Hardware choices - Part 2

Hardware choices - Part 2

1) Why this hybrid inverter (Deye SUN-6K-SG03LP1-EU)

I chose a hybrid inverter because I wanted three modes in one device:

  • Self-consumption (PV to house during the day)
  • Shifting with battery (PV to battery to night loads)
  • Backup capability (grid outages, with the right setup)

The model is Deye SUN-6K-SG03LP1-EU, mainly because it is cost-effective and not tied to an expensive proprietary battery ecosystem.

2) The real sizing constraint: PV string voltage

My panels are JA Solar JAM72S30-565/LR. The key datasheet values are:

  • Voc (open-circuit voltage) about 50.5 V
  • Vmp (voltage at max power) about 41.7 V

The inverter constraints that matter are:

  • Max DC voltage (absolute ceiling)
  • MPPT operating range (where it can actually track power)

So you do not “choose 12 panels”, you choose a string length that stays safe in winter and still sits inside the MPPT range.

Final string design: 6 + 6 (two independent strings)

With 12 panels, the clean configuration is:

  • String 1: 6 panels in series to MPPT1
  • String 2: 6 panels in series to MPPT2

Now the useful numbers:

String Vmp (operating):

Vmp,string=6×Vmp,panel=6×41.7≈250 VV_{mp,string} = 6 \times V_{mp,panel} = 6 \times 41.7 \approx 250 \text{ V}

String Voc (STC):

Voc,string=6×Voc,panel=6×50.5≈303 VV_{oc,string} = 6 \times V_{oc,panel} = 6 \times 50.5 \approx 303 \text{ V}

This is safely below typical 500 V-class hybrid inverters, and about 250 V is a comfortable MPPT operating voltage.

Winter check (cold makes Voc higher)

Voc increases when temperature drops. The proper way to verify is:

Voc,cold=Voc,stc×(1+∣βVoc∣×(25−Tmin))V_{oc,cold} = V_{oc,stc} \times \left(1 + \left|\beta_{Voc}\right| \times (25 - T_{min})\right)

Where:

  • beta_voc is the panel Voc temperature coefficient from the datasheet
  • Tmin is your local minimum temperature assumption

Even with a conservative cold uplift (rule of thumb about 10-15%), a 6-panel string stays well below a 500 V ceiling. That is why 6 in series is the sweet spot here.

What I did NOT do:

  • 12 panels in one string (voltage too high)
  • Paralleling strings into one MPPT (unnecessary and can break current limits)

3) Cable sizing for a 40 m DC run (and the math)

The inverter is not next to the roof. Distance from array to inverter is about:

  • L = 40 m one-way, so the electrical loop is roughly:
Lloop=2×40=80 mL_{loop} = 2 \times 40 = 80 \text{ m}

I used 6 mm2 PV cable, because with long runs, voltage drop becomes real.

Resistance of the cable

Using copper resistivity:

ρ≈0.0175 Ω⋅mm2/m\rho \approx 0.0175 \, \Omega \cdot \text{mm}^2/\text{m} R=ρ×LloopA=0.0175×806≈0.233 ΩR = \rho \times \frac{L_{loop}}{A} = 0.0175 \times \frac{80}{6} \approx 0.233 \, \Omega

Voltage drop at operating current

For one string, current is about the panel Imp (series does not change current):

I≈13.6 AI \approx 13.6 \text{ A} ΔV=I×R=13.6×0.233≈3.2 V\Delta V = I \times R = 13.6 \times 0.233 \approx 3.2 \text{ V}

Percentage drop

At string operating voltage about 250 V:

%ΔV≈3.2250×100≈1.3%\%\Delta V \approx \frac{3.2}{250} \times 100 \approx 1.3\%

That is a good result (a common target is staying under about 2-3% for DC runs).

Power lost as heat (for completeness)

Ploss=I2×R=13.62×0.233≈43 WP_{loss} = I^2 \times R = 13.6^2 \times 0.233 \approx 43 \text{ W}

So the cable is not just cost; it directly impacts efficiency and long-term reliability.

4) Practical takeaway

  • The correct 12-panel design with this inverter is 6S + 6S, one string per MPPT.
  • The critical safety check is winter Voc (cold uplift), and 6 panels gives a big margin.
  • With a 40 m run, 6 mm2 keeps voltage drop around 1 to 1.5%, which is exactly where you want to be.