Quectel EM05-G 4G Cat4 and Quectel Antennas: The Split vs. Bundle Comparison That Matters

I’m not an RF engineer. I’m the office administrator for a 40-person hardware company, and I handle component purchasing across about 18 vendors. That puts me in the middle of a lot of conversations I don’t fully own: “Is this module compatible?” “Why does this gateway have low throughput?” “Can we use the cheaper antenna?” I can’t answer all of those. I can answer the procurement question underneath them.

The part I keep ordering is the Quectel EM05-G 4G Cat4. It appears in our mobile computing and gateway designs. It is also a part where the phrase “it’s the same module” can hide a lot of cost.

The comparison: matched Quectel package vs split sourcing

I’m not comparing Quectel to another module manufacturer. I’m comparing two buying decisions.

Option A: order the Quectel EM05-G 4G Cat4 plus Quectel antennas from a distributor we know, with traceable lot codes and return support.

Option B: order the same-looking module part number from the lowest-priced brokerage listing, then choose an antenna based mostly on connector type and price.

Both options can work. I’ve seen Option B work once when we had an RF engineer on staff who could tune and test everything in-house. I’ve also seen it burn more engineering time than the purchase-order savings ever justified. The difference shows up in four places.

1. RF ownership when a signal drops

If a field-return unit comes back with poor LTE performance, the fastest path is one vendor that owns the RF chain. When I order the Quectel EM05-G 4G Cat4 with a Quectel antenna, the engineer has a documented starting point. Quectel publishes module specs and antenna specs in the same vocabulary. If the design fails a radiated test, we can ask one supplier to explain the mismatch.

With split sourcing, the conversation becomes the module vendor pointing to the antenna and the antenna supplier pointing to the module. Nobody owns the link budget. The team starts testing with a different antenna, then with a different module, and the project schedule slips. What looks like a smart purchasing decision on the spreadsheet becomes an engineering problem after the parts arrive.

2. Blue chip only helps if the serial trail is real

Quectel is often described as a blue chip supplier in the IoT module space. I agree, but not because the logo looks reassuring. A blue-chip supplier gives you traceability. The Quectel EM05-G 4G Cat4 family has regional sub-SKUs, and the suffix on the part number matters. Last year I almost ordered the wrong suffix because I only looked at the headline LTE speeds. Our distributor caught the mismatch against the supported bands before it shipped. That check is worth real money.

Consumer phones don’t prepare you for this. A phone is a sealed radio system. The modem, RF switches, antennas and firmware are tuned together, and you never have to justify one element. In an embedded design, the module and antenna are separate decisions. Treating them as separate is normal. But if the split happens only to save money, you are accepting an RF test burden that may be bigger than the savings.

3. Verification: no-name meter vs Klein multimeter is the wrong tool question

I keep seeing tool questions like “no-name meter vs Klein multimeter” in electronics forums. A Klein multimeter is a solid meter. It is still the wrong tool for deciding whether a cellular module is good.

What a multimeter can do is check continuity on an antenna cable. That is useful. It does not tell you whether the antenna is resonant at 700 MHz, whether the module can hold transmit power, or whether a labeled Quectel EM05-G 4G Cat4 is actually an original Quectel part. I once watched a colleague test an antenna cable with a Klein and declare it good because the resistance read close to zero. That only proves the cable isn’t open. It doesn’t prove the antenna is matched to the radio.

Gray-market listings worry me for the same reason. Most brokers are honest, but the “same module, half the price” listing with a stock photo and no lot trace is a gamble. Modules pulled from old phones, tablets and other consumer devices can look fine. If a module was reflowed from another host board, the pins look okay and a multimeter passes the supply pins. The RF switch or crystal can still be damaged. The first time you know is in a test fixture or in the field.

4. The invoice is not the total cost

In 2024 I ran a small split-sourcing test because engineering asked me to challenge our distributor pricing. I don’t remember the exact number, but the broker quote was something like 11% lower on 200 units. We applied our normal incoming inspection and the first problem showed up in under an hour: four units would not register on the test network. Another five had firmware versions that our test tool couldn’t read. Engineering spent about 14 hours deciding whether the issue was firmware, hardware, or our configuration.

At a loaded internal rate of $100 an hour, that’s $1,400. Even before return shipping and restocking fees, the saving no longer looked like 11%. I’m not saying every trial purchase ends that way. But a real saving that drops from 11% to maybe 2% is not worth the delay in a product launch.

When I’d still split the purchase

This is where I try to be honest about the limits of the easy answer. I’d still consider Option B in a few cases.

If your enclosure requires a custom antenna, standard Quectel antennas may not fit. Forcing a puck antenna into an industrial housing with a metal cap can make RF performance worse than a properly designed custom antenna. The module is good; it doesn’t solve every enclosure constraint. A good supplier should say “that antenna is not the right fit” instead of selling you one anyway.

If you have RF engineers and access to a chamber, you can manage the test burden. You can also justify a custom antenna that gives a better radiation pattern. At that point, buying a bare module from a different source might make sense. But I’d keep the module supply traceable. Traceability is not a feature I want to trade away for a few dollars per unit.

What I tell our engineers

Don’t buy a module. Buy a working radio, with someone accountable when it doesn’t work. The blue-chip route isn’t about prestige; it’s about knowing exactly what you received. If the supplier can tell you the regional band variant and the antenna is specified for the same bands, the project is off to a better start than a spreadsheet with the lowest number.

I’d rather work with a supplier who says “this is what we can support, and here is the part that matches” than one who answers every question with “yes, compatible.” Professionalism has boundaries. A purchase order is easier to sign when both sides respect those boundaries.