Cheap DC-DC converter for QRP with batteries

My recent outtings with the mcHF on battery power have made me realize, once again, that there is always a compromise to be made when matching battery type to the rig's maximum safe input voltage level. Quite often, a 3S battery will be a couple of volts less than the radio's maximum level but a 4S pack will exceed the radio's rating.

The ramifications are minimal and may not seem to make a significant real-world difference but consider the 25% variation in maximum permissible input voltages of several QRP rigs:

• mcHF - 16v
• KX2, KX3 - 15v
• Anan-10E - 14v
• MTR-3B - 12v

A battery that is safe to use with an MTR-3B will underpower my mcHF, decreasing its 10-watt output to 6 watts.

I wanted a way to use the same battery with any of these rigs (and for charging my phone, tablet at 5V) so I built a 4S4P battery pack from 16 Li-Ion 18650 cells.

And to make it variable, a $15 buck converter that will convert a 5-32V input to a user-adjustable 1-30V output. It seems to do exactly that as long as the input voltage is at least half a volt higher than the desired output voltage.

My main concerns with using such a device had to do with resolution, accuracy and RF noise.

Two small pots allow output voltage to be adjusted and a current limit to be set. The voltage adjustment is a 20-turn pot, the current limit is 10-turn. Both were sufficient enough in resolution that I could easily set either in tenths of a whole number. And it stays there after repeated Offs and Ons of the device.

An onboard illuminated LCD agreed with my Fluke 87 for both voltage and current values. Two small pushbutton switches allow the converter to be turned on and off, and another switch allows either input or output voltage & current to be displayed. A long press of this switch causes power, in watts, to be displayed for either input or output.

My first test on the converter was to set the output to 13.8V and the current limit to 3 amps. I then used my battery analyzer as a variable DC dummy load, increasing the current draw until the converter tripped offline at 3.1 amps.

Next, I set the analyzer to draw 3 amps from the 17V battery pack and let her run. As the battery's voltage decayed over several hours, the converter maintained the output at 13.8V. This continued until the battery voltage dropped to 14.2V.

So far, so good - but what about RF noise?

DC-DC converts work by rapidly switching the incoming DC on and off, generating AC (square waves) which are then rectified into the desired DC level. Square waves are rich in harmonics, so how well would this thing work as a DC supply for a radio?

I set up my RSPduo near the converter, with a short lead attached to the receiver's High-Z antenna input and proceeded to tune around while the converter did its thing.

Sure enough, harmonics were found, and they were so intense that I could not distinguish exactly where the fundamental was. Fortunately for me, most of the noise is just below and in the AM broadcast band. From about 75 kHz up to 4 MHz, a signal is generated every 14.8 kHz, decreasing in signal strength as frequency increases.

These signals were completely gone by 6 MHz. This made me think that the converter would be okay for powering a radio as long as I didn't operate on 60 or 80 meters.

.

500kHz - converter OFF

.

500kHz - converter ON

.

Zoomed in a bit on 900kHz - converter OFF

.

Same display settings as above, converter ON, showing harmonics 15kHz apart

.

Upper part of 80m band - converter OFF

.

80m band - converter ON

So I did just that - hooked up the mcHF to my 80m dipole and powered the rig with the converter.

Tuning around, I could neither hear nor see any difference anywhere on the band between the converter being on or off. Of course the 80m dipole is much further away from the rig than the wire draped over it was. Further up the band from 80 meters - same thing, no problems or any indication of noise at all.

mcHF connected to 80m dipole. Left - converter OFF; Right - converter ON

I'm quite happy with the converter, especially given its cost and the degree of versatility it will allow in a variety of applications.
.
.