Introduction

This document contains the release notes for WITE revision 1.44 from Guzik Technical Enterprises. Revision 1.44 fixes problems with the SSI4935/4937 PRML chip, high-frequency ADC calibration, spinstand alignment and some head amplifier chips.

All discussion in this document uses WITE version 1.43 as a baseline for comparison. For information about features and modifications prior to 1.43, please refer to the WITE user’s manual and to the release notes for previous revisions of WITE.

New features in WITE 1.44

Mirror head loading mechanisms are supported.

The SSI 32R1606AR head amplifier is supported.

Bugs Fixed in WITE 1.44

The "Cannot calibrate delay" error message could appear during the ADC delay calibration if the frequency was higher then 160 MHz. As a result, partial histograms of SVD could not be built. ADC delay calibration works correctly at all frequencies in version 1.44.

If the acoustic sensor was activated during spinstand alignment, incorrect jogging could be performed that could cause a head crash. WITE 1.44 interrupts alignment correctly.

If the distance between the reset radius and loading position was too small, the spinstand could stop responding to software commands. This problem is fixed in version 1.44.

When testing head stacks with TLS24M558 head amplifier, the head selection worked incorrectly. This bug was introduced in WITE 1.4 and fixed in WITE 1.44.

Head stacks based on Marvell 81M3006 head amplifier were not initialized properly. These head stacks are correctly initialized in version 1.44.

Chip optimization for the SSI 4935A and 4937 PRML chips intermittently produced poor values in WITE 1.43. In WITE 1.44, the optimization procedure is stable and produces correct results.

Known bugs in WITE 1.44

Some bugs were found too late to be fixed for the 1.44 release and will be fixed in a later release. This section contains all known bugs that you are likely to encounter. A bug’s presence here means that we are aware of the problem and actively working to solve it. If you encounter a problem not on this list, please report it to Guzik customer support.

If the track profile test is used with the "use fast offset" option enabled and more then one iteration specified, one or more points may fall below the real TAA value. If you need more than one iteration, do not use the "use fast offset" option.

The "retest with next setup" option in the "Production | Grd/Norm" menu does not register results to the database after the first setup. It should not be used.

Some DOS applications interfere with communication with Guzik hardware and can result in unstable test results. Please exit all DOS applications before running WITE.