DC-DC Converter Module Manufacturers China (Isolated)
Wholesale DC-DC converter modules from top Chinese manufacturers. Buck, boost, and flyback topologies with CE/UL certifications. Custom OEM available.
Published · Updated
Isolated vs. Non-Isolated: When Topology Choice Is a Safety Requirement
The decision between an isolated and non-isolated DC-DC converter power supply unit (PSU) is not merely a cost decision — it is a foundational circuit architecture decision strictly driven by high-voltage safety, signal noise, and critical system grounding requirements. Getting this wrong at the early design stage is exceptionally expensive to correct during mass production.
Non-isolated converters (buck, boost, buck-boost voltage regulators) share a direct common ground between input and output. They are highly appropriate when:
- Input and output ground rails are intentionally connected (single-supply systems, on-board PCB mounting regulation)
- Output voltage is lower than or extremely close to input voltage (buck), or input is strictly lower than output (boost)
- There is absolutely no risk of high-voltage transients natively appearing on the input rail — for example, a 24V industrial PSU feeding an on-board 3.3V rail for a sensitive microcontroller
Synchronous buck converters sourced from top Chinese manufacturers operating at a fast 300–600kHz switching frequency consistently achieve 90–95% efficiency at moderate loads. The massive efficiency advantage directly comes from elegantly replacing the older Schottky rectifier diode with a second sophisticated synchronous MOSFET, entirely eliminating the diode forward voltage drop (~0.4–0.6V at full load). At a high 5A output, this smart recovery of 2–3W as latent heat is incredibly significant in a highly compact module.
Isolated converters deliberately introduce a highly secure galvanic barrier completely separating the input and output. Physical isolation is rigidly required when:
- The input rail is mains-derived or dangerously carries a hazardous voltage (>60V DC or >42.4V AC peak per strict IEC 62368-1 definitions)
- The output ground must freely float relative to the metal chassis or input ground — typical in industrial control systems and precise sensors measuring delicate signals safely referenced to earth while the main power system has a completely separate earth
- Nasty common-mode noise injection forcefully originating from the noisy input rail would massively corrupt a highly sensitive analog measurement circuit (strain gauges, precision ADCs, pH sensors)
- Strict creepage and clearance distances mandated by IEC 62368-1 or IEC 61010-1 rigidly require true physical separation
Topology mapping:
- Flyback — The absolute most common isolated topology for <100W total output power. It utilizes a single transformer winding (primary + secondary). Highly cost-effective but inherently generates higher output ripple than advanced forward converters. It is the gold standard in tiny 1W–30W industrial sensor power supplies.
- Forward converter — A single-ended forward topology; natively offers lower ripple than standard flyback designs at equivalent power levels. Considerably more common in medium 20–100W modules. It rigidly requires a dedicated transformer reset mechanism (RCD clamp or advanced two-switch design).
- Push-pull / full-bridge — Heavily used well above 100W where simple flyback magnetizing current rapidly becomes fundamentally impractical. Much less common in off-the-shelf modular catalog products; it typically exclusively appears in expensive, custom-designed power supplies.
Chinese module manufacturers very commonly vaguely label flyback modules simply as “isolated” without explicitly specifying the crucial isolation test voltage or the exact creepage/clearance distances. Always request the official isolation voltage test specification (typically a 1,000V DC hipot for 1 full second, or a stronger 1,500V AC for 1 minute) and the continuous working voltage rating. A basic module only rated for 1,000V DC isolation dangerously used in a 230V AC mains-derived application provides dangerously inadequate safety margin — IEC 62368-1 strictly requires reinforced insulation of 2× peak working voltage plus a massive 1,000V minimum.
For major industrial IoT applications where cleanly isolated power feeding RS-485 or highly sensitive analog sensors is a rigid system architecture requirement, we deeply evaluate all modules directly against the target IEC standard heavily during our rigorous inspection process well before any bulk order release.
Efficiency at Real Load Points: What “92% Efficiency” Actually Means
DC-DC module datasheets predominantly report peak efficiency — typically only measured under ideal conditions at 50–75% of rated output current at one specific input voltage. This heavily marketed number merely tells you the absolute best-case conversion loss but is fundamentally irrelevant for properly sizing crucial thermal management in most actual real-world applications.
Load profile heavily matters vastly more than peak efficiency. Consider a standard 5A/5V synchronous buck module boasting a published 93% peak efficiency:
| Load current | Typical efficiency | Power dissipated |
|---|---|---|
| 0.25A (5% load) | 72–80% | 0.31–0.43W |
| 1.25A (25% load) | 88–91% | 0.21–0.34W |
| 2.5A (50% load) | 91–93% | 0.27–0.33W |
| 5A (100% load) | 89–91% | 0.69–0.83W |
Crucial light-load efficiency disastrously collapses in older non-synchronous designs largely because heavy gate drive and constant quiescent current losses vastly dominate at very low output power. A poorly designed module constantly consuming 15mA quiescent continuously at 24V input hopelessly dissipates 360mW completely regardless of the actual output load — at a tiny 100mA output load on a 5V rail (only 500mW output), this massive quiescent loss alone shockingly represents a 42% wasted overhead.
