SAW vs BAW Filter Selection: Walsin’s Catalogue Holds 8 Parts, Not a Range

SAW vs BAW filter selection in Walsin’s own export comes down to 8 real part numbers: 3 SAW and 5 BAW, sharing one 14-column schema, with exactly 1 part matching a stated 2,400–2,483 MHz, 50 Ω, 3.0 dB requirement on both counts. The two files also disagree on how they write “no second band.”

---

Venatech Limited, an authorised Walsin distributor in Hong Kong, holds every parametric export file published by Walsin Technology (Walsin), including Walsin_SAW_Filter.xlsx and Walsin_BAW_Filter.xlsx, the two files this post is measured from in full. Every part number below is real and came out of one of those two files. We hold no RF datasheet of any kind, so every figure here is the export’s own value.

How many SAW and BAW filters does Walsin’s export actually offer?

Eight, total. We opened both files row by row: Walsin_SAW_Filter.xlsx holds 3 part numbers, all in a single 1109 case size, and Walsin_BAW_Filter.xlsx holds 5, across three case sizes — 1109 (3 parts), 1411 (1 part) and 1814 (1 part, a duplexer). The two files share an identical 14-column schema, column for column, even though the underlying technologies are physically different — a surface acoustic wave device and a bulk acoustic wave device are not described any differently by this export.

File	Part numbers	Case sizes present	Frequency columns populated
Walsin_SAW_Filter.xlsx	3	1109 only	Frequency 1 on all 3; Frequency 2 on none
Walsin_BAW_Filter.xlsx	5	1109, 1411, 1814	Frequency 1 on all 5; Frequency 2 on 1 (the duplexer)
Total	8	4 distinct sizes	—

We have not seen these two files measured against each other before, including by us.

Which of the 8 parts actually matches a 2,400 to 2,483 MHz requirement?

Exactly one. We set a stated requirement — a single-band filter at 2,400–2,483 MHz, 50 Ω, insertion loss 3.0 dB or better — and checked all 8 parts against it, band by band, rather than by name or technology.

Part number	File	Size	Band stated	Bandwidth	Insertion loss	Match / mismatch vs 2,400–2,483 MHz, 50 Ω, ≤3.0 dB
BA11092442AE1T	BAW	1109	2,401 – 2,483 MHz	82 MHz	2.5 dB max	Match — band, impedance and insertion loss all clear
BA11092350B40E1T	BAW	1109	2,300 – 2,400 MHz	100 MHz	3.0 dB max	Mismatch — band sits entirely below the requirement
BA11092593B41E1T	BAW	1109	2,496 – 2,690 MHz	194 MHz	3.2 dB max	Mismatch — band above the requirement; IL also over 3.0 dB
BA14112593B41E1T	BAW	1411	2,496 – 2,690 MHz	194 MHz	3.0 dB max	Mismatch — band above the requirement; larger body
DA18141747B3E1T	BAW	1814	1,710–1,785 / 1,805–1,880 MHz	75 / 75 MHz	3.7 / 3.7 dB max	Mismatch — a duplexer at a different band pair entirely
SF11091582GC2T	SAW	1109	1,559 – 1,605 MHz	46 MHz	2.5 dB max	Mismatch — GNSS band, not Wi-Fi/Bluetooth
SF11092350B4004T	SAW	1109	2,300 – 2,400 MHz	100 MHz	2.8 dB max	Mismatch — band sits entirely below the requirement
SF11092595B4108T	SAW	1109	2,535 – 2,655 MHz	120 MHz	2.8 dB max	Mismatch — band sits entirely above the requirement

One match out of eight, and it is a BAW part. We will not tell you SAW and BAW are equivalent technologies for this band — we will show you that, on this export, only one specific part number clears all three fields you actually asked for.

Do the SAW and BAW files use the same missing-value convention?

No, and this is the fact we have not seen counted anywhere. We read every cell in both files’ Frequency 2, Impedance for Frequency 2, Insertion Loss for Frequency 2 and Attenuation for Frequency 2 columns. All 3 SAW rows write the literal string - in every one of those cells. All 4 single-band BAW rows leave the same cells genuinely blank — a real null, not a character.

