DC Input - Bipolar Output Switcher PC Boards for Small Projects

[mediatechnology]

For evaluation I'm ordering a couple of small PC board bipolar output switchers for use with DC wall-warts or USB power.

https://www.amazon.com/gp/product/B0752TRXDC



https://www.amazon.com/gp/product/B07SFLVGMJ



There is a third one available but it has no mounting holes which I consider a serious limitation.

We'll see if they're quiet enough...

[AnalogJoe]

We'll see if they're quiet enough...

If not, you can always spend tons on LC filtering....

[mediatechnology]

At least these, being PC boards, don't have to go in a hiccup mode to meet some dubious efficiency requirement.

That killed the otherwise excellent Meanwell triple output supply when they went from the P25A-series to the "green" GP25A.

[mediatechnology]

This one looks very promising. Still running some tests. The output ripple is low when preloaded at about 60 mA. The sweet spot for ripple is either 5V or 12V in. At 15V input over a narrow range ripple is higher: It looks like it's trying to makeup its mind whether to boost or buck. When unloaded its in hiccup mode.

The real test is to hook it up to a phono preamp.

https://www.amazon.com/gp/product/B0752TRXDC

[mediatechnology]

https://www.amazon.com/gp/product/B0752TRXDC



The above is the "DROK" brand switcher.
I used two 220Ω loads across the +15 and -15V rails to load it in 68 mA.
The specified minimum is load 30 mA but >60 mA yielded better results.
The input supply was a linear bench supply.
The input and output grounds are not galvanically isolated.

Some ripple oscillograms:


DROK Switcher Ripple 5V 460mA Input 68mA Load Eff 88.6%


DROK Switcher Ripple 12V 200mA Input 68mA Load Eff 85%


DROK Switcher Ripple 24V 100mA Input 68mA Load Eff 85%

The RMS ripple over a 4 MHz bandwidth is <7mV; at 100 kHz BW <3 mV.

The second supply in the original post has much higher ripple and I didn't test it.

[mediatechnology]

I scanned the small leaflet that came with the DROK supply and created a data sheet.

The note for the ripple specification "on bare mental" is something I've never seen before.



PDF: https://proaudiodesignforum.com/images/ ... asheet.pdf

[mediatechnology]

I was able to check the DROK switcher from Amazon with the Flat Moving Coil Preamp.

Gain 61.5 dB input termination 10Ω.
The measurement bandwidth was 48 kHz.
(Over a 20 kHz bandwidth the output noise is around-78dBu at 61.5 dB gain.)

The DROK was pre-loaded in 60 mA plus the FMCP load which add 80 mA to the positive supply and 50 mA to the negative.
The total currents are thus 140 mA on the +15 and 110 mA on the -15V.

The switcher supplying the DROK is 12V 2A.
Not every generic 12V 2A switcher will run the DROK; the startup current is too high for some and they go into overcurrent.
For this test I used the Meanwell SGU25U12 2A wall-wart type.
I would recommend sticking with Meanwell unless you want to hand-pick generic supplies.

There is a small hump from 25-30 kHz induced by the switcher making it appear to be spread-spectrum.
Some of the hump is conducted and some is radiated by the DROK switcher magnetics.

The 60 Hz and it's harmonics are environmental and radiated.
They were physically nulled by angling the board.

As you can see the FFT looks very good making the DROK a contender provided a quality 12V input switcher is used.
I also tried a Meanwell 5V 1A which would run it but it was highly-loaded with the output around 3.5V.
A higher-current Meanwell 5V should run it fine.



The FFT looks as good as if it were being run on the bench supply.


https://www.amazon.com/gp/product/B0752TRXDC

[emrr]

Nice find

[mediatechnology]

Thanks Doug.

I measured the peak start-up current for the DROK and it's quite high.

With only the pre-load the peak current is about 3.2A for 40-50 ms.
The FMCP preamp raises the current to 140 ma on the positive rail - with the added load the startup drops to about 2.3A.

Although the steady-state current with a 12V input is around 400 mA, or 4.8W, the switching supply feeding it would need to be about 3A to 4A or 36W to 48W to assure start-up.

Most of the 2A supplies allow overcurrent from 110-140% of rated load but all except the 2A Meanwell went into over-current protection.

Using a 24V supply only shortens the duration of the start-up current but not the peak value.
5V supplies have even higher peak currents.

EDIT:

Tried it with a generic 12V 4.5A "brick" switcher I had used with a Mini-PC and it would start-up.
Also tried a generic 12V 800 mA linear wall-mount supply and it would also start.

Have ordered a Meanwell 12V 3.75A switcher and a Triad 12V 600 mA unregulated linear wall-wart for testing.

My conclusion is that if you significantly oversize the switcher to around 4A this supply is useable.
There's not a lot of EMI radiated - the greater EMI that the preamp sees is ambient magnetic fields.

[scott2000]

Wonder how the old xbox bricks are? I run my security camera setup with one for years..pretty beefy..
Also some of the direct tv bricks are beefy too...And iirc were loaded with some nice parts.. I can pull them out tomorrow and see what they say??.
these are 12v btw..