For remote battery-powered IoT modules continuously operating in long sleep cycles, you must rigorously always request the critical light-load efficiency curve (specifically typical at 1% and 10% of rated load) directly from the manufacturer. Reputable top-tier Chinese suppliers (such as MORNSUN or CINCON) actively publish comprehensive full efficiency-vs-load curves directly in their detailed datasheets. If a random supplier lazily provides only a single misleading efficiency figure with absolutely no load specification, immediately treat that vague datasheet as completely incomplete and untrustworthy.
Where to critically apply efficiency data:
Thermal dissipation precise calculation: P_loss = P_output × (1 − η) / η. A 20W output module operating at 90% efficiency dissipates roughly 2.2W as waste heat. At a slightly lower 85% efficiency, the exact same output fiercely dissipates 3.5W — a staggering 60% massive increase in pure thermal load that can easily dangerously push a borderline natural-convection design straight into requiring loud forced-air cooling territory.
For mass OEM applications where the raw module will be strictly certified as part of a much larger total system, precise efficiency data — measured at 25%, 50%, 75%, and 100% load — feeds directly into the crucial end-product energy label compliance calculations (strict DOE Level VI for external power supplies, aggressive ErP Lot 6 for the EU). Our dedicated sourcing team can rapidly provide authentic test reports clearly showing the complete full load-efficiency curves completely originating from the factory’s actual production test, vastly surpassing just the marketing datasheet claim.
Thermal Derating and PCB Layout: The Margin That Disappears in Enclosures
Standard DC-DC converter modules are vastly rated favorably at a cool 25°C ambient with infinite natural convection. In a tightly sealed enclosure absolutely packed with other heavy heat-generating components, the harsh internal ambient temperature directly at the module may easily aggressively reach 50–70°C — completely halving or entirely eliminating the heavily relied upon available thermal headroom completely before severe derating automatically kicks in.
Critical thermal derating curves vividly specify exactly how the raw output current must be strictly reduced as the surrounding ambient temperature unavoidably rises. A standard 3A-rated buck module may severely derate linearly starting from 3A at a warm 40°C straight down to a mere 1.5A at a hot 85°C. If the overworked module constantly operates at a heavy 2.8A continuous inside a tightly sealed enclosure where the stifling internal ambient reliably reaches 65°C, it is dangerously operating highly near or actively beyond its heavily derated rating — which absolutely causes devastatingly premature electrolytic capacitor aging and rapid, eventual power inductor saturation particularly under heavy transient conditions.
Aggressively request the actual derating curve, not merely the vague temperature range. “Operating temperature: -40°C to +85°C” only vaguely means the module will technically function — absolutely not that it can robustly deliver its full rated current cleanly across that entire extreme range. The crucial distinction drastically matters for high-reliability industrial deployments.
Strict PCB layout requirements for bare non-isolated modules:
The central controller IC and the heavy power inductor are strictly the massive primary heat sources naturally inside a dense synchronous buck converter. Premium module manufacturers strictly specify a mandatory minimum copper pour area natively on the host PCB — typically a massive 10–25 cm² of thick 1oz copper securely connected directly to the exposed bottom thermal pad or heavy mounting pad. Woefully insufficient copper area severely reduces natural heat spreading and rapidly creates a devastating hotspot that permanently degrades the power inductor’s crucial DC resistance heavily over time.
Critical, non-negotiable layout rules for successfully integrating DC-DC modules directly onto a host PCB:
- Precisely place large input and output bulk capacitors as physically close as humanly possible directly to the module pins; sloppy long traces between the bulk capacitor and the module fiercely create massive parasitic inductance that aggressively increases dangerous voltage overshoot particularly during sudden load transients
- Strictly separate the noisy high-current switching loop (the main MOSFET, power inductor, and output cap) cleanly from highly sensitive analog measurement traces — enforcing a 5mm minimum rigid separation, preferably with a solid ground plane strictly poured carefully between them whenever possible
- Absolutely do not ever dangerously route quiet signal traces directly through or directly under the noisy switching node area; the extremely active switching node on a powerful buck converter violently swings full rail-to-rail constantly at the high switching frequency and aggressively couples capacitively directly to any adjacent sensitive traces
Forced airflow vs. infinite natural convection. A standard 10W module (dissipating approx. 1.1W at a solid 90% efficiency) relying on natural convection strictly requires approximately an enormous 25 cm² of external heatsink surface or raw PCB copper pour merely to maintain a safe <20°C temperature rise safely above ambient. Strategically adding just 1 m/s of forced airflow directly across the bare module surface massively reduces the strict required heatsink area by a staggering approximately 60%. For highly dense assemblies completely packed with multiple high-power modules, proactively budget essential airflow incredibly early in the mechanical design phase — painfully retrofitting forced airflow into a fully sealed enclosure later is obscenely expensive.
For power electronics products destined for industrial or outdoor environments, our inspection service includes thermal imaging of sample modules operating under full rated load inside a representative mock enclosure, not only at bench level with open airflow.
