Don, thanks for posting!
Since your jpg will disappear in a few days, I have given it a permanent home at:
?
Looks like a very good solution to the problem of limiting voltage to the QMX at 12v.
I believe that circuit will have a quiescent current of around 1ma.
?
Could change the resistor divider to give 9v out, for example? ?R3=40k and R4=50k,
or any resistors of the same 4/5 ratio.
Given that the LM2950-5 is a 5v reference, we need that ratio to give
? ? ?9v*50k/(40k+50k) = 5v on the base of Q2.
If the 9v output exceeds that value, the base of Q2 rises above the base of Q1?
and raises the voltage at both emitters, resulting in no current through Q1
and a rise in voltage at the Q3 gate.
?
Q1, Q2 could be any small NPN's, such as a 2n3904 or the QMX's BC817.
Q3 is an enhancement mode PFET, it must be able to dissipate any excess voltage?
at the expected current.? For example, with 17v in, 12v out, and a 2a current,?
it will be dissipating (16-12)*2 = 8 Watts.
May need to be bolted to the back cover to dissipate the heat.
The LP2950-5 is a 5v voltage reference, could be replaced by a 5v zener and a 10k series resistor
at slightly increased quiescent current and somewhat poorer regulation.
?
Perhaps we could have a PC board with pads for all the parts.
Gets mounted to the back panel by the tab of Q3.
?
I doubt many of us have signed up for SPRAT, perhaps we should.
?
Jerry, KE7ER
?
?
?
On Fri, Mar 14, 2025 at 02:09 PM, Don, ND6T wrote:
Another option is a Very Low Drop Out regulator like G4COL described in the latest SPRAT (issue 201) page 12. More efficient than a buck regulator, all analog, and capable of huge currents.
