[ARC5] Switching DC-DC Power Supplies.

[Ben Hall]

Good evening all,

On 10/14/2015 6:04 PM, Phillip Carpenter wrote:
> Can you share the schematic for Martin's 350vdc converter?

While I'm replying to Phillip's message, I've read the following 
comments in this thread.

I'm not a lawyer and I'm inclined to think fair use allows limited 
sharing for educational purposes...but...let me get it built and see if 
there is anything worth sharing before we go down that path.  Sound 
reasonable?

In other news, my Chinese DC-DC converters from the e-place arrived.

The units I purchased:  <http://r.ebay.com/MfTEUQ>

Initial impression...the box is extruded aluminum, thin, but functional. 
  The end plates don't electrically bond due the white plastic that 
surrounds them, and the screws holding the end plates are so short only 
about one thread engages.  (yeech!)

The circuit board looks to be all-thru hole.  I can't see the bottom, so 
it is possible that there are some I can't see.  Workmanship on the 
electrical assembly is actually fairly good.

Unfortunately, the input/output terminals are marked in Chinese, as is 
the 8-pin chip visible from the top.  Some minor sleuthing was required 
to figure out what was high voltage output and what was low voltage 
input.  Polarity was indicated, however.

Hooked it up, cranked it up, the LED came on, and no smoke.  Holy cow it 
was set to 425 volts, which is too close to the 450VDC rating of the 
"brand I have never heard of" 10uF output capacitor for me!  I promptly 
adjusted it down to 250VDC.  ;)

Regulation appears good.  249.3 VDC no load, 249.1 VDC at 40 mA load 
(purely resistive).

At 249.1 VDC 40 mA output, input current was 1.11 amps at 14.0 VDC.

That's 9.96 watts output, 15.54 watts input, an efficiency of ~64%.

On the scope, the output was absolutely clean with no load, and had a 
6.8 Vpk-pk ripple with 40 mA load on the 465M with high voltage probe. 
(yikes!)  That ripple was a cross between a triangle wave and a sine 
wave (rising edge faster than training edge), with some high frequency 
crap riding on it.  The "sine-wave-ish" waveform was about 300 Hz.  The 
high frequency crap was very "impluse-like", IE: scope trace looked like:

____|\____|\____|\____

(yes, the bottom is flat, it should be curved, doing the best I can with 
ascii art, hahahah)

With a frequency of about 50kHz.  Those spikes ride about 3 volts peak 
on top of the 300 Hz waveform.

After about five minutes at 40 mA / 250 VDC output, the switching 
transistors were absolutely cool and the transformer was slightly warmer 
than room temperature.

The 300 Hz wave concerns me.  I almost wonder if that's something wrong 
with the feedback circuitry and this unit may not be stable under all 
input/output currents and loads.

I will play with it some more.  I expect to build Martin's circuit this 
weekend.  I have the magnet wire ready.  :)

thanks much,
ben, kd5byb