Top Chinese DC-DC Converter Manufacturers: Verified vs. Re-labeled
The DC-DC module market inside China splits into two tiers: established manufacturers with internal certification programs and strong engineering support, versus small trading companies re-labeling unverified modules under unfamiliar brand names.
Tier 1 premium domestic manufacturers:
MORNSUN (金升阳) is unequivocally the absolutely largest dedicated DC-DC module manufacturer natively in China strictly by massive volume, proudly headquartered deeply in Guangzhou. Their sprawling catalog massively spans from a tiny 0.25W completely up to 300W heavily isolated modules proudly carrying genuine, verifiable UL/CE certification strictly on each exact model number. Their published efficiency specs are highly conservative and easily reproducible. Their critical OCPT (overcurrent protection threshold) is exceptionally well-characterized directly in the main datasheet. Standard lead time directly from the massive factory is reliably 15–25 days for the standard catalog; complete custom label printing is readily available entirely at a highly reasonable 2,000-unit MOQ. Premium pricing is predictably roughly 15–30% safely above totally unbranded, risky alternatives.
Where the input is mains AC rather than a DC rail, the equivalent sourcing decision shifts to an AC-DC power module, which carries its own creepage, isolation, and safety-approval considerations.
CINCON Electronics (台湾, China manufacturing) — A well-established OEM supplier for reliable DC-DC modules in the 1W–75W range. They offer significantly better, highly detailed documentation than most mainland competitors; their crucial CE/UL certifications are strictly third-party independently tested, absolutely not merely self-declared. They are highly commonly specified extensively by strict European industrial equipment integrators. Standard reliable lead time is 20–35 days.
Tier 2 functional alternatives (Murata / RECOM equivalents): Several agile Guangdong manufacturers rapidly produce strict pin-compatible, highly capable alternatives squarely targeting RECOM’s legendary R-78 series (non-isolated, reliable 1A SIP) and actively targeting Murata’s highly popular MEJ and MGJ series (fully isolated, 1–2W). These highly aggressive alternatives are shockingly 40–60% vastly cheaper than the original premium brand. However, the exact functional parity wildly varies: the exact switching frequency, vital internal input capacitance, and crucial thermal performance limit can massively differ significantly straight from the premium part actively being replaced.
Strict verification checklist thoroughly required before dangerously ordering from any unknown supplier:
- Aggressively request the official full CB test report (specifically IEC 62368-1 or older IEC 60950-1) — absolutely not merely the basic CE Declaration of Conformity completely alone. The simple DoC is entirely self-declared; the massive CB report strictly requires an independent Notified Body. Vigorously verify the specific test report issuer is genuinely an active IECEE-recognized CB Testing Laboratory.
- Specifically request the exact UL file number and independently meticulously verify it strictly at ul.com/database. Any claimed UL certification stamped on a Chinese module that completely does not precisely appear directly in the official UL database is entirely fabricated, fake documentation.
- Strongly ask for the specific isolation voltage daily production test record — rigidly confirming that absolutely each individual unit is actively hipot-tested directly on the active line, definitely not just the single golden type-approval sample.
- Rigidly request exactly five physical samples well before ever blindly committing to massive bulk. Carefully manually measure the precise output voltage accuracy, actual active switching frequency (utilizing an oscilloscope directly at the raw switching node), and the exact output ripple heavily at both 50% and 100% full load. Meticulously deeply compare all findings directly against the claimed datasheet values.
Our audit service covers factory visits to MORNSUN, CINCON, and shortlisted domestic alternatives to verify production test coverage, PCB material traceability, and component BOM compliance before you commit to tooling or bulk purchase.
MORNSUN is headquartered in Guangzhou, CINCON manufactures in Guangdong, and most trading companies for smaller modules are based in Shenzhen. For a broader sourcing workflow, see our guide to sourcing electronics from China.
Common questions
When is an isolated DC-DC converter required instead of a non-isolated buck? +
Isolation is required when the input rail is mains-derived or hazardous voltage (>60V DC), when the output ground must float relative to the input ground, when common-mode noise from the input would corrupt sensitive analog signals, or when safety standards mandate creepage and clearance distances. Non-isolated bucks are appropriate for on-board regulation where input and output share a ground and no high-voltage transients are present.
Why is light-load efficiency more important than peak efficiency for IoT applications? +
Battery-powered IoT devices spend most of their time in sleep mode at very low load. A module marketed at 93% peak efficiency may drop to 72–80% at 5% load, wasting a disproportionate share of the battery budget. Always request the efficiency curve at 1%, 10%, 25%, 50%, and 100% load before selecting a module for a low-power design.
How do I verify a Chinese DC-DC module's UL certification is genuine? +
Request the UL file number and verify it independently at iq.ul.com. Also request the CB test report from an IECEE-recognized lab, the production hipot test record, and at least five physical samples. Measure output voltage accuracy, switching frequency, and output ripple at 50% and 100% load to compare against datasheet values.
Related knowledge
Have a sourcing project in mind?
Tell us what you need. We respond within 24 hours, including weekends.