A 2.5kW "Decca Clone" Class D Amplifier for 136kHz

 

A 2.5kW Class D Amplifier for 136kHz.

This is my version of the Decca Navigator Power Amplifier as featured in the RSGB Low Frequency Experimenter's Handbook.

Additional information together with circuit diagrams of the original unit is available from G3YXM's website.

 

My design basically follows that as per the above references, with the exception that I replaced the MOSFETs with high current- high voltage IRFP460 devices so that the unit could be powered directly from rectified a 230VAC mains supply instead of the 60 Volt supply used by the Decca amplifiers. (The 60 Volt supply would have been a Battery Source to provide a no-break service in the event of a mains power failure). The amplifier was fed directly from my 136kHz transmitter without any other signal conditioning with the exception of a 6dB attenuator to keep the MOSFET gate voltages to about +/- 10 volts.

This unit was used for my 2001/2002 Trans-Atlantic Tests during which it behaved flawlessly, although the power output was reduced to 1500 watts output so that it could run unattended throughout the night. (I like my sleep!).

This amplifier uses 4 MOSFETS in a full bridge arrangement - identical to a higher power switched mode power supply - with the gate drive arranged via a transformer. In my case this happened to be a ferrite ring core just because I had a suitable one available, although there is no reason why a conventional split core transformer could not be used.

One of the advantages of the bridge topography is that the MOSFETs will not be subjected to more than the supply voltage. With a push-pull design, the MOSFET's Drain will see twice the supply voltage during the "off" period, and thus either very high Vds MOSFETs must be chosen (expensive) or use "standard" switched mode power supply devices having Vds's in the range 400 to 600 Volts and limit the supply voltage to less than half of the MOSFET's Vds (i.e. 200 to 300 Volts), which invariably requires a heavy transformer due to the power levels. Feeding the Amplifier directly from the rectified A.C. mains supply nice avoids the need for a heavy transformer, and also avoids volt drops as the transformer is loaded. After all, switched mode power supplies operate directly from the rectified A.C. line so why not a transmitter amplifier? The high DC supply (~320V) also reduces the current drain through the switching MOSFETs which again simplifies matters. (During the Trans-Atlantic Tests I was using transformers to provide supply isolation so that I could monitor the MOSFET drains without damaging the oscilloscope).

During initial testing I did loose a couple MOSFETs from over dissipation. Normally the switching of the MOSFETs in this design is very clean, going from the on-state to off-state and vica versa very quickly. Consequently the power dissipation from the MOSFETs is negligible and the heatsink, upon which the complete amplifier is assembled, remains barely warm to touch. However, if the load happens to be reactive, such in the event of a poorly matched or off-tune antenna, then the switching waveform as observed on a oscilloscope will have a discernable slope - which means the MOSFETs are starting to dissipate power and become hot. Once this had been recognised, I didn't loose any other devices. (The blue capacitors across the output transformer to the left in the photo are to provide a crude but effective matching network - due to the relatively low frequency, I prefer to use electrical power terminology and consider them as Power Factor Correction Capacitors).

I also lost a couple of MOSFETs due to failure of the protection diodes placed across the MOSFETs to limit voltage excursions either above the positive supply rail, or below the negative supply rail. I replaced the diodes with higher current devices and that problem was solved.

For further details of my 136kHz Trans-Atlantic tests, see my 136kHz pages.

Return to the Amplifiers Page

Page last updated 8 Jan 2002