SF2321D Selection Guide for 1090 MHz Receivers
Choose an SF2321D SAW filter by validating frequency coverage, RF loss, board integration, and traceable sourcing.
Last updated: November 2026
Bottom Line: Choose an SF2321D filter by checking the wanted frequency band, the receiver noise and blocker budget, and the exact mechanical and sourcing identity of the offered component. This is a 1090 MHz SAW filter, so a nearby catalog entry for a different frequency is a comparison candidate rather than an automatic replacement. Use the manufacturer drawing for the supplied revision, validate the assembled RF path with calibrated measurements, and qualify production lots against the same acceptance limits. A familiar part number and an attractive unit price cannot replace evidence that the filter fits the actual receiver and production process.
Start with the wanted band and the controlled drawing
Frequency coverage is the first selection gate for an SF2321D receiver design. The RFMi-authored SF2321D datasheet identifies a 1090 MHz center frequency and a specified insertion-loss band from 1075 to 1105 MHz. Its source and load reference impedances are 50 ohms. Those figures describe a particular documented device under its stated measurement conditions; they do not describe every filter whose ordering code begins with SF.
Write down the wanted signal band, the largest expected nearby interferers, and the receiver architecture before comparing products. A narrow wanted channel can still need margin for frequency tolerance, temperature drift, and the rest of the analog signal path. Conversely, selecting a broader filter just because it passes the carrier may admit blockers that the following amplifier cannot handle.
The catalog record for SF2321D provides a useful starting point for an inquiry. Treat the attached drawing revision, reel label, supplier manufacturer identity, and electrical acceptance requirements as the controlled procurement documents. Legacy catalog naming can differ from the name on a later manufacturer document, so reconcile that difference before approving a substitute or issuing an order.
A good design requirement separates passband performance from stopband rejection. Ask what must pass, what must be rejected, and at which test reference planes each limit applies. This prevents a supplier from answering a receiver-level requirement with an isolated component characteristic that excludes board loss or mismatch.
Budget insertion loss before choosing filter placement
Filter placement determines whether its insertion loss mainly affects sensitivity or blocker handling. A passive filter ahead of the first amplifier attenuates the wanted signal before gain is available. A filter after an amplifier preserves more front-end sensitivity but leaves that amplifier exposed to unwanted signals.
For an illustrative room-temperature cascade, a passive input loss of 3 dB has a linear loss factor of about 2. If an amplifier following it has a 1 dB noise figure, the combined input noise figure is approximately 4 dB before accounting for the rest of the chain. This is a generic cascade example, not a measured SF2321D receiver result or a guaranteed system specification.
Put a maximum loss budget beside the sensitivity target and reserve room for connectors, protection devices, traces, and assembly variation. When the measured receiver misses its target, compare component-only and assembled-path measurements before attributing everything to the filter. A damaged launch or incorrect calibration can produce the same apparent loss problem.
An input filter is often justified when nearby transmitters would otherwise overload the first amplifier. A post-amplifier filter may be preferable when the environment is quiet and sensitivity dominates. If both requirements matter, evaluate a staged architecture and document the sensitivity cost alongside the improvement in blocker tolerance.
Use a real blocker test to choose between these architectures. Increase the unwanted input while holding the wanted signal at a defined level, then record the receiver metric that matters to the application. Filter selection becomes much more reliable when the acceptance condition is tied to successful reception instead of an isolated screenshot of transmission loss.
Check impedance and the board reference planes
A nominal reference impedance does not make the surrounding PCB electrically invisible. Controlled-impedance routing, return-current continuity, connector transitions, and ground connections all affect the measured response. Follow the package drawing and layout guidance belonging to the actual offered revision.
The source datasheet describes an internally matched device for its stated source and load conditions. That does not justify attaching arbitrary reactive networks or long unmatched traces without simulation and measurement. Extra matching components can improve one operating condition while making another worse, particularly when board parasitics were omitted from the original model.
Calibrate a vector network analyzer to the intended reference planes and document whether the fixture is included or removed. Use identical fixtures when comparing supplier samples. Otherwise, a difference between two test boards may be mistaken for a difference between component lots.
Record both transmission and reflection behavior. A filter can appear acceptable at the center frequency while creating an undesirable impedance condition near a passband edge. Include the intended frequency sweep, fixture description, instrument settings, and acceptance limits in the validation record so another engineer can reproduce the conclusion.
Evaluate rejection against your actual interferers
Stopband performance should be evaluated at the frequencies that threaten the receiver. A single rejection number without a frequency interval can conceal an important weakness. Select the interfering transmitters and harmonics relevant to the installation, then map their levels through the front-end chain.
Start with an antenna-input spectrum measurement or a justified interference model. A nearby signal may be outside the wanted band yet strong enough to compress an amplifier or create intermodulation before the filter can help. The right architecture depends on where that nonlinear behavior occurs.
Use separate requirements for ordinary operation and the worst credible environment. An engineering prototype that works on a quiet bench may fail close to another transmitter. Equally, designing for an unrealistic blocker can force unnecessary loss, cost, and complexity into a product that never encounters it.
Filter rejection and receiver linearity are complementary controls. Do not claim that changing one component guarantees immunity. Repeat the interference test on the complete assembly with its enclosure, antenna connection, power supply, and operating software because those elements can change the coupling paths.
Qualify the footprint and assembly process
The package drawing must control footprint approval and assembly setup. Catalog dimensions are helpful for screening candidates, but lead geometry, ground-pad arrangement, orientation, and tolerances determine whether an actual board can be manufactured reliably. A family resemblance does not establish footprint compatibility.
Request the manufacturer land pattern and identify the terminal numbering before routing. Make the grounding implementation explicit in the review checklist. Small RF packages can exhibit degraded performance when a nominal ground connection uses a long return path or a layout copied from a mechanically different device.
