If you encounter repeated or missing keystrokes on your UHK 60, and the general troubleshooting instructions haven't helped, it may be worth lowering the i2cBaudRate value.

Install the latest Agent, go to the About page, click the Agent icon at the top 7 times, go to the newly shown Advanced settings page, and enable the "I2cRecovery debugging" option. The output should look like this:

2022-03-19 08:18:39 I2cRecovery: 0

The I2cRecovery value increments when communication errors occur within the UHK. You can trigger such errors by reconnecting the left half or the modules, and then the value will increase:

2022-03-19 08:24:32 I2cRecovery: 1
2022-03-19 08:24:36 I2cRecovery: 2

Under normal circumstances, the I2cRecovery value should be low and not increase. But when the UHK is exposed to electrical noise or electromagnetic interference (EMI), it continues to increase, potentially resulting in repeated keystrokes (chatter), missing keystrokes, or phantom keystrokes.

If the I2cRecovery value keeps incrementing, start with the simplest setup you can create. Merge your UHK halves, connect your UHK directly to your computer, disconnect other USB devices, disconnect other power appliances from power sockets, and disable the backlighting in Agent. Then watch how the I2cRecovery value is affected as you split your UHK halves, add more USB devices, connect power appliances, and generally restore your environment. Please also try the suggested points at the top of this page while watching the I2cRecovery value to find the cause of the issue.

If the I2cRecovery value keeps increasing, no matter what, then you can reduce the communication speed of the internal I2C bus of the UHK, making it more stable. The default value is 100000. You can halve the communication speed by running the following smart macro commands, preferably in the $onInit macro:

set i2cBaudRate 50000

which should make communication more stable, but the smaller the value, the less responsive your UHK will get, which you will notice below a certain value.