Release Note
C0FE0024_xxxx.sff

Fix/Solution:
In FPGA version x0346 and later, an ioAbort event still forces gpio "output" pins to "input" mode, which allows external resistors to pull pins from either high or low for a "safe" state. The resistors may or may not be present depending on the PCB assembly. However, during an ioAbort event, service writes to the gpio configuration register no longer cause momentary "output" direction. After the ioAbort event is cleared, gpio pins will resume their previous "output" state without software intervention.

Fix/Solution:
The FPGA code was corrected in version x0346 and later releases. This fix is required for systems using the Dual String or IN Port String topologies. This fix is NOT required for systems using normal OUT Port String or Ring topologies.

Description:
0x00 will be read for dataType NULL in the status and feedback packets.

Fix/Solution:
The fault recovery problems was corrected in FPGA build versions x0346 and later. The rcvMiiBuffer will correctly clear itself after an overflow. The xmitMiiReg will release xmitHoldRcvBuf within 8 clocks after a fault, preventing an rcvMiiBuffer overflow. 8 clocks will guarantee that there is minimum packet spacing between the packets. sqPll will ignore the second SYNQ packet after fault recovery (in case packet delays were effected in the cycle where fault recovery occurred). The rcvMii will correctly handle a valid packet with data length of 0. No software changes are required.

Description:
Add support for multiple packets to the rxMiiBuffer. This improves tolerance to network jitter. Prior to version x0341, the rxMiiBuffer could absorb up to 32 clocks (1.28 µS) of delay for a single packet, but only 8 clocks for multiple packets with minimum spacing. This change is required to meet theoretical worst-case network jitter in large networks (32 nodes). The improvement may also help at network startup, since the node's PLL phase error is larger than normal (as the PLL pulls into nominal lock). No software changes are required.

Description:
Add BranchRev register. This extends the existing 16 bits of FPGA version number with 8 bits of branch and 8 bits of revision ID.

Problem/Cause:
The Drive serial FPGAs would force the last word of upstream data from the drive to 0. Also, they may not have detected checksum errors in the serial data from the drive. In some drives, the last word and/or checksum were not used (these errors would have no effect). In the CD drive, the last word of the data was the upper word of the service channel response data, so the upper 32 bits of any service channel read were forced to 0. The FPGAs using the driveInterfaceSerial module are: C0FE0018_xxxx.sff, C0FE0024_xxxx.sff, C0FE0027_xxxx.sff, C0FE002E_xxxx.sff. The last word problem was introduced at build version 0x0202. The checksum problem existed in all build versions prior to 0x0321.

Fix/Solution:
This problem has been fixed in FPGA versions 0x0321 and later. The FPGA code has been modified to correct the last word and checksum error logic.

Problem/Cause:
Service channel data sent to or from a drive, may have been corrupted. This was caused by an FPGA bug in crossing clock domains from driveClk to the internal 25 MHz clock. This problem could have occurred in any driveParallel designs (C0FE0014, C0FE0019, C0FE0031) prior to version 0x0320.

Fix/Solution:
The FPGA code was modified to prevent clock domain problems. This problem was fixed in FPGA versions 0x0320 and later.

Description:
Expand the dedicatedIn word from 16 bits to 32 bits and add the status bits Feedback 0 Fault and Feedback 1 Fault. Note: Although the FPGAs now supports additional bits, the MPI does not yet enable them.

Problem/Cause:
Clearing the driveWatchdogFault and driveChecksumError required two service comands—a write to start the clear, and a write to end the clear. MPI releases prior to the 03.01.00 did not clear the bits.

Fix/Solution:
Corrected the Drive Watchdog Fault and Drive Checksum Error bits to clear with a single service command. MPI releases 03.01.00 and later support this new behavior and clear the bits after a network reset.

Description:
Set unused pins to FLOAT after the FPGA configuration process. After a power-on, but prior to FPGA configuration, SynqNet FPGA pins should be high-impedance outputs.

NOTE: The mode pins can enable internal pullups, but this is not recommended for SynqNet.

After configuration, unused pins will be high-impedance outputs with no internal pull-up or pulldown. In earlier versions, unused pins would default to weak internal pulldowns. This was not ideal since it could have conflicted with an external pull-up.

Problem/Cause:
There was a problem where SqMax NodeStatus cyclic status bits may have been corrupted after reading some of the FPGA registers (via the service command). The bits that could have been corrupted are: analogPowerFault, ioAbort, nodeAlarm, and nodeDisable. This issue has existed since release x0100.

Fix/Solution:
This problem has been fixed.

Description:
Support has been added for a number of different flash component vendors. Flash components that are currently supported are:

Problem/Cause:
There was a problem with the inverted polarity option for the FPGA inputs nodeDisable and analogPowerFault. Prior to this fix, only the default polarity was functional. The problem happened because the hardware invert logic was inserted after the hardware latch function. Therefore, when "invert" was selected, the input would typically latch the negated state, and would not report a fault unless the latch was cleared by a software command. If the "invert" option was NOT used, the input would function correctly. This issue did not exist until the x020C release, where these inputs were first latched.

Fix/Solution:
This problem has been fixed.

Problem/Cause:
There was a problem in the sqPll function of the sqMac. With certain timing values, one node in the network may have fail to hold lock and then aborts cyclic operation. For a typical 3-node network, this timing typically occurred with TxTime set to 95%. Other TxTime values may fail in other network configurations.

Fix/Solution:
This problem has been fixed.

Description:
The drive current of GPIO pins has been increased to +/-24mA. Other outputs which may drive opto isolators (LED outputs, nodeAlarm, ampEnable, brakeApplied) were already configured for 24mA drive current. See individual FPGA specifications for complete details of I/O characteristics.

Problem/Cause:
There was a problem in the receive data unload function of the sqMax. In certain situations where the packet size did not match the configured packet size, invalid sqMax outputs were observed to cause "glitches" in the I/O output signals. The glitches would occur at the transition from cyclic to asynq operation, but the ioAbort signal would normally prevent these glitches from reaching external logic.

Fix/Solution:
This problem has been fixed.

Description:
Improvements have been made to the secondary encoder features. These changes allow the use of encoder faults and capture with secondary encoders. Secondary feedback for drives has also been added. The changes are NOT backwards compatible with software releases prior to 20031222 (secondary encoder feedback is not fully functional with prior software).

Description:
Improvements have been made to the ADC features to use the new "nodeIo" conventions. These changes allow up to 8 ADC channels per node (prior to this, the default was 1 ADC channel per motor). The changes are NOT backwards compatible with software releases prior to 20031222 (ADCs are not functional with prior software).