Use the offered component's documented storage, handling, and reflow requirements to establish the manufacturing process. If a supplier proposes a different revision, compare its process limits before accepting the change. A component that fits the board can still require a different assembly profile or inspection method.
For pilot production, keep representative samples and measured data from the accepted lot. That reference set is useful when diagnosing a later yield change. Compare soldering quality and test-fixture behavior before rejecting a replacement lot solely because its first measured response differs from the prototype.
Resolve sourcing identity and lifecycle before production
A controlled sourcing decision connects the physical component to the documentation used by engineering. The internal SF2321D catalog currently carries a legacy manufacturer label and an obsolete lifecycle field, while the linked RFMi document identifies its own issuer. These records justify a verification request; they do not establish a universal statement about every current sales channel or revision.
Ask the supplier to state the manufacturer shown on the reel, complete ordering code, packaging format, date code, lot code, and documentation revision. Obtain photographs and traceability documents when they are material to the purchase. Resolve any mismatch before approving the commercial offer.
For lifecycle-sensitive production, ask whether the offer represents new manufacture, authorized remaining inventory, or another documented supply source. These cases require different replenishment plans. A single available reel does not establish a sustainable production supply, and an old catalog status does not prove that every possible source has the same lifecycle classification.
Compare total landed cost using the required quantity, delivery schedule, inspection effort, and qualification burden. Published or historical unit prices are not live offers. Use a dated quotation with explicit quantity and packaging details, then make the technical acceptance conditions part of the order documentation.
Compare catalog candidates without assuming interchangeability
The following catalog entries help organize a family-level review. They are distinct devices whose roles must be checked against their own drawings; the table does not present them as qualified replacements for one another.
| Product | Catalog frequency information | Interface information | Mechanical verification | Price basis | Best use in the review |
|---|---|---|---|---|---|
| SF2321D | 1090 MHz target device | Datasheet-controlled RF ports | Exact offered revision | Dated quantity quote | Starting point for the wanted receiver band |
| SF2049E | Catalog describes 915 MHz | Verify its own matching requirements | Separate footprint review | Dated quantity quote | Different-band architecture comparison |
| SF2098H | Catalog describes 915 MHz | Catalog describes a nominal RF impedance | Separate package review | Dated quantity quote | Another different-band screening candidate |
The different frequency targets are a decisive reason to avoid copying a substitute from a search-results list. A purchasing system can place these products in the same category while the receiver requires fundamentally different frequency behavior. Keep category similarity separate from engineering approval.
When a supplier proposes another candidate, compare the full response envelope, mechanical drawing, handling limits, and supply identity. Ask for the evidence needed to close each gap. A replacement becomes usable only after its application requirements and assembly constraints have both been verified.
Follow a repeatable selection decision flow
If the offered drawing does not cover the wanted band, reject the candidate for that receiver and continue the search. If frequency coverage passes, calculate the front-end loss budget and identify the blocker environment. Then choose whether the filter belongs before gain, after gain, or in a staged path.
If the modeled architecture meets both sensitivity and interference targets, verify the footprint and build a controlled test assembly. If measurements disagree with the model, check calibration, fixture loss, grounding, and source/load conditions before changing the component. Once performance passes, repeat the important checks across the required operating conditions.
If the electrical and manufacturing checks pass, review sourcing identity and replenishment risk. Obtain a dated offer that names the accepted ordering code and packaging. Release the component only when the engineering approval, supplier documentation, and purchasing record all refer to the same controlled item.
Keep the decision record short enough to reuse. A requirements sheet, a measured response file, an approved drawing, and a traceable supplier offer are more actionable than an unstructured collection of screenshots. Update that record whenever the proposed source or component revision changes.
Frequently asked questions
Can a different SF-series filter replace SF2321D?
A shared series prefix is insufficient evidence of compatibility. Compare the wanted band, transmission and reflection limits, package drawing, terminal functions, and sourcing identity before treating another product as a replacement. A different-frequency filter can be useful for a separate receiver design while being unsuitable for this one. Approve the substitute against the actual application requirements and verify it on the assembled board.
Should the filter be placed ahead of the low-noise amplifier?
Place it ahead of the amplifier when blocker protection justifies the added input loss. Place it after gain when the sensitivity budget dominates and the amplifier can tolerate the incoming interference. These are architecture choices rather than universal rules. Evaluate both options using a receiver-level sensitivity and blocker test, including the antenna interface and protection network that will exist in the finished product.
What should purchasing request with an SF2321D quotation?
Request the complete manufacturer and ordering identity, packaging format, quantity, date and lot information, delivery terms, and supporting traceability documents. Ask which drawing revision controls the offered goods and whether the supply is repeatable for production. Include the agreed inspection and electrical acceptance requirements. A clear quotation should identify what is being delivered and what evidence will establish that it matches the approved design.
How should the engineering team handle conflicting catalog information?
Treat the conflict as an open documentation item and resolve it before release. Ask the supplier or manufacturer for the current controlled drawing associated with the physical goods, then compare that drawing with the approved design record. Avoid inferring a new manufacturer identity, package revision, or lifecycle conclusion from one search snippet. Keep the resolution with the part approval so purchasing and manufacturing use the same evidence.
Turn the approved design into a traceable inquiry
Choose the SF2321D around the receiver's actual passband, loss budget, interference conditions, and manufacturing constraints. Complete the sourcing review with the same care as the electrical review. The best purchase is the documented component that passes the assembled-board requirements and can be supplied under agreed acceptance conditions.
Use component search to review catalog candidates, then request a quotation with the full ordering code, quantity, accepted drawing, and delivery requirements. Those details make engineering review and supplier comparison more efficient while keeping the final decision tied to the device your receiver actually needs.
