Bare Circuit Card, Module, or Fully Enclosed Unit? 4 Questions That Decide How Much RF Integration to Buy

By Pat Sherlock, NewSpace Growth Lead, Triad RF Systems

Most conversations about RF amplifiers start with the numbers: gain, output power, efficiency, frequency band. Those matter. But one of the most consequential decisions on the RF side of a small satellite program usually gets made before anyone compares datasheets, and often without much discussion at all.

Do you buy a bare circuit card and integrate it yourself, or do you buy a fully enclosed unit that bolts in and connects?

In practice, it isn’t a choice between just those two. The options run from a bare circuit card, to a connectorized module, to a fully enclosed and tested unit. Each step up moves work from your team to your supplier. Each step down saves money on the purchase order and adds work somewhere else in your program. Neither direction is right by default. The mistake is treating it as a pricing decision when it’s really a resourcing and risk decision.

After 26 years working as an integrator and with integrators, the last 8 helping SmallSat teams make this call, these are the four questions I’d ask before you decide.

1. Does your team have the time and people to integrate RF hardware?

A bare circuit card is the cheaper line item. It’s rarely the cheaper cradle-to-grave cost. Someone still has to design the housing, manage the thermal path, handle shielding and grounding, and test the assembly. If you have an RF engineer with bandwidth and a bench to do it on, that may be a good trade. If your RF engineer is also covering multiple other link designs, power, and half of systems engineering, the savings on the purchase order can disappear into schedule.

I worked with one customer on an extremely aggressive schedule who needed an amplifier that already existed off the shelf. The catch was that it existed as a bare circuit card, because that’s what the original customer had required for their platform. Rather than work with us on an enclosure design, or bring us into the mechanical integration, their message was clear: “Get us the units as fast as possible, and we’ll handle the rest.” In the end, they spent more time on mechanical integration, and on calls with our engineers, than it would have taken us to design an enclosure that worked out of the box.

The honest test: write down who will do the integration work, how many hours you think it will take, and when it has to be finished. If any of those answers is “we’ll figure it out,” that tells you something.

2. Where is the heat going to go?

On a small spacecraft, the power amplifier is often one of the larger heat sources on board, and there’s no convection to carry that heat away. Everything depends on conduction and radiation: from the device, through the board, into a housing or mounting surface, and out to the rest of the thermal design.

When you buy a bare circuit card, you own that entire path. When you buy a finished unit, the thermal interface should already be engineered, modeled, and tested, and your job shrinks to mounting it correctly.

Thermal management on a SmallSat is a platform-level problem. There are multiple heat sources, often from different suppliers, but there has to be one system that manages all of them. That means several suppliers and integrators have to work together to define their piece of the puzzle. Only once every piece is defined does the platform integrator have what they need to design the thermal management system. When you buy a finished unit with defined thermal tolerances, you have your piece of that puzzle figured out from day one.

If your thermal analysis isn’t mature yet, that’s a reason to lean toward more integration from your supplier, not less.

3. Who owns the testing?

Testing a board and testing a flight-ready unit are different jobs. At the unit level, you’re looking at performance across temperature, vibration, and the other environmental conditions your mission will see. Somebody has to plan that testing, run it, and document it.

If you buy the circuit card, that somebody is probably you, and it may mean outside test facilities, additional fixtures, and time on the schedule you didn’t budget. If you buy the finished unit, much of that work should already be done and documented.

This question also connects to flight heritage, and the level of integration matters more here than most people expect.

In my view, “flight heritage” on a fully enclosed unit holds more water than it does on a circuit card. With a finished unit, the mechanical enclosure, the connectorization, and the thermal management have all been verified in space, together. With a circuit card, we often don’t know what the integrator did with our board to make the mission a success. We don’t know what the enclosure and shielding looked like, or what thermal interfaces were required.

So before you choose, get a clear answer from your supplier on exactly what has been tested, at what level, and what documentation comes with it. A heritage claim means the most when the configuration that flew matches the configuration you’re buying.

4. How much do you expect the design to change?

A finished unit locks down the mechanical envelope, connectors, and electrical interfaces. That’s a benefit when your spacecraft design is stable: you know exactly what’s arriving and how it fits.

A circuit card gives you more room to adapt as the design moves. That flexibility is real, but it comes with more ways for problems to surface late, when they’re most expensive to fix. If your layout, power budget, or mission profile is still in flux, ask whether the flexibility is worth the extra exposure. And if you do go with a finished unit, make sure your mechanical interfaces are settled first.

I’ve seen what happens when they aren’t. We had repeat customers, working with a new integrator for the first time, find out that the mounting hole pattern had to change after the units were already with their integrator. They considered drilling new holes into the enclosure of a fully tested unit, which would have been a serious risk. Instead, they had to pay us to change the hole pattern, manufacture new metalwork, and build, test, and deliver a completely new unit. It was a major hit to their schedule. That’s the tradeoff: a finished unit gives you far less room to change course late in the program.

Make the decision on purpose

None of this means a bare circuit card is the wrong choice. For teams with strong in-house RF capability, established integration processes, and their own test resources, it can be exactly the right one. But that team has to be top-notch. Almost all of the circuit card designs we’ve sold successfully have gone to very established names with very strong engineering capability.

The problem is when the decision gets made by default, based on unit price or on whatever the supplier happens to sell. That’s how integration work ends up landing on a team that didn’t plan for it.

That’s also why we offer every level, from bare circuit card to connectorized module to fully enclosed unit, and why we’ll design an enclosure around an existing card when that’s what a program needs. Our job is to give engineers what their program actually calls for, not to steer them toward whatever we happen to have on the shelf. With 600+ delivered payloads, we’ve seen this decision play out across a lot of programs, and the right answer depends on your team and your mission, not on our catalog.

If you’re working through this tradeoff on your own program, I’m happy to talk it through.

About the author

Pat Sherlock is the NewSpace Growth Lead at Triad RF Systems. He leads the company’s go-to-market strategy for LEO, SmallSat, and CubeSat RF amplifier customers and works directly with engineering teams on mission-specific RF front ends for NewSpace missions.

Triad RF Systems builds mission-specific RF subsystems for LEO constellations, satellite IoT networks, and NewSpace missions, with 600+ payloads delivered and flight heritage across UHF through Ka-band.