While the simple protection circuit described here will not prevent transistor failure in all cases, for AM use with carrier, it will save you 99% of the time. Ready? Install one 20 amp fast acting fuse inline feeding the DC into the center tap of each output transformer. With an 8 transistor amplifier you will be adding 4 fuses. Connect the load side of each fuse to the input of a digital "AND Gate". The AND gate is an IC that requires all inputs to be high before the output will turn on. You'll use something like a 10K dropping resistor to feed the AND gate inputs from the fused 12 volt lines. Put the resistors on the protection board with the gate, so .01uf bypass caps can take advantage of the resistors to remove unwanted RF, prior to the gate input.
The output of the AND gate should be buffered with a transistor just like the one used in the RF keying circuit if it's an NPN supplying ground to the relay. Add the second transistor between the keying transistor and relay ground side. With an NPN keying circuit, you should keep the original tied to ground since it senses RF to ground. If the keying transistor is in the positive leg of the relay, just add the NPN being driven by the gate, into the negative side.
Now all output transistors can be individually current limited. Since only one at a time draws current in each bank, you fuse as though only one transistor was in each bank. While the old 2879 was rated at 25 amps, you need a 20 amp fuse because they do not open at 20 amps and the next size fuse at 25 amps, will cause the transistor to "replace" the fuse. Because the fuse has to melt, it should not be expected to protect against extreme changes in current like we see on SSB or those who use AM with little to no carrier...
You need to "warm" the fuse with the AM carrier to keep it "ready" to open on brief excursions beyond the safe point. This was my earliest form of protection circuit learned through a digital electronics course in high school. It passed all of the testing I could give it on AM without any problem and almost did the same on SBB, tripping every time I brought the drive up too much. But, one key on SSB with too much drive and one sharp peak being the first thing to hit the amp, blew one of 16 transistors and earned the circuit an F- for SSB performance!
Protection can be dramatically improved if the amplifier uses an internal driver that is not being maxed out. Apply the same protection to the driver stage but use a smaller fuse that will only handle the current drawn, when properly driven. Now, the driver will be the first to trigger the protection, long before the finals are placed under any stress. When any bank draws more than the safe current, a fuse opens. One input to the AND gate goes low. The output of the AND gate turns off, unbiasing the NPN transistor and releasing the keying relay to "barefoot" operation. As a teenager, this circuit almost got me beat up by the local technician when I started selling it to his "frequent flyer customers".
The output of the AND gate should be buffered with a transistor just like the one used in the RF keying circuit if it's an NPN supplying ground to the relay. Add the second transistor between the keying transistor and relay ground side. With an NPN keying circuit, you should keep the original tied to ground since it senses RF to ground. If the keying transistor is in the positive leg of the relay, just add the NPN being driven by the gate, into the negative side.
Now all output transistors can be individually current limited. Since only one at a time draws current in each bank, you fuse as though only one transistor was in each bank. While the old 2879 was rated at 25 amps, you need a 20 amp fuse because they do not open at 20 amps and the next size fuse at 25 amps, will cause the transistor to "replace" the fuse. Because the fuse has to melt, it should not be expected to protect against extreme changes in current like we see on SSB or those who use AM with little to no carrier...
You need to "warm" the fuse with the AM carrier to keep it "ready" to open on brief excursions beyond the safe point. This was my earliest form of protection circuit learned through a digital electronics course in high school. It passed all of the testing I could give it on AM without any problem and almost did the same on SBB, tripping every time I brought the drive up too much. But, one key on SSB with too much drive and one sharp peak being the first thing to hit the amp, blew one of 16 transistors and earned the circuit an F- for SSB performance!
Protection can be dramatically improved if the amplifier uses an internal driver that is not being maxed out. Apply the same protection to the driver stage but use a smaller fuse that will only handle the current drawn, when properly driven. Now, the driver will be the first to trigger the protection, long before the finals are placed under any stress. When any bank draws more than the safe current, a fuse opens. One input to the AND gate goes low. The output of the AND gate turns off, unbiasing the NPN transistor and releasing the keying relay to "barefoot" operation. As a teenager, this circuit almost got me beat up by the local technician when I started selling it to his "frequent flyer customers".
Last edited:
