Moving string voltage from 1000 V to 1500 V has been one of the most effective cost reductions in solar over the past several years: longer strings, less cable, fewer combiner boxes, less racking. It also pushed every device on the DC side into a new stress regime. The parts that fail first are rarely the large switches in the inverter bridge — those get audited repeatedly. They are the diodes that got specified along the way: the blocking diode on each combiner box input, the reverse-blocking device at the inverter DC input, the one that sits reverse-biased all night on a mismatched string. The failure modes are remarkably consistent: devices that open up on the cold morning when string open-circuit voltage spikes, or a batch of combiner boxes coming back for repair after lightning season. Behind both are three calculations that rarely get done properly — voltage margin, blocking loss, and surge verification.

HXY Electronics HC3D25170H is a 1700 V / 25 A silicon carbide Schottky diode with a 1.40 V forward drop, TO-247-2L package, 200 µA reverse leakage, and a 225 A surge current rating. It has no reverse recovery current and no forward recovery voltage; its switching behavior is temperature-independent, and the positive temperature coefficient of VF supports passive current sharing in parallel. It targets combiner box blocking diodes, inverter DC-side reverse blocking, and Boost rectifier positions in 1500 V PV systems. Same-class alternatives include HC3D10170H (10 A, VF 1.50 V, IFSM 55 A), HC3D50170H (50 A, VF 1.50 V, IR only 5 µA, IFSM 150 A), and HC1D80170S (SOT-227, 80 A, VF 1.45 V, IFSM 225 A).

1. The Real DC Stress in a 1500 V System: Start With Cold Open-Circuit Voltage

Many engineers read "1500 V system" as "1500 V bus" and select parts against that number. That reading is half right, and the wrong half is the dangerous one.

The 1500 V on the inverter nameplate is a hard limit, not an operating point. The maximum DC input voltage of a 1500 V central or string inverter is typically exactly 1500 V, while the MPPT operating window is commonly 875 V to 1300 V. What actually sets device stress is the string open-circuit voltage on the coldest morning of the year — module Voc rises as cell temperature falls, with a temperature coefficient β_Voc typically around −0.27 to −0.30 %/°C for PERC, and somewhat lower in magnitude for TOPCon and HJT.

The industry string-sizing expression is:

Voc_max = Voc_STC × [1 + |β_Voc| × (25 − T_min)]

Run it with real numbers. 540 W module, Voc_STC = 49 V, β_Voc = −0.28 %/°C, site extreme low temperature T_min = −30 °C:

Correction factor = 1 + 0.0028 × [25 − (−30)] = 1.154

Twenty-six modules is the critical string length at these parameters, with 2% margin — exactly the lower bound of conventional practice. Two more modules and that cold morning pushes the inverter and every device on the DC side past 1500 V.

The operating bus voltage during the day is not 1500 V either. At low winter temperatures the maximum power point voltage reaches 80%–85% of open-circuit voltage: a 1470 V Voc corresponds to roughly 1200–1250 V at MPP, inside the 875–1300 V MPPT window. In other words, the steady-state bus sits near 1200 V, and only extreme conditions — open circuit, night, mismatch, stacked transients — approach 1470–1500 V.

2. How Much Voltage Margin: Why This Class Has to Be 1700 V

With the worst-case reverse voltage established, the device class follows. The standard derating practice for power devices is: working voltage, spikes and transients included, no more than 80%–90% of the rating. Substituting 1470 V:

A 1200 V device in a 1500 V system is not "short on margin" — it is simply not qualified; the worst-case reverse voltage exceeds the rating by 22.5%. That is why the industry settled on 1700 V as the baseline device class for 1500 V systems.

And the margin at 1700 V is genuinely tight: 1470/1700 = 86.5%, already against the upper edge of the derating window. Two design constraints follow directly.

Cold-temperature string verification is never optional. In the table above, a 28-module string pushes reverse voltage to 1583 V and utilization to 93.1%, straight out of the window. Any error in string design, module selection, or site minimum temperature lands on the DC-side diode first.

Transient overshoot needs its own budget. If the diode is not in a pure DC blocking position but sits at the Boost or freewheel position on the inverter DC side, switching overshoot stacks on top. Published data for two-level 1500 V platforms puts it plainly: at a 1300 V MPP voltage, a 1700 V device has only about 400 V of permissible overshoot left, and with fast SiC switching and commutation loop stray inductance that routinely exceeds 100 nH in a two-level layout, 400 V is not generous. Low-inductance laminated busbars, symmetrical commutation loops, and slowing the switching edge where necessary are mandatory homework at this voltage class. Selecting 1700 V does not by itself close the issue.

3. Blocking Diodes: Position, Loss Accounting, and Leakage Budget

Start with the boundary of what a blocking diode does, because it is routinely assigned jobs that are not its own.

