When I experimented by connecting a large load to the DPS605U, there was quite a bit of ripple noise on it. Please see here for the review and load experiment.
I modified this Chinese stabilized power supply DPS605U and almost got rid of the ripple, so I would like to show you how to do it.
Ripple noise of DPS605U
In the previous experiment, the ripple noise when the