Troubleshooting Flex-6500 Band Following With Tokyo Hy Power HL-2.5Kfx Amplifier

I’m a Flex-6500 user and am trying to enable frequency band tracking with my Tokyo Hy Power (THP) HL-2.5Kfx amplifier using strictly a serial cable solution between the Flex and THP amp. Thus far, I’ve had limited success, in that the THP will band track the Flex, but the 6500’s performance is so degraded as to make it unusable. Eventually, it locks up and must be rebooted. The THP has a 15-pin serial connector on back, and uses cables supplied with the radio. Cable schematics are provided in the manual: http://www.ae8i.com/THP25.pdf

I’m using a Kenwood cable of the type shown in the schematic in the above link, as the Flex uses Kenwood protocol. This cable is mated to a null modem, which in turn mates to this USB-to-serial converter cable: http://www.usconverters.com/usb-serial-adapter-xs880?zenid=6fa9b412ffddf82e069045f0cbf46f6e

This is the null modem I’m using: https://www.amazon.com/gp/product/B075XHWVSJ/ref=oh_aui_search_asin_title?ie=UTF8&psc=1

The null modem has been found to be necessary by those who have had success band tracking a Flex with the smaller THP HL-1.5Kfx amplifier, which has a 9-pin serial connector on the back panel. The schematic for the Kenwood cable supplied with that radio is described at the top of this page in the manual: http://www.ae8i.com/THP15.pdf

This is a screen capture of my Flex-6500 USB cable data log while the Flex is attempting to work with the THP amp: http://www.ae8i.com/DataTHPUSB.png

If the null modem is removed from the cabling, the 6500 will not band follow and there is no degradation of its performance.

I have successfully implemented band following between my Flex and Palstar AT-AUTO tuner (Kessler Engineering version).  This was as simple as connecting the Flex to the Palstar with the above USB-to-serial converter cable and making sure the Palstar was set to expect Kenwood protocol. The USB cable data log shows this: http://www.ae8i.com/DataPalstarUSB.png

When comparing the data logs where band following is completely successful and where it is not, it appears that a flood of TX data may be overwhelming the Flex and degrading its performance, possibly due to some fault in the RX response.