What it does: inserted between each string‘s positive (or negative) terminal and the combiner busbar, it blocks current that would flow backward into that string. Three conditions trigger it — the string is not generating at night or under shading while neighbors sit at higher voltage; mismatch between parallel strings drives circulating current; or polarity is reversed during installation. It fails by reverse overvoltage breakdown, or by reverse fault current exceeding its surge capability.

What it does not do: it does not address PID (potential-induced degradation). PID is driven by high negative bias between the cells and the module frame, causing ion migration; it lives at the level of module construction and inverter grounding/bias scheme. A series diode cannot be pressed into service as a PID fix. The two topics get conflated in design reviews often enough to be worth separating up front.

Where to put it: on each string input inside the combiner box — one per string — rather than a single device at the inverter DC input. The per-string device blocks inter-string circulating current; the single DC-input device only blocks reverse feed from the inverter into the array. They protect different things, and the standard practice in 1500 V combiner boxes is to populate every input.

Now the loss accounting, which is the continuous loss present at this position, and where SiC does not win:

Be honest about this: at a pure DC blocking position, reverse recovery is not the selling point — VF and surge are. A 1700 V SiC Schottky has a higher forward drop than a silicon rectifier at the same voltage class, costing roughly 4 W of extra heat per string per year — on the order of 60 W across a 16-input combiner box. What you get in return are the two items below, and those are the actual selection criteria.

The first is controllable high-temperature reverse leakage. Combiner box interiors routinely exceed 70 °C in summer, and module backsheet temperatures run higher. Leakage loss at full reverse voltage (1470 V), using the listed room-temperature IR:

At room temperature all are fractions of a watt and unremarkable. But leakage rises along the datasheet curve at temperature, and a device reverse-biased all night deserves a pass through that curve. Where the leakage budget is genuinely constrained — insulation monitoring thresholds, nighttime self-consumption audits — the 5 µA of HC3D50170H is clearly the better bin.

The second is surge capability, which is the next section.

4. Surge Verification: Reverse Fault Current and Lightning Surge Are Two Different Things

First: reverse fault current, which sets IFSM. This is where blocking diodes most often die. Take a 16-input combiner box with 13.5 A Isc per string. When one string is reversed or shorted, the remaining 15 feed current into it through the combiner busbar:

15 × 13.5 A = 202.5 A (theoretical)

Derated to 0.75 for line impedance and fuse current limiting, that is 152 A. Clearing time is set by the gPV fuse, on the order of 10 ms — which lines up exactly with the 10 ms half-sine convention for diode IFSM.

Using the industry conversion i²t = 0.005 × IFSM², capability across the 1700 V class is:

The most useful row in that table is HC3D50170H: it is the 50 A class, yet its IFSM is only 150 A — lower than the 225 A of the 25 A part. In other words, IFSM and rated current are not linearly related, and surge capability at a blocking position must be read off the datasheet rather than inferred from IO. The 50 A class is marginally non-compliant in this particular check; its value lies elsewhere — 5 µA leakage, and 50 A of continuous current for high DC/AC ratios and bifacial gain pushing string current upward.

Second: lightning surge, which sets overvoltage coordination. The diode does not absorb this one; the SPD does. For 1500 V systems, DC SPDs carry a Ucpv (maximum continuous operating voltage) of 1500–1600 V — deliberately above the diode‘s worst-case reverse voltage. But what actually determines whether the diode is protected is Up, the voltage protection level: the engineering requirement is Up ≤ 0.8 × Uw, the impulse withstand of the protected equipment. For a 1700 V diode, Uw = 1700 V, so Up must be 1360 V or lower.

Type II DC SPDs used in 1500 V systems commonly list Up around 2.5 kV — far above the 1700 V device rating. The conclusion from that gap is direct: overvoltage protection at the semiconductor cannot rely on the SPD at the combiner box or inverter entrance alone. Local clamping is needed across the diode (TVS/MOV with an RC network), and the device itself must retain adequate surge and overvoltage margin. The SPD protects cable and insulation; it does not protect silicon.

One more caution: IFSM is specified at a defined pulse width and starting case temperature, and the elevated-temperature figure is meaningfully lower. Verify against the surge curve at the worst-case case temperature (per the datasheet).

5. Reference Circuit: Signal Chain and Assembly Thermal Path on the 1500 V DC Side

Place the device in the full chain, from module to grid.

DC power path: module string → string cable → combiner box input terminal → gPV fuse → blocking diode (the subject here) → combiner busbar → SPD (shunting to ground) → DC disconnect switch → inverter DC input → DC EMI filter and bus capacitors → Boost/MPPT stage → inverter bridge → transformer → grid.

