to study your electrical diagram & predict how it behaves under various conditions of PV panel output and my battery bank state-of-charge.

 

Are you tracking..........use the excess power to track.

 

You could try extending the usable day light by series connection of the panels at dawn/dusk.

I thought about that, but it's counterproductive to my overall goal. Say I have one primary PV panel hooked to the 10-amp MPPT doing it's regular thing. Dusk comes along, PV power output drops to (lets say) half, (so we're at 5-amps that the MPPT is processing), and the second panel connects instantaneously to supplement. (I can't connect the second panel too "early" because the combined potential energy would surpass 10 amps). Now we're back up to 10 amps until the sun sets, figuratively speaking, but what's the auxiliary panel doing the remainder of the daytime, while the primary panel is connected and the circuit is fluctuating between 5 to 9 amps? In order to make use of the secondary PV output during mid-day hours, it would be necessary to wire in a second charge controller, another bank of batteries... so on & so. Without the add'l components, it would be sitting in the sun with open-circuit voltage producing no usable energy since it's connected to nothing.

Both PV panels have identical specs. If it's slightly overcast and the MPPT is processing a steady 7 amps with the primary panel, the auxiliary panel has a 7-amp potential output at that same instant. I'm at a net negative of 2 amps at my battery bank (because load consumption is about 9 amps). The auxiliary PV potential for power output is identical to the primary one at all times during the day. If I connect too early, it'll overload the MPPT. I have to wait for the magic 5-amp mark before two PV's can be simultaneously be connected (when output drops) or disconnected (when output increases) to make full use of 10-rated amps at the MPPT. Series connect too early and input voltage will be too high. Parallel connect too early, voltage will be OK, but current will be too high.

As a side note, I closely monitor my battery bank voltages and ease up on load as needed so that the net negative 2 amps doesn't drain the bank. When I see consumption surpass supply, I shut down various loads to let batteries recover. Call it manual load balancing by human intervention. My 12-volt battery bank is rated at 250 Amp-Hours, in case you're wondering about capacity. They're deep-cycle golf-cart type, not automotive. Just info...

Maybe I'm making things too complicated, but I like exploring alternatives. Thinking outside the box offers improvement to technology when results are positive.

 

Are you tracking..........use the excess power to track.

I have amp-hour meters hooked up to both the supply (charging) side, and on the demand (load) side. Both of them display present and cumulative supply & consumption numbers. I monitor the supply and demand numbers daily (as well as battery bank voltage) and increase or decrease load as necessary to keep the battery bank at a happy charged level. I think that answers your question (?)

 

panel tracking.......the sun.

 

panel tracking.......the sun.

No, my PV panel is on a fixed, non-moving base. If you're inferring to have the panel perpendicular to the sun's rays at all times to maximize output, I'm familiar with concept. Before permanently mounting the PV, I did some power output measurements by tilting the PV's angle in relation to the sun by hand and wasn't too impressed with the numbers. Perhaps if I want to squeeze out every last milliamp from the PV's, I'll look into automated tracking. But at this time due to installation logistics, I'm better off having them on a fixed base.
They are, however, angled slightly towards the south (I live in Florida, USA). So at least during noon time, the sun's rays are near perpendicular relative to the PV..........I know.... angle of attack changes according to season, too. Let's just say I mounted them the best I could to capture the rays best they can.

 

You only have 2 choices.
1- you can shunt the excess current to your lamps.
2- you can reduce the voltage to the charge controller.
Since you don't want to use PWM that would force you into some analog design. But 20 amps at 12-18 volts is a lot of power.
Maybe you could use a digital design where you turn on ever increasing numbers of bulbs based on the current.
Say 1 ant 9.5 amps, 2 at 9.55 amps, 3 at 9.6 amps and so on.
Lot of "stuff."
If you have specs on the lamps we could have a look at the power in the analog switching device.