A script that checks for missing data with a null test such as .isna() will correctly flag all 4 single-band BAW rows as having no second frequency, and will miss all 3 SAW rows, because a cell containing the text - is not null to that test. The two files describe the identical real-world condition — “this part has only one passband” — with two incompatible representations of “nothing here.”

File	How “no second band” is written	Rows affected	Caught by .isna()?
Walsin_SAW_Filter.xlsx	Literal string -	3 of 3	No
Walsin_BAW_Filter.xlsx	Blank cell (null)	4 of 5	Yes

Does the one duplexer row expose a gap even where both bands are filled in?

Yes. DA18141747B3E1T is the only row in either file with two real, populated frequency bands — 1,710–1,785 MHz and 1,805–1,880 MHz — and both Insertion Loss for Frequency 1 and Insertion Loss for Frequency 2 are filled in, at 3.7 dB max each. But Impedance for Frequency 2 is still blank on that row, exactly as it is on every single-band part. We checked this cell by cell: impedance is recorded once per part number in this export, not once per frequency band, even on the one row built specifically to carry two bands. Insertion loss is the only field that actually tracks both bands separately.

What does the export give you, and what does a real SAW vs BAW decision still need?

Band edges, impedance (mostly), insertion loss and a rough attenuation mask as free text — and nothing past that. We listed every column in both files against what an engineer choosing between SAW and BAW actually asks.

What the file gives	Present?	What it does not give
Passband (as two edge frequencies)	Yes, all 8 rows	A computed centre frequency or bandwidth — we calculated both above ourselves
Impedance	Yes, 50 Ω on every populated cell	No field distinguishes balanced from unbalanced ports
Insertion loss	Yes, all 8 rows	No stated test temperature except where it appears inside the free-text attenuation string
Rejection / attenuation mask	Yes, as an unstructured multi-line text block	No columns for individual rejection bands — the 5 to 8 stop-band specs per part are one unparseable cell
Case size code	Yes — 1109, 1411, 1814	No millimetre dimensions anywhere in either file; the code is not decoded
Q factor, power handling, operating temperature	No	Not a single column exists for any of the three in either file
AEC-Q200 or automotive qualification	No	No column, and no series name implies it

Zero of the three fields a design-in review usually asks for first — Q, power handling and operating temperature — exist in either file. Insertion loss and impedance are not the whole decision; they are what this particular export happens to capture.

What documentation does Venatech hold for SAW or BAW parts?

None. We hold 28 Walsin datasheets, 451 pages — 26 MLCC and 2 chip resistor — and checked our own _Walsin_Datasheets folder again while writing this post: it has an MLCC subfolder and a Resistor subfolder, and nothing else. Zero datasheets, zero application notes, cover any SAW or BAW part.

What should a buyer choosing between SAW and BAW do with this?

State the band, the impedance and the insertion-loss ceiling before you ask which technology is “better,” because on this catalogue the technology label decides almost nothing by itself — the specific part number does. Of 8 total RF filter parts classed as SAW or BAW, only BA11092442AE1T clears a stated 2,400–2,483 MHz, 50 Ω, 3.0 dB requirement, and it is a BAW part, not because BAW beats SAW in general, but because it is the one row whose band happens to sit where you need it.

1. Write your band, impedance and insertion-loss ceiling down before you look at the catalogue. With 3 SAW and 5 BAW candidates, there is no room for a vague request.
2. Treat a “-” cell as unconfirmed, not as zero. We found it only by opening every cell; a null check alone will not catch it.
3. Cross your adjacent MLCC and passive BOM lines through the free cross-reference tool. Paste up to 500 part numbers or drop your Excel or CSV; it is matched against Walsin’s published catalogue and states every difference from your original.
4. Check what is physically in Hong Kong on ready stock, a dated snapshot as at 23 September 2026, before committing a filter choice to a production slot.
5. Ask us for the SAW or BAW datasheet by part number. We hold none today; your request is what puts it on our list to Walsin.

For how the rest of the RF line is organised, see the Walsin RF front end: switches, BAW, SAW and multiplexers and Walsin RF balun, filter, coupler, multiplexer. The full range is on the Walsin RF product page.

