吃瓜tv

Compare an old and a new PCB BOM

Export UTF-8 CSV or paste tab-separated rows from your spreadsheet. Required headers are RefDes and MPN. Supported optional headers are Manufacturer, Value, Footprint, Population and Quantity. 鈥淒esignator鈥� and 鈥淢anufacturer Part Number鈥� are also recognized. Fields outside this list are not compared. Convert boolean DNP/Fitted columns to an explicit Population column before comparing; a Yes flag can have opposite meanings in different exports.

Prepare comparable exports, not just similar filenames

Begin with the actual manufacturing baseline. A file called 鈥渓atest鈥� is not enough: record the assembly number, board revision, BOM revision, variant and release date. Compare a released baseline with the proposed next release. Two exports from different product variants may correctly contain hundreds of differences, but that report does not explain an engineering change.

Keep designators explicit. A grouped row may contain R1 R2 R3, or a quoted CSV cell such as "R1,R2,R3". Expand a range such as R1鈥揜3 before running the tool. Quantity, when present, must equal the number of references on that row for one board. Reel size, order quantity, attrition allowance and the number of boards in a build belong in purchasing records instead.

Do not put a distributor stock code in a column labeled MPN. Preserve the manufacturer's complete orderable part number, including relevant package, tolerance, temperature and finish suffixes. Include the manufacturer separately because a short part number alone may not establish identity. A blank MPN generates a warning; it cannot establish that the component is equivalent.

Worked example: four locations, three kinds of change

The built-in example uses fictional components. Revision A has R1 and R2 populated with the same 10 k惟 resistor, C1 populated with a capacitor, and U1 populated with a controller. Revision B keeps R1, changes R2 to a 22 k惟 resistor, removes C1, adds C2, and changes U1's population instruction to DNP. Both lists contain four reference designators, so a total component count alone would miss the actual work.

The expected result is one added reference, one removed reference, two changed references and one unchanged reference across the union of both lists. R2 requires a circuit review because the electrical value changed. U1 requires confirmation of the intended assembly variant; a part can remain in the BOM while being intentionally unpopulated. C1 disappearing and C2 appearing could be a rename or a real circuit change. The tool reports both facts and does not guess which explanation is correct.

Read the difference list as a review queue

An added part may affect placement access, soldering sequence, component supply and test coverage. A removed part may change a filter, protection function or calibration condition. Review these consequences before approving the revised build. The difference list tells you where to look; the schematic and engineering change record explain why the change is acceptable.

When an MPN changes but the value remains the same, inspect the complete specifications. Two capacitors with the same nominal capacitance may have different voltage ratings, dielectric behavior, dimensions or approved applications. Two controllers can share a package name without sharing a pinout or firmware compatibility. A footprint field that looks unchanged cannot establish that every physical and electrical detail is compatible.

Population changes deserve their own review. A DNP instruction may alter a product option, debug interface or power path. Check the placement file and assembly drawing as well as the BOM. Some CAD exports omit unpopulated components entirely; others retain them with a flag. Choose one convention for this comparison so that a changed flag is not confused with a removed circuit element.

Text differences can also be editorial. The tool trims cell edges and normalizes reference designators to uppercase, but it does not treat 鈥�10k鈥�, 鈥�10 k惟鈥� and 鈥�10000鈥� as equivalent. Manufacturer spelling and package descriptions are compared as supplied. These differences are intentionally visible so the reviewer can distinguish a formatting cleanup from a material change.

What this tool can and cannot establish

The browser tool supports a flat assembly BOM, not a multilevel mechanical product structure. It has no access to a schematic's unique component identifiers, PCB coordinates, procurement systems or manufacturer lifecycle feeds. If reference designators were reannotated between releases, use a CAD comparison or an explicit old-to-new reference map before interpreting the report.

Duplicate references stop the comparison because matching would be ambiguous. A missing required header, inconsistent CSV column count, unterminated quoted cell or quantity mismatch also stops it. Missing optional fields are reported rather than silently treated as evidence that a feature stayed unchanged. Correct the export and run the tool again; do not bypass the validation by deleting a confusing component.

The CSV export includes all compared locations and old/new values for supported fields. The on-screen change preview is capped at 200 rows for readability. Changing either input invalidates the prior result and disables its download. Spreadsheet-sensitive prefixes are escaped in the exported text; inspect imported cell types if you intend to use the file in another automated system.

There is no automatic pass or fail verdict for manufacturing. The tool does not examine circuit behavior, solderability, component authenticity, compliance, lifecycle status, stock, price or whether a substitution is approved. Treat a zero-difference result as 鈥渘o difference found in these supplied fields,鈥� not as permission to release an assembly.

Turn reviewed changes into a controlled assembly handoff

Assign an owner to each material change. Record the decision, approval evidence and intended cut-in point: a new purchase order, build lot or serial range. Existing stock and work in progress need explicit disposition. If a replacement is approved only for one batch, do not silently promote it to the product's permanent approved parts list.

Send the final BOM together with the matching fabrication data, placement file, assembly drawing and test instructions. Identify the variant and revision in the handoff message. Attach the reviewed difference report when it helps explain the change, but keep the approved manufacturing files as the source of truth. Ask the assembly partner to acknowledge open questions before production starts.

For the next step, use engineering change control to define approval and cut-in, or alternative part approval when the change involves substitution. If you are still preparing the first BOM, start with the assembly BOM field guide instead of comparing incomplete lists.

吃瓜tv can discuss the PCB assembly scope using your approved files. You can contact the team at any stage; a completed comparison is not required. The contact form is separate, so attach the files and reviewed report yourself when appropriate.

BOM comparison questions

Method and references

This original 吃瓜tv tool compares exported field values and does not connect to an external BOM service. For release and variant concepts in an ECAD workflow, see . The fictional example and matching rules above describe this tool's own behavior, not a reproduction of another tool. Technical reference checked 4 October 2026.