Rewind test report packages, assembled from your shop records for your qualified person to review and sign.

A rewind makes its paperwork in pieces. The incoming test data and core loss printouts sit with the shop order. Winding data is on the sheet the winder filled in while taking out the old coils. The balance record comes from the machine shop and final run data from the test panel. Someone in the office gathers all of it, puts it in the form the customer asked for, and chases whatever never got written down before the motor ships.

In the test, AI works from copies of those records and assembles the report: core loss test, winding data and insulation class, resistance and surge test fields, balance record, final run data, and a list of the records it couldn't find. Every value is copied as recorded, with the page it came from. It doesn't judge any result, and nothing gets signed.

The trade says what a full file holds. ANSI/EASA AR100, the recommended practice for repairing rotating machines, has a shop open a record when each machine arrives and keep it at least three years: nameplate data, electrical tests and mechanical measurements from before and after the repair, original and final winding data, the parts replaced, and the primary cause of failure where it can be found. Its 2025 edition adds that photographs of all sides help when comparing the as-received machine at final inspection. A report can leave out a before value and still look complete, so the draft sets each item against that list and names every gap.

Core loss is a before-and-after record. AR100 wants cores tested before burnout and again after the winding is out, with the results compared, and keeps core temperature to 700 degrees F. Tulsa's Water and Sewer Department caps its burnout at 650 degrees F, wants the figure in watts loss per pound, and asks for a second reading after burnout to confirm the same data as before. The Department of Energy's 1999 model specification has exciting current and watts loss recorded both times at a stated flux density, and holds the second test until the core is cleaned and dried. Pair the wrong two readings and whoever signs is misled, so each value keeps its stage, its unit and its date.

Which tests ran depends on the winding and the buyer. AR100 calls for insulation resistance, surge and winding resistance on stators and wound rotors, insulation resistance before any high-potential test, and every result kept. It recommends a ten-minute polarization index of at least 2.0 for Class B and higher, citing IEEE 43, the recommended practice for insulation resistance, and notes that a random winding may need a 60/30 second ratio instead. Erie County, New York, wants surge comparison before and after the shop test run on a rewind, and before dismantling on anything else. So a blank field can mean nobody ran the test or that it didn't apply. The draft reports which, where your records show it, and never marks a reading good or bad.

Winding data is the part a customer can't check. Tulsa wants wire size, number of turns and cross-sectional area measured off the old winding and put in the final report, with equal or greater wire area in each slot. The Department of Energy's model specification bars changing turns per coil, span or connection without the purchaser's authorization. On insulation, AR100 asks for a system equal to or better than the original, a 1999 guide from the Electrical Apparatus Service Association notes that many shops upgrade from Class B to Class F or H, and Tulsa and Erie County both call for Class H. Temperature sensors may be left out only with the customer's consent, noted in the record. The draft sets old and new side by side and flags any difference without saying which is right.

The last pages carry the signature. AR100 balances to the level the customer names, otherwise to grade G2.5 at 2,500 rpm or slower and G1.0 above that. Erie County asks for dynamic balance from 25 HP up, with any coupling half back where it came from. Tulsa's run list covers no-load current for each phase, bearing temperature, vibration in three directions at both ends, speed and magnetic center, and above 1,000 volts it wants new time-current coordination curves or a certification that the old ones hold. AR100 expects the instruments behind those numbers to be calibrated at least annually, traceable to NIST. Then someone signs: Erie County wants as-found and as-left values reviewed and signed by the shop's engineer or foreman, and AR100 has the person the shop authorizes enter something like OK to ship in the record. Your reviewer, usually that person, checks that every value landed on the right motor and stage and every gap is real. None of this is legal or engineering advice.

With your reviewer's yes, each new rewind gets the same assembled report for about 30 days, while there's still time to find a missing record.

A qualified person reviews and signs every report, and nothing reaches a customer on a draft alone.

You'll have a fixed price for those reports in hand before you sign. Any further job is priced on its own, with monthly support after.