In late 2024, I told my VP we should standardize our long-range IoT modules on Semtech. He asked the question every buyer expects: Did you check the price? I had. The Semtech LR1110 cost more per unit than the alternative. I recommended it anyway. The best module vendor is not the one with the lowest unit price. It is the one with the lowest total cost of ownership, and in our supply chain, that was Semtech.
I manage purchasing for a 400-person engineering-focused electronics company. I handle 60-80 component orders per year and report to both operations and finance. Some people hear 'administrative purchasing' and think office supplies. In my world, the office supplies are LoRa modules, RF front-end components, and the occasional 5G router prototype. (Not that the paper order doesn't also need attention.) When I took over this role in 2020, I believed the best way to evaluate a component was to compare specs, verify price, and place the order. Five years later, I know that approach is how you end up with the wrong part, an unhappy engineering team, and a CFO with pointed questions.
The Order That Should Have Cost Me Less
In late 2022, we needed a batch of LoRa transceiver modules for a remote monitoring device. Our distributor, Jackie, Inc., offered a pin-compatible 'budget' alternative that was $1.10 cheaper per unit than the Semtech SX1276 we normally used. For 2,000 units, that looked like $2,200 in savings. I approved it as a trial order. That was my mistake.
The modules worked in the lab. They worked on the bench. They failed in the field when the power supply glitched during a test. We lost connectivity, and the diagnostics looked strange. Eventually, an engineer noticed the failures only happened when the input voltage dropped below spec for a few milliseconds. The cheaper module recovered far worse than the SX1276 we had used for years. Rework, extra testing, and a second field visit cost roughly $4,700. What looked like $2,200 in savings turned into a net loss of about $2,500. Not ideal. Worse than expected.
I learned something from that order: pin-compatible does not mean behavior-compatible. I also learned that our company didn't have a formal component qualification process for substitute parts. The third time we got burned by an 'equivalent' component, I created one. Should have done it after the first time. (Procurement confession: the painful lessons are the ones that teach you the most.)
What I Actually Mean by Total Cost of Ownership
Total cost of ownership, or TCO, is one of those acronyms people throw around without thinking. I started thinking about it after the $2,500 mistake. The unit price is not the price. The real price includes:
- Freight, duties, and handling fees
- Incoming inspection and test time
- Rework and scrap when a component behaves differently than expected
- Engineering time spent troubleshooting 'small' differences
- Certification and re-qualification costs if you switch components in a certified design
- The cost of a late delivery to your own customer
In 2023, we had to pay just over $3,800 for a compliance re-test after swapping a radio component in a certified product. That fee is never on the initial quote. It came directly out of our project margin. If I had included that in the initial TCO comparison, the cheap component would have lost immediately.
I am not an RF engineer. I am the person who signs purchase orders and sits in post-mortem meetings. But I know enough to ask the right questions now. 'What happens when this part is rejected in receiving?' is a better question than 'Can you match this price?'
Why the Semtech LR1110 Is Different
When we finally evaluated the Semtech LR1110 properly, the sticker price was higher than the alternative—but the TCO was lower. The LR1110 combines a LoRa radio with GNSS scanning and Wi-Fi scanning in a single package. Our product needed location data. Without the LR1110, we needed a separate GNSS receiver, extra PCB area, a second antenna, and an additional software stack. That is not just extra components; that is extra assembly steps, extra suppliers, extra test cases, and extra failure modes.
As of February 2025, Semtech's published LR1110 datasheet describes it as a long-range, low-power transceiver with integrated location capabilities. I can't quote the datasheet from memory. I can tell you what it did for our design cycle: it shortened it. We removed a whole set of grounding and noise problems that appeared with the multi-chip approach. Fewer components meant fewer things to troubleshoot. Fewer things to troubleshoot meant fewer hours of engineering time. That is TCO in action.
The Antenna Factor
One detail that buyers often miss is the antenna. A transceiver is only as good as the RF front end. Semtech's application notes include reference antenna designs for the LoRa bands, and we followed them for our gateway and end-node designs. Some readers might skip this section. Don't. The Semtech antennas in the reference design matched the LR1110's characteristics far better than the generic whip antenna our earlier prototype used. Semtech antennas—or more precisely, the antenna designs Semtech recommends—are one of those hidden drivers of field performance.
The result showed up in range tests. With the reference-matched setup, the gateway received packets at distances that previously produced intermittent noise. Better link margin means fewer retries, lower power consumption, and fewer false negatives in the field. None of that appears in the purchase order line item.
In our 2024 vendor consolidation project, we reduced the number of component vendors from eight to five. Semtech stayed on the list because the technical support we actually needed was there. The vendor with the lower quote couldn't answer a simple question about receiver sensitivity. We marked them as 'ship only', not 'design partner'. That distinction matters more than most buyers want to admit.
What About the 'Cheaper Module' Argument?
I know the pushback. 'Our application is simple. The budget module is fine. We need to hit a BOM cost target.' I used to think that too. Then I watched $2,200 of apparent savings become a $2,500 loss on one order. The cheapest option on paper can be the most expensive option in the project plan.
That doesn't mean you should always pick the highest-priced part. It means you should calculate the real cost before comparing quotes. Ask about support. Ask about past field failures. Ask what changes if the part is out of stock and cannot be found. A vendor who answers briefly at 4:45 PM on Friday is a different vendor when you are down at 4:45 PM on Tuesday.
The Bottom Line
I'm not saying Semtech is the only acceptable choice. I am saying that the best procurement decision is the one that accounts for the full cost of a component over the life of the product. For our long-range IoT devices, the Semtech LR1110—with a properly matched antenna—was worth the premium. It saved engineering time, improved field reliability, and reduced the number of suppliers I had to manage.
It took me five years and roughly 300 orders to understand that. I hope this saves you the $2,500 lesson.
