5w Power Amp

Here's my Mitsubishi RD06HHF1 power FET amplifier as I use on QRSS. It requires no amazing construction technique at HF, and I have even use it at 70Mhz. The RD06HHF1 FET is actually capable of up to 20w output when given a drive level of about 250mw and can be run on a higher voltage, up to 30v. It's quite safe. and perfectly permissible to do so. The "06" in the specification refers to the maximum power output that the device will survive when operated into a bad load, short or open circuit.. It is NOT the maximum power output, which is nearly 20 watts.

At ALL times, even at QRP levels, ensure that you have adequate heat sinking (I use the aluminium box).  Above 5 watts I recommend a small snail fan blower to blow in colder air. They usually run on 24v DC and can be purchased on Ebay for about £5 delivered.  Seek too keep the heat sink temperature to  a maximum of 65 degrees Centigrade. Just about ALL semiconductors start to fail above 70-80 degree Centigrade, even military spec ones. You've been warned !  These figures and all tests have been when using the amp in beacon mode at almost 100% duty cycle, for 24 hours a day, 7 days a week and for several years now.  I think it's safe to say that this amplifier passes the stress test.

My amp has a variable voltage PSU in it capable of 1 amp at 30 volts DC.  I like to use LM338 5 amp regulators for better heat dissipation. A voltage and current monitor (used about 450ma at 15v on 10m) if I push it to 5 watts. (I don't use it this way, I only used it to boost the tiny output power of the TX to the 500mw level). The output stage uses QRP Labs plug in filters.  Cheap polyvaricon capacitors are used in the ATU (they'll handle 20 watts into a match load, but 10w into jut about anything else).  An Led-O-Meter is used a basic output level indicator.  NOTE THAT THE OUTPUT USES AN UN-UN AND NOT A BALUN.  My antenna wires on QRSS connect directly into the black and red speaker terminal sockets as seen on the top right. An SMA patch cord allows me to break into the path between the output of the PA and the internal ATU should I require it for testing and measurement purposes. The SO239 socket on the top is the input to the amplifier, that also goes through an on board attenuator. I can see the input to either 0 / 6 / 10db attenuation to make it easier to match the levels of several of my transmitters. 

I use about 2.4 volts bias on the input gate for optimum results. Unlike IRF510's, the required gate voltage is remarkably consistent with each device. IRF510's can vary a LOT as other have discovered, so they all tend to require each device to have it's own select on test gate voltage on each build. Never use excess gate voltage bias or you'll lower the drain to source resistance too low and it can draw too much current and destroy itself. This also applies to the RF input level. 5v peak to peak should really be the maximum.

If you keep the peak to peak RF input levels to within the 2.4 volt Gate bias then the amplifier will operate in a linear mode. I have tested this by feeding an AM transmitter into it and it sounds great.  

Some notes on gate bias arrangements.  You can ruin a project by having a poorly designed gate bias arrangement. It is essential that RF that is supposed to be fed into the gate does NOT travel up the feed of the gate bias line. My technique is to use something like a 2mH choke (good down to about 1.5Mhz) in series with a 10K Ohm resistor.  Fit these physically  right up close to the input gate.  The choke should have a highly inductive reactance at the lowest operating frequency, aim for 500 - 1000 Ohms. The resistor increases that RF stopping effect, as well lowering the current drawn by the gate should there be a fault condition.

Note, ignore the warning about insulating the tab with a Mylar fixing kit. This is wrong since the tab is Earth anyway. You can safely bolt the tab to chassis ground.

Here's a general picture of two of my QRSS rigs up in the loft. One for 40m and the other smaller one is for the 10m band.

You could also build a similar amplifier using IRF510's (see notes above regarding bias). IRF510 have a reverse Drain and Source pinout.  For Mitsubishi data sheets, IRF510 data sheets and constructing un-un for output stages have a look at the resources link here.