1
\$\begingroup\$

Given that Class E RF amplifiers, even those that generate less than a few watts, can easily generate voltage spikes approaching (or exceeding) 100 V, how can they be homebrew built and experimented on safely?

enter image description here

Generally, the rule for anything over 30 V is to use it only in a well protected enclosure that prevents skin contact. This isn't possible for simple homebrew projects, which:

  • Lack such enclosures
  • Need to be measured and perhaps tweaked while operating
  • Involve exposed wires, contacts, terminals

This question isn't limited to my design. Take, for example, the popular QCX+ transmitter. According to its docs, it uses a Class E amplifier that I'd estimate creates voltage spikes exceeding 100 V:

enter image description here

L4 is a 1.4uH inductor, 19 turns on T37-2 core.

\$\endgroup\$
1
  • \$\begingroup\$ Did you check the BS170 datasheet? I think you'll have a hard time developing quite that much voltage here. \$\endgroup\$
    – Tim Williams
    Commented Jun 26 at 23:19

2 Answers 2

Reset to default
1
\$\begingroup\$

OP's circuit does generate voltage spikes at MOSfet drain that can considerably exceed the DC supply voltage. Switching current off quickly to an inductor is the cause.

The circuit fragment of the few-watt QCX+ transmitter has cut off the critical section that limits voltage spikes: the low-pass-filter (LPF). Only one of the LPF filter capacitors is shown (C30+C27 are in parallel).
Low-pass-filter inductors and capacitors are chosen carefully so that when those three BS170 switches simultaneously switch off, drain voltage arcs up to perhaps +40V before arcing back down to 0V. The 40V peak is indeed above the +12V DC supply, but within BS170 voltage limits.

Things can go awry if the 50 ohm load at the low-pass-filter's end is missing. When missing, energy temporarily stored in the low pass filter's inductors and capacitors can potentially appear at MOSfet's drain as a much larger spike.

\$\endgroup\$
2
  • \$\begingroup\$ I understand how having a load resistor, esp. with relatively low impedance, prevents an excessive voltage build up. But how does putting a LPF in front of the load contribute to avoiding voltage build ups? Can it do so without decreasing output power? \$\endgroup\$
    – SRobertJames
    Commented Jun 27 at 3:16
  • \$\begingroup\$ Class-E transistors operate as a switch: discontinuously on/off. LPF not only attenuates harmonics, but also smooths discontinuous operation. It can also transform impedances. Current in QCX+ L4 comes in pulses, voltage across C30 in pulses too, but current in 50 ohm load is continuously sinusoidal. Pulse power entering L4 is efficiently delivered to 50 ohm load. \$\endgroup\$
    – glen_geek
    Commented Jun 27 at 13:15
2
\$\begingroup\$

basically be very carefull Follow the same advice given to those working with vacuum tubes. rubber shoes and one hand behind your back.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.