Some products you build to sell. Some products you build to learn. The Solar PCU — Su-Kam’s three-phase Power Conditioning Unit, conceived in 2004 and deployed in 2006 — was both. It was a genuine commercial product, installed and running at over fifty sites across India. And it was also the most intensive education I ever received about what it actually takes to put solar power to work in the real world.

In 2004, when I first sat down with my engineering team to draw the architecture of what would become the Solar PCU, solar photovoltaic in India was not a market. It was barely an idea. The National Solar Mission would not be launched for another six years. The cost of a solar panel was so high — roughly ten times what it is today — that every installation was a bespoke project, every site a negotiation between the economics of what customers needed and the reality of what solar could deliver at those prices.

I built it anyway. Not because the numbers were comfortable, but because the direction was correct.

What the Solar PCU Was: A Complete Hybrid Power System

SLIDE 01 · Solar PCU System Architecture — PV Panel → Hall Effect sensor → MPPT → Bidirectional Inverter → 3-Phase Y/Y Transformer → Changeover Section (Relays) → Customer Load. Input Grid with CT sensing on all three phases (R, Y, B). Battery Bank at 360V DC. A complete self-contained hybrid power conditioning unit designed and engineered at Su-Kam Gurgaon R&D, 2004–2006.

The architecture diagram in this slide is one of the most technically dense things Su-Kam’s R&D team ever drew. It is not a product brochure. It is an internal engineering specification — the kind of document that lives in the R&D room, covered in handwritten annotations, revised a dozen times before the first unit is built. Looking at it today, I see two years of thinking compressed into a single page.

The Solar PCU was a complete hybrid power conditioning unit integrating five major subsystems:

The PV input stage used Hall Effect current sensing on the solar panel output — a non-contact measurement method that was more accurate and more reliable than resistive shunting in the high-current, outdoor environment of PV installations. This fed into an MPPT (Maximum Power Point Tracker) — the intelligence that extracts maximum available power from solar panels as irradiance and temperature change through the day.

The MPPT output connected to a bidirectional inverter — an inverter that could work in both directions, converting DC from the solar panels and battery to AC for the load, and also converting AC from the grid to DC to charge the battery bank. The inverter drove a three-phase Y/Y isolation transformer, which stepped the voltage up to three-phase grid levels and provided galvanic isolation — critical for safety and power quality in commercial and industrial installations.

Finally, a relay-based changeover section managed the switching logic between solar/battery power and the input grid, with Current Transformers (CTs) sensing all three phases (R, Y, B) on both input and output sides for accurate power measurement. The battery bank ran at 360 volts DC — a high-voltage design choice that reduced current at a given power level, improving efficiency and reducing cable losses across the installation.

The 11-Point Sensor Architecture: What the System Had to Measure

SLIDE 02 · ADC & Sensor Requirements for In-House MPPT — 11 measurement points covering AC input/output (voltage and current, all three phases), DC battery side, PV side, thermal monitoring, and ambient temperature for battery voltage compensation. This list defined the intelligence layer of the PCU.

The second slide is the intelligence specification — a list of eleven parameters the PCU’s control system needed to measure, process, and act on in real time. This list is what separates a sophisticated power conditioning unit from a simple inverter-charger. Every measurement point had a reason:

The note at the bottom of this slide is telling: “Need Energy meter in Output side for true power measurement.” This is the kind of annotation that only appears in a real engineering document — the recognition, mid-design, that apparent power measurement (voltage × current) is not the same as true power (accounting for power factor), and that a proper energy meter is needed for accurate billing and ROI calculation. Real engineering thinking, in real time.

The Economics That Stopped Mass Production

Fifty-plus installations might sound modest. In the context of 2006 Indian solar economics, it was extraordinary. Each installation required a site survey, a custom panel array design, cabling that no standard electrician had seen before, and commissioning by an engineer who had to learn on the job — because there was no job description for “solar PCU installation engineer” in India at that time. We wrote the playbook as we went.

What the high panel costs prevented was scale. The PCU itself was well-engineered and worked. The barrier was always on the generation side — customers who needed reliable power could afford the PCU, but the panel cost to generate meaningful energy made the total system cost difficult to justify against diesel generator alternatives, which were the dominant comparison at the time.

I made the decision not to mass-produce the Solar PCU not because it had failed, but because continuing to invest in scaling a product constrained by an external cost barrier — solar panel pricing — made less strategic sense than developing the next layer of technology that would serve customers better when costs eventually fell. That decision led directly to Su-Kam’s in-house MPPT development in early 2007.

What Going to Sites Taught Me That No Laboratory Could

I want to be honest about something that no engineering specification captures: installing these solar PCUs across India was genuinely hard. Not hard in the way that solving a circuit problem is hard. Hard in the way that the real world is hard — unpredictable, uncooperative, and completely indifferent to what your diagram says.

The Pivot: In-House MPPT and Why We Moved On

This is the pattern that I believe defines genuine technology companies, as distinct from companies that sell technology: every product you build teaches you enough to make the next product obsolete. The Solar PCU taught us thermal management, field commissioning, three-phase integration, high-voltage DC safety, and — critically — exactly what an MPPT controller needed to do to work well across Indian irradiance conditions. That knowledge became the foundation of Su-Kam’s own MPPT development. When we had built our own, the Solar PCU’s third-party MPPT component was no longer the best option we had. So we moved forward.

Why This Matters: Su-Kam Was Building Solar India in 2004

India’s National Solar Mission launched in January 2010. By that point, Su-Kam had already been designing, installing, and servicing solar systems for six years. We had made every mistake that the industry would later learn to avoid. We had developed practices, procedures, and product designs that the market would eventually standardise around. We had trained engineers who became some of the most experienced solar system designers in the country.

None of this was visible from the outside. The Solar PCU was not a mass-market product. It did not generate headlines. The fifty-plus installations were quiet, site-by-site, customer-by-customer deployments that built capability without fanfare. But that quiet accumulation of real-world experience — on rooftops and in substations, in summer heat and monsoon humidity, with customers who were betting real money on a technology that most of their peers considered premature — is what made Su-Kam’s solar division genuinely expert rather than merely technically qualified.

When I look at the Solar PCU architecture diagram today, I do not see a discontinued product. I see the foundational curriculum of Su-Kam’s solar engineering education — the document that started a six-year learning journey that nobody in India had taken before us.