Control loop: the inverter DSP runs the MPPT algorithm (perturb-and-observe or incremental conductance) plus the grid-tie double closed loop. MPPT step size and tracking speed must avoid parking the string near its open-circuit voltage — held at open circuit, DC-side devices sit at 1470 V reverse bias for extended periods, which is simultaneously the worst case for voltage stress and for leakage loss. Insulation monitoring (IMD) continuously checks pole-to-ground impedance and blocks grid connection on a fault.

Protection loop: this is where the diode lives, and it is a three-layer coordination problem. The gPV fuse interrupts sustained fault current in the string — its clearing time sets how long the diode sees surge current. The SPD shunts lightning energy — its residual voltage sets the transient overvoltage the diode sees. The blocking diode itself blocks reverse current and withstands reverse voltage. All three must be verified together: fuse clearing time × fault current → diode IFSM; SPD residual voltage Up → diode overvoltage margin.

Power and thermal path — where 1500 V differs:

Budget the heat this way: the blocking diode dissipates 18.2 W (13 A × 1.40 V). With a combiner box interior at 70 °C and a junction temperature target of 125 °C or below, the allowable rise is 55 K, so the required total thermal resistance is Rth(j-a) ≤ 55 / 18.2 ≈ 3.0 K/W. A TO-247-2L needs a real heatsink to hit that (the device‘s own RthJC is per the datasheet). SOT-227, having removed the insulating pad layer, works with considerably more margin at the same budget, at the cost of package volume and price. Also worth stating: creepage and clearance on a 1500 V DC side are substantially tighter than at 1000 V, and tighten further with altitude as air density and breakdown strength fall. Size against the applicable clauses of IEC 62109 or equivalent, and against the certifying body‘s determination.

6. Five Things to Get Right With 1500 V DC-Side Diodes

Verify the cold-temperature string voltage every time, and put it in the design file. Compute worst-case open-circuit voltage with Voc_STC × [1 + |β_Voc| × (25 − T_min)] and leave at least 2% margin; recompute whenever the module type or site changes. Two miscounted modules push the diode to 93% utilization.

Read IFSM off the datasheet; never infer it from rated current. The 225 A of the 25 A class against the 150 A of the 50 A class proves the two are not proportional. At a blocking position, verify against the reverse current the remaining parallel strings can drive into a faulted string, coordinated with gPV fuse clearing time, using the surge curve at the actual case temperature (per the datasheet).

SPD Up does not protect silicon; add local clamping. A 1700 V device is only protected when Up is 1360 V or lower, while Type II SPDs for 1500 V systems commonly list Up around 2.5 kV. The combiner box SPD protects cable and insulation; the device still needs a TVS/MOV with RC across its terminals.

On TO-247-2L the tab is an electrical node, and the insulating pad is part of both the thermal path and the insulation system. Choose a thermally conductive insulator that satisfies both the thermal requirement and the 1500 V insulation withstand, and check creepage through the screw bushing at the same time. SOT-227 has an internally isolated baseplate, but its insulation system still needs to be confirmed against the system voltage.

Account for nighttime reverse-bias leakage; do not budget on VF alone. Multiply the listed IR by worst-case reverse voltage for the room-temperature figure, then walk the high-temperature curve in the datasheet. Where the leakage budget is tight, prefer the lower-IR bin (such as the 5 µA of HC3D50170H).

7. Selection Summary

Close the chain: cold string verification sets worst-case reverse voltage; worst-case reverse voltage sets the voltage class (1700 V for 1500 V systems); reverse fault current sets the IFSM class; leakage and thermal budget set the specific part and package. Work those four steps and the selection converges on its own.

HXY Electronics is a power semiconductor brand with ten years in the industry and more than 5000 catalog part numbers. Its silicon carbide diode line covers 122 devices across 650 V, 1200 V, and 1700 V classes, with rated current from 2 A to 250 A, forward drop from 1.3 V to 1.7 V, and surge current from 18 A to 225 A, in more than ten package formats including TO-247-2L, TO-220C-2L, TO-263, TO-252-2L, DFN8X8B, and SOT-227. HXY staffs field application engineers for selection support, and parts are available through sample request, online ordering, volume quotation, lead-time inquiry, and scheduled (forward) ordering.

One sentence to close: when a diode opens up in a 1500 V system, the cause is usually not device quality — it is a cold open-circuit voltage calculation that was never done. Put string length, worst-case reverse voltage, and reverse fault current into the design verification sheet, and the 1700 V class becomes a comfortable choice.

Disclaimer: Industry comparison figures in this article are drawn from public sources and are provided for selection reference only; they do not constitute a procurement or design commitment. Refer to the latest official HXY datasheet for electrical parameters. For selection and technical support, contact HXY Electronics at sales@hxymos.com.

声明 / Disclaimer
本文部分参数引自华轩阳产品手册或第三方规格书，数据有更新可能，仅供选型参考，最终设计请以最新版规格书与实测为准。
联系 Contact：sales@hxymos.com