Specification articles are the easiest to get wrong, because the figures have more than one possible source: the specification sheet, the test record, the equipment nameplate, and also second-hand statements and further editing. If any link in that chain is copied with one digit wrong, or is rewritten in passing while editing, the figures the reader ends up with will not match the machine in front of them. Wrong figures are not the problem; the problem is that the reader has no way of telling which one is the correct one.
First fix the basis: one figure has one source
The test bases and criteria cited on this site are fixed at four items: GB/T 7600 (water content), GB/T 423 (gas content), GB/T 507 (breakdown voltage), NAS 1638 (cleanliness). When an article is written, every figure is taken from one place only, and once taken it is noted which document it comes from. The article is published only when the specification sheet, the test record and the equipment nameplate agree with each other; figures that do not agree are left out rather than replaced by one that happens to look suitable.
Take the ZYD-100 as an example: processing capacity 6000 L/h, working vacuum -0.08 ~ -0.0999 MPa, ultimate vacuum ≤5 Pa, filtration accuracy ≤1 μm, water content ≤3 ppm, gas content ≤0.01%, cleanliness NAS 1638 grade 5, breakdown voltage ≥65 kV, total power 72 + 5 kW, power supply three-phase 380V / 50Hz, overall dimensions 1200×1400×1800 mm, weight approx. 720 kg. These figures have one single authoritative source, and when they are written into an article they are not reworded, not converted and not rounded, and only then do they mean something to a reader checking them against a machine. For the same model, individual parameters may vary between production batches according to the configuration, and each machine is still governed by the test record supplied with it; the figures in the article correspond to the version on the specification sheet and the nameplate.
When a mistake is found, handle it in these four steps
First, withdraw: when a wrong statement appears, first remove the disputed conclusion from the body text, so that it is not cited further. Second, verify: go back to the specification sheet, the test record and the nameplate, and establish which source is wrong — a copying typo, an inconsistent basis, or a different machine version. Third, correct in place: correct the figure in the body and keep the original title, rather than concealing it by changing the title or deleting the article, and do not leave two versions of the statement in the same place. Fourth, state what was changed: write in the article which figure was corrected and on what basis, so that readers of the earlier version can match it up.
How a reader checks a machine that has arrived
Before leaving the factory the machine completes a no-load and load test, the test record and certificate of conformity are supplied with it, and the control cabinet carries the equipment nameplate and serial number marking. After delivery, check in this order: first see whether the model and serial number on the nameplate and the certificate correspond, then compare the values in the test record with the nameplate parameters line by line, and finally confirm whether the test bases in the record are the four that concern you. If the serial number does not match or the record is incomplete, ask before the machine is installed: if the problem is only found after the position and the pipe connections are complete, what has to be reworked is the whole pipe section.
Some differences fall within the normal range and some do not. The criterion is the basis: only values measured for the same indicator under the same basis are comparable, and values from different bases cannot be subtracted directly. When two values do not match, first establish whether the test basis differs, whether the sampling position differs or whether the machine version differs, instead of concluding at once that the machine is faulty. Stating the basis clearly is more useful than adding one vague adjective.
State the limits rather than adjectives
For the water content, write ≤3 ppm and state the basis GB/T 7600, rather than writing that the water content is extremely low; for the vacuum, write the working vacuum -0.08 ~ -0.0999 MPa and the ultimate vacuum ≤5 Pa, rather than writing that the vacuum performance is excellent. An adjective provides no basis for judgement, whereas a figure together with its basis does. Errors in specification articles are not rare, and public corrections are not rare either; putting the source and the basis in the article is the precondition for the reader to be able to check it independently.
| Item checked | What is looked at | What is done first when they do not agree |
|---|---|---|
| Nameplate and certificate of conformity | Whether the model and serial number correspond | Confirm with the supplier first, do not install first |
| Supplied test record | Whether the values and the test bases are complete | Require the missing records to be completed |
| Specification sheet | Line-by-line comparison with the nameplate | Check against the same specification sheet as the reference |
| Article body | Whether the values and the basis are stated | Judge on the basis corresponding to the test basis |
- Every figure has one source; note where it came from as you take it, and check once more when the draft is finished.
- Anything that cannot be reconciled across the specification sheet, the test record and the nameplate is not written into the article.
- The correction stays at the place of the original text and the original title is kept, with a statement of which item was changed and on what basis.
- Discuss the matter with values measured under the same test basis, and do not substitute adjectives for the basis.
A wrong figure has to be corrected, but the way it is corrected should make it verifiable, not make it look as though it never happened.