Believe me or not, having numerous small loads turn sequentially on and off depending on power availability was my first choice for the branch circuit. Kind of like a cascading power meter. But as each bulb turns on, it'll cause a voltage drop on the system so I need to do the hysteresis (?) calculations just right. Otherwise, the bulb will turn on & off as the voltage jumps over & under the selected thresh-hold. Not sure how the MPPT will approve of a wildly fluctuating PV panel voltage input when it's used to seeing a relatively stable voltage.
Anyway, the bulbs I'm using are regular automotive type. I hooked them up to my battery bank & got the following baseline readings on my multimeter: 13.4 volts battery voltage, and 2.8 amps current draw.

 

...furthermore, hooking up 12-volt bulbs is only a band-aid, if eventually I'll be looking for efficiency rather than a simple branch dump circuit. "Julian Llett" (youtube) has an interesting vid on how 12-volt bulbs and PV panels are a mis-match when hooked up directly. PV's put out max power around 18 volts. So, (but don't quote me), his 55-watt 12-volt bulb glowed quite dimly even with a 100 (?) watt panel in bright sunlight. I replicated his experiment with similar results. The 12-volt bulb vs the 18-volt panel is a big mis-match. There's so many dynamics... I find it fascinating to conquer !

 

...furthermore, hooking up 12-volt bulbs is only a band-aid, if eventually I'll be looking for efficiency rather than a simple branch dump circuit. Adam Welch (youtube) has an interesting vid on how 12-volt bulbs and PV panels are a mis-match when hooked up directly. PV's put out max power around 18 volts. So, (but don't quote me), his 55-watt 12-volt bulb glowed quite dimly even with a 100 (?) watt panel in bright sunlight. I replicated his experiment with similar results. The 12-volt bulb vs the 18-volt panel is a big mis-match. There's so many dynamics... I find it fascinating to conquer !

Yes, bulbs are very non linear and have no specs except on.
The heater is a better load.

 

...furthermore, hooking up 12-volt bulbs is only a band-aid, if eventually I'll be looking for efficiency rather than a simple branch dump circuit. Adam Welch (youtube) has an interesting vid on how 12-volt bulbs and PV panels are a mis-match when hooked up directly. PV's put out max power around 18 volts. So, (but don't quote me), his 55-watt 12-volt bulb glowed quite dimly even with a 100 (?) watt panel in bright sunlight. I replicated his experiment with similar results. The 12-volt bulb vs the 18-volt panel is a big mis-match. There's so many dynamics... I find it fascinating to conquer !

Yes, it's a bit of fun. https://forum.allaboutcircuits.com/...rolled-battery-array.32879/page-3#post-975215

I have a diversion power function in my system that uses a PWM signal to control the power to a dump load (heater.)

The Solar monitor system calculates real-time energy flows and redirects excess energy to the load.

CCEFF_DIFF is a variable that's calculated on the efficiency of the charge controller process to move the estimated energy from the panels to the loads or batteries. As this CCEFF moves lower (0->100%) a power offset signal is generated to control the PWM system to increase the solar panel dump load power before the charge controller reducing excess power and causing the efficiency of the charge controller movement process to increase until the control loop is re-balanced.

code fragment in C18 for pic18f8722

void pv_pwm_calc(float slope) // calc a duty-cycle from the PV power excess supply, CCEFF_DIFF is a global variable

static int16_t CCEFF_DIFF_tmp;
static float power_exp;
power_exp = PWM_EXP + slope;
CCEFF_DIFF_tmp = (int16_t) (99 - CCEFF); // power offset
if ((CHARGERL == R_ON) && !PWMTEST) { // set all PWM to zero if charger is on and not testing
pv_pwm_shutdown();
CCEFF_DIFF_tmp = 0;
}
if (CCEFF_DIFF_tmp > PWM_LIMIT) CCEFF_DIFF_tmp = PWM_LIMIT;
if (CCEFF_DIFF_tmp > PWM_SLOPE) {
CCEFF_DIFF = (int16_t) lp_filter(pow((float) CCEFF_DIFF_tmp, power_exp), LP_PWM, TRUE); // control power
} else {
CCEFF_DIFF = (int16_t) lp_filter((float) CCEFF_DIFF_tmp, LP_PWM, TRUE); // control power
}
if (CCEFF_DIFF > 100) CCEFF_DIFF = 100; // limit max value to 100% power
}