15V 6A (90W) power supply

[JBeale]

I got a "15V 6A" DC power supply that appears to be this one "Model RC-218":
https://www.amazon.com/Adapter-100-240V-Switching-Charger-Voltage/dp/B073S4M1YV

it came bundled with a separate LiPo smart charger. The charger is fine, but the power supply gave up during the second use, at moderate load (about 2 amps).  Now it has no power light, no AC draw and no DC output. Not worth my time to fix it, but I'm still curious.

The circuit inside looks very similar to this one: https://datasheet.lcsc.com/szlcsc/Sifirst-Tech-SF1560SGT_C261340.pdf  except that it uses a different DIP-8 PWM controller marked CR6842 (datasheet: http://www.agspecinfo.com/pdfs/C/CR6842T.PDF ).  The circuit board has space for a line input choke, but it's not stuffed. The input full bridge & filter cap are fine. There is one power FET to drive the switchmode transformer, a STF10NM60N and I checked that it's fine for DC performance at least (Vds = 0.53 ohms at Id = 1.0 A, Vg = 5.5 V which matches the datasheet nominal).  However plugged in, I see the CR6842 pin 7  Vdd = 6 V and the chip wants 15 V nominal, so something's wrong there.

It's a bit tricky to understand how this circuit starts up because if there's no FET switching action, there's no AC voltage on any coil of the transformer, and there should be no power to the PWM controller at all.  I guess at the moment, it gets the 6V just from the leftover AC power line ripple (?)  I suppose there's an initial surge when plugged in though (there's no external power switch).

[shakalnokturn]

Check:
R2A, R2B (start)
D6, C3, R8 (run)

[mariush]

I sincerely doubt that power supply did 6A ever in its life.

It's most likely a power supply recycled from a laptop adapter and put into a new case.
You can see on the bottom on the circuit board the date, 2013-12-18  and model RZC-218 ... 218 would make me think it's meant to be a 2A , 18v power supply
It looks like they messed with R13 and R17 and maybe other things... there's flux residue and extra solder (and seems shinier) there.

You can buy wallwart adapters with the suitable barrel jack (yours is probably 2.1mm or 2.5mm, standard diameters) from around 8$: https://www.digikey.com/short/pnmt2v - pick your voltage and total power and the right connector.

 

[JBeale]

Thanks for the replies folks!  All the labelled passive components and diodes appear to be OK and/or within spec, as far as I can test them in-circuit. R2A, R2B are marked 750k and they are within 1%, measuring 744k and 743k.

For the price, I didn't think 90 W was likely. The DC output cable uses 20 AWG wire, not a great choice for 6 A, but at least some internal components are plausible:

The input bridge rectifier is KBP310 rated for 3A average current.  At 120 Vac input that could exceed 90 W (although might need it's own heatsink?)
http://vakits.com/sites/default/files/KBP310%20datasheet.pdf

The power FET is STF10NM60N rated for Vds = 650V and Id = 10 A and it's heatsinked.
https://www.st.com/resource/en/datasheet/stp10nm60n.pdf

The transformer I don't know anything about. On the other side, the DC output rectifier is SG20TC10M rated for 100V 20A, also heatsinked.
https://www.mouser.com/datasheet/2/366/J533_SG20TC10M-473177.pdf

The other parts AFAIK mostly don't affect total power rating, maybe allowable ripple voltage on the input & output caps.  I admit the actual parts might be cheaper knockoffs that don't match the mfr numbers printed on them, but so far I can't point to anything obviously giving the lie to the label specs.

EDIT: the transformer secondary resistance is 10.88 milliohms (54.4 mV at 5.00 Adc).  If we had 9A RMS current, then I2R loss in the secondary winding would still be less than 1 watt.  On the output side it's a half-wave rectifier design.  Vf across the output rectifier diode is 0.65 V at 6 A DC, dropping over time as temperature rises.  Per its datasheet, this diode will have about 4 watts dissipation in this design, assuming 6A DC output. The diode heatsink fin has 44 cm^2 of area. By the 6 cm^2 / W rule of thumb for open air heatsinks, that's good for 7 watts, but this is inside a larger sealed case, so there's a question there. There's an LC output filter, the L1 resistance is 7.5 milliohms, so that's about 1/4 watt at 6A, it looks large enough for that.

As best I can tell this design might well be capable of 6A DC output for a short time, but long term it is likely thermally limited by dissipation of the output diode. Looks like the designers knew that because there is actually a footprint on the PCB and a mounting hole on the heatsink for a second output diode in parallel, which is not stuffed on my unit.

[JBeale]

Followup: I did not find any other problems with the supply. Today, finally I received some replacement CR6842T DIP-8 controller chips (ebay). The new chip has the same part number, but the markings are a different style. I removed the old one, swapped in the new chip, and the supply sprang back to life. I tested it under load and measured 15.59 V at 0 A, 15.42 V at 2 A (34.6 W AC in, 30.9 W DC out = 89% eff), 15.26 V at 3.67 A (94.3% eff), 15.09 V at 5.34 A (85.4% eff) and 15.01 V at 6.09 A output with 110 W AC in, for 83.1% overall efficiency at full load.  (note: my load was a PWM controller driving a power seat-adjuster DC motor, repurposed to turn a wheel. The full speed setting drew 3.67 A at 15V so I used a friction brake against the wheel to increase the load to get up to 6 A).

So I am a little bit surprised but I have now confirmed the label claim was true: this cheap supply really can deliver 15 V at 6 A.  I don't know for how long, but for a few seconds at least, and nothing smoked.  Looks like peak efficiency is reached around 3.5 to 4 A load.