The short version

SAW vs BAW filter selection on Walsin’s own data is a choice between 3 and 5 real part numbers, not between two technologies in the abstract. We opened both files in full: one BAW part, BA11092442AE1T, matches a stated 2,400–2,483 MHz, 50 Ω, 3.0 dB requirement on all three fields, and the other seven miss on band, impedance capture or both. The two files write “no second band” two different ways — a literal - string in SAW, a genuine blank in BAW — so a null check alone misses half the catalogue. The one duplexer row proves impedance is tracked once per part, not once per band, even where two real bands exist. We hold no SAW or BAW datasheet at all.

Send us your band, your impedance and your insertion-loss ceiling, and we will tell you which of these 8 parts actually clears them — then tell you which was in Hong Kong on 23 September 2026. And if your design needs a Q, power or temperature figure neither file states, say so: that is the question we would rather put to Walsin than guess at.

---

Frequently asked questions

<!– wp:rank-math/faq-block {"questions":[{"id":"faq-question-01","title":"How many SAW and BAW filters does Walsin actually offer?","content":"Eight part numbers in total: 3 in Walsin_SAW_Filter.xlsx, all in a single 1109 case size, and 5 in Walsin_BAW_Filter.xlsx, across three case sizes including one duplexer. We opened both files row by row to get this count.”,”visible”:true},{“id”:”faq-question-02″,”title”:”Which Walsin SAW or BAW part matches a 2,400 to 2,483 MHz, 50 ohm, 3.0 dB requirement?”,”content”:”Only BA11092442AE1T, a BAW filter at 2,401 to 2,483 megahertz, 50 ohms, with an insertion loss of 2.5 decibels maximum. We checked all 8 SAW and BAW rows against this stated requirement band by band; the other seven miss on frequency, insertion loss, or both.”,”visible”:true},{“id”:”faq-question-03″,”title”:”Do Walsin’s SAW and BAW export files use the same convention for a missing second frequency band?”,”content”:”No. The SAW file writes a literal hyphen character in the Frequency 2, impedance, insertion loss and attenuation cells for its three single-band parts. The BAW file leaves the equivalent cells genuinely blank on four of its five rows. A null check such as isna finds the BAW gaps and misses the SAW ones, because a hyphen is not a null value to that test.”,”visible”:true},{“id”:”faq-question-04″,”title”:”Does Walsin’s BAW duplexer record impedance for both of its frequency bands?”,”content”:”No. DA18141747B3E1T has two real, populated frequency bands and two populated insertion-loss figures, but the Impedance for Frequency 2 cell is blank, exactly as it is on every single-band part. We checked this cell by cell: impedance is captured once per part number, not once per frequency band.”,”visible”:true},{“id”:”faq-question-05″,”title”:”What does the SAW and BAW export not tell you that a real selection decision needs?”,”content”:”Q factor, power handling and operating temperature range appear in neither file, as a column or otherwise. Package dimensions are reduced to a case-size code with no stated millimetre value, and the rejection or attenuation mask is one unstructured block of text per part rather than separate columns.”,”visible”:true},{“id”:”faq-question-06″,”title”:”Does Venatech hold a datasheet for any Walsin SAW or BAW filter?”,”content”:”No. We hold 28 Walsin datasheets, 451 pages, covering multilayer ceramic capacitor and chip resistor families only. Nothing covers SAW or BAW parts, so every figure in this post is the parametric export’s own value, unconfirmed against any datasheet.”,”visible”:true},{“id”:”faq-question-07″,”title”:”What is the fastest way to second-source around a SAW or BAW choice?”,”content”:”State your band, impedance and insertion-loss ceiling first, since the catalogue only holds 8 candidates between the two technologies. Cross the rest of your BOM through the free cross-reference tool, which takes a pasted list of up to 500 part numbers or an Excel or CSV file and states every difference from your original, then check the dated ready-stock snapshot before committing to a build.”,”visible”:true}],”listStyle”:””,”titleWrapper”:”h3″,”titleCssClasses”:””,”contentCssClasses”:””,”sizeSlug”:””} –>

---