Juma PA-100D Version: 4.00a

[5B4AIY]

NOTE: AS THIS VERSION USES AN ADDITIONAL WORD OF EEPROM, IT WILL CAUSE A CHECKSUM ERROR ON INITIAL POWER-UP! MAKE SURE YOU NOTE YOUR EXISTING CALIBRATION AND CONFIGURATION DATA PRIOR TO LOADING.

This version of the firmware introduces three new features:

1. The Low-Voltage Pre-Limit is now adjustable;
2. The amplifier's band change can now be controlled by the Yaesu 5-Byte Binary command protocol.
3. The frequency sense menu selection now has a Manual mode.

Adjustable Pre-Limit
Previous versions of the firmware gave a Low Battery warning if the input voltage was 100mV above the Low-Voltage trip setting. There have been some instances where this setting could be considered slightly too close to the final trip voltage, and, rather than simply increasing it by an arbitrary value I decided to make it user-adjustable. You can now choose in the Calibration menu to set the pre-limit to be anywhere from +100mV to +800mV above the final low-voltage alarm setting. Note also that if you make a change to the low-voltage alarm then this will automatically update the existing pre-limit. For example, if the initial setting of the low-voltage alarm is 11.00V, and the pre-limit is set to 11.20V, then if you change the low-voltage alarm to 10.50V you will find that the pre-limit setting has now been updated to 10.70V, 200mV higher, as was previously set. Of course, you are free to modify this setting as you wish.

Yaesu 5-Byte Binary
The previous versions of the firmware could accommodate the Juma TRX-2 protocol as well as the Elecraft KX3. This latter transceiver uses the ASCII protocol, and the standard frequency query command in this protocol is FA; Thus this setting can be used with any transceiver that correctly responds to this command. I have now added the Yaesu 5-Byte Binary protocol which is used by the Yaesu FT-817 and FT-818 transceivers. You will have to use the CT-62 level converter to connect to the transceiver's serial port and convert its TTL voltage levels to standard RS-232 levels.

I have also closed a 'window of opportunity' whereby it was possible for the amplifier to remain with an incorrect band selected. If you are using any of the three serial protocols, Yaesu, KX3, or Juma, and the transceiver failed to respond, then it was possible for an incorrect band to be selected, especially if a manual band selection was made. In this version, if there is more than a 200mS delay between received characters then this will cause the amplifier state to change from OPERATE to STANDBY, and the RF will be shut off. In addition, if a manual band change is made, this will force a frequency query command to be sent to the connected transceiver and the response from this command will over-ride any manual frequency change. Similarly, if the mode is changed from MANUAL to AUTO or vice-versa then this will also force a frequency query command to be sent, and the response used to select the correct band. This will occur even if polling is disabled. The 200mS time was chosen to comply with the Yaesu CAT commands for the Yaesu FT-817/818 which state that there can be up to 200mS delay between characters.

In the case of the Elecraft KX3, it is possible to set the transceiver to AUTO-INFORM, and in this case it will eventually send its new band information to the amplifier, but this can take several seconds. In this version of the firmware, if either the Yaesu or the KX3 mode is selected, then this will automatically enable polling and set the interval to 2 seconds. You can of course change this as you wish. As the Juma TRX-2, using my firmware, automatically sends band information whenever a band change occurs, as well as every time the transceiver enters the transmit state, then polling is unnecessary and if this mode is selected, then polling is automatically turned off. Again, you can enable it if you wish.

The debug messages that were sent if a serial terminal was connected whenever either the Calibration or the User Configuration settings were saved to the EEPROM have now been removed. They were really only there to confirm the operation of the Cyclic Redundancy Check feature and once that had been verified, these messages no longer served any useful purpose. Indeed, if the amplifier was connected to the Juma TRX-2 and its serial port was also in the TEST mode these messages would provoke a rash of error responses, which would cause the amplifier to respond similarly and a 'Deadly Embrace' lock-up could occur which would eventually cause a stack over-run and crash the operating systems. Since these messages no longer serve any purpose and would only be of interest if you dumped the EEPROM and compared the CRC with what had been saved, then I decided that they had now outlived their usefulness.

There has also been a minor change to some of the default values that would be loaded on initial start-up or if you had to restore the default settings. In addition, the filter frequency change points have been altered to coincide with those of the Juma TRX-2 transceiver.

As the incorporation of an adjustable pre-limit required an additional word of EEPROM memory, THIS VERSION WILL CAUSE A CHECKSUM ERROR ON INITIAL START-UP, AND THE STANDARD DEFAULTS WILL BE LOADED! Please make a note of your existing Calibration and User Configuration settings prior to loading this firmware. There are pages at the back of the User Manual set aside for this data. This checksum error should only occur once on initial start-up. Indeed, the CRC used is extremely robust for the amount of data being saved, and if you were to ever encounter a checksum error after a normal start-up this would indicate a serious error or corruption of the EEPROM data memory. I have only ever heard of two instances of this in over 8 years of supporting the Juma TRX-2 and PA-100D, and in both of these cases it was caused by power line glitches. The equipments were being powered by switch mode power supplies that had only a short hold-up time, and this caused a fast reset which initiated a wild write to the EEPROM, corrupting its data. It is easy to recover from this, but it pays to ensure that your equipment is powered from a power bus that is not shared with transient inducing loads such as industrial compressors, or refrigerators.

Build 3 removes an inconsistency when selecting a manual band. I discovered that when using the Elecraft KX3 if the 6m band was accidentally selected, then in the AUTO mode the amplifier correctly displayed the "?" symbol indicating an unknown band, and the amplifier would not go into transmit. However in the MANUAL band select mode it would select the last known manual band. This has now been fixed. In all cases, if an illegal or out-of-band frequency is selected on the connected transceiver, then the amplifier will display the "?" symbol, or the "!?" symbol, and if it is in the OPERATE state and an attempt is made to transmit, then the "Err" (Error) symbols will be displayed.

Build 4 adds a Manual frequency selection mode. With all the built-in protections against an inadvertent incorrect band selection it was almost impossible to actually manually select a band. To be honest, I cannot imagine why you would ever need to manually select a band with all the automatic protocols available, but, just in case there is actually a QRP transceiver that cannot utilize any of the available automatic modes, I have added a purely manual band selection mode. This mode should only be used as a last resort, as there are NO protections available. In particular, if you select a low-frequency band and you transmit on a higher frequency then you are likely to severely damage the low-pass filters, and possible the PA transistors. You have been warned!

Adrian, 5B4AIY

(I obviously had far too much time on my hands in the COVID-19 lock-down!)