• Resolved

DCP020515D: Incorrect output voltages

Part Number: DCP020515D

Hi, I'm using DCP020515D power module to get +-15V from USB +5V.

My schematics consists of only 4 elements and as I understand from the datasheet it should be enough.

On the input I use 2.2µF ceramic X8L capacitor between +Vs (USB 5V) and -Vs (USB GND)

On the output I use 1µF ceramic X7R capacitor between COM and +Vout and also another 1µF X7R cap between COM and -Vout

Measuring voltages with multimeter shows 28.40V between COM and +Vout and shows -25.7V between COM and -Vout

These voltages seems wrong because DCP020515D datasheet tells that max should be ±15.75V. So am I connecting something wrong or the measurement itself is not correct? Could you help me on this?

  • Antanas:

    DCP020515D is a dual isolated power converter. The output voltage is dependent upon output loads . Each specified output voltage  load range at 75% load / 50mA load  ±14.25 ±15 ±15.75 with 5Vin at +/-1%.

    Unload +/-Vout DCP020515 output voltage with no output loads can be as high as  25-30V on each polarity.

    Attached is one method to limit +/- Vout to about 16 V with a 16V zener  clamp and series resistor. Tom

    DCPfilteroutput clamp16Vdual 2015.pdf

  • In reply to Tom Guerin:

    Hi, thanks for the response! What about if instead of zener diodes I used LDOs (TPS7A47 for 15V and TPS7A33 for -15V)? What would be cons and pros for this approach comparing to zeners?

  • In reply to Tom Guerin:

    I'm having a problem with output voltages as well, but on the low end. I'm using an LM1117 LDO to provide the input 5V, and have wired the DCP020515D as prescribed in the datasheet, but the output voltages are just a few volts and unbalanced. I have not tried the filtered clamp, partially because I am not even approaching the Zener voltage yet. At first the outputs were unloaded, but I tried both 10k and 350-Ohm loads to see if there was a difference.

    With the 10k loads, I tried three different DCP020515Ds, and got the following output voltages:
    +Vout, -Vout
    +6.2, -4.2
    +3.7, -2.9
    +3.9, -3.3

    With the 330-Ohm loads (just under a 50-mA target at 15V), I got the following voltages:
    +3.8, -3.4
    +1.86, -1.54
    +4.3, -3.3
    In this test, the input voltage drooped to 4.86V and both the LM1117 and DCP020515 were running warm (but not too hot to touch).

    My initial thought from the first chip was that I damaged the device soldering the SOIC chip to a breakout board, which is why I tried the others as well. My soldering skills are not pro, but I do rank myself as generally competent with these bigger SMDs, so I am inclined to rule that out since three separate chips are behaving badly.

    Any other thoughts on what may be causing the low, unbalanced output voltages?

    Thanks in advance...

  • In reply to Michael Prairie:

    Hi Mike,

    Is there anything else on your board that can be loading these outputs?

    Can you send me the schematic? You can send it directly to me at jarrigo@ti.com .

    Thanks, Jason

  • In reply to J Arrigo:


    As we discovered off line, I was using the "Typical Application" schematic in Fig 12 of the datasheet, but I did not catch that it was for the single-output version for which -Vout was also the GND terminal. I fixed it by connecting a capacitor between each of the two outputs of my dual-output device and GND instead of between each other, correctly shown in Fig 10 of the datasheet.

    Using the cap between +Vout and -Vout effectively locked them together and resulted in two 370kHz, ringing square waves with an offset of about 10 or 12V. My original measurements were with a handheld volt meter that displayed the average value of each signal. Connecting two capacitors correctly resulted in good +/-15V dc rails.