I'm still pondering the output power of my transverter. It's designed around an E class PA in theory, so it was time for me to learn a few more things. For your Advanced license you're only required to be familiar with the basics of A, B and C classes of amplifiers. However, out in the real world people also regularly consider D, E and F.
So before I post up about some findings with my transverter, here's a run down (by a newbie) on the various classes of amplifiers.
First, you can split the six classes into two groups:
Linear amplifiers: A, B and C
Saturation amplifiers: D, E and F
The term linear here is used in a slightly different fashion compared to when people refer to A being linear and B and C being non-linear. Here, were simply meaning linear amplifiers operate for the key part of their time in the linear region of the amplifying device (BJT or FET). Conversely, the opposite to the linear amplifiers are those that focus on operating from cutoff to saturation – sometimes also as a result called switch mode amplifiers.
With linear amplifiers the main differentiation is the duration for which there is current flowing from the output:
A – 100% of the time
B – greater than 50% of the time
C – less than 50% of the time
That's all pretty simple. For saturation amplifiers they're primarily classified based on their output network. But before we get on to that, what exactly do we mean by saturation amplifiers?
Classes D thru F rely on three key things for their input/driving waveform:
1. A square wave (ideally);
2. With minimum voltage just below the active devices threshold – so for an IRF510 FET the Vgs(th) is 4V so commonly the minimum voltage is 3V; and
3. A maximum voltage just past the saturation point of the active device – so again, for an IRF510 FET this is 8V
In this scenario then, you can have a 5V Vpp square wave set with a bias of 3V driving an IRF510 FET and you'll have either a class D, E or F. Which of those it is, depends on the output network:
D – Treated no different to if it was a class C setup (the key here is that the way the amplification device is driven is different to a normal class C);
E – The output is made up of a inductor on the drain and a capacitor in parallel so as to form a tuned network (thereby a low impedance circuit) taking into consideration the FETs output capacitance (Coss), all of which then feeds into a normal LPF;
F – This one replaces the LPF with a combination of tuned circuits to capture at least the 3rd harmonic (and potentially the 5th) and then feed the voltage back in phase with the fundamental – fiddly stuff and understandably a bit hard to get going (but apparently used in commercial kW transmitters).
And a reminder that one of the key drivers for the increasing number of classes of amplifiers is efficiency. A class is at it's theoretical best 50% efficient, but normally in the 40s. When you get to C you can get around 70%. With E and F there are real world examples at 90% efficiency and above.
Of the switchmode amplifiers, E is probably the most popular with amateurs. F is great for commercial setups, and D was originally targeted at HiFi systems with a PWM input source. That said though, with D being so similar to C it is considered a good starting place to start playing with these types of amplifiers.
All one needs to get started with class D, is work on generating a square wave as the input waveform rather than the sine wave we normally use. This can be done with various CMOS and TTL chips.
That should be enough for an overview, and being new to this myself there could be several errors. But, search the internet and you'll find far more authoritative sources. For me, I relied on a document by Paul Harden NA5N – "The Handiman's Guide to MOSFET 'Switched Mode' Amplifiers" (originally published in the journal "QRPp"). I was also hoping there'd be information in the ARRL book "Experimental Methods in RF Design", but there was only a very very short mention that class D and E amplifiers exist – need to check the ARRL handbook yet.
With this in mind, I'll post up some interesting bits on why I think my transverter is producing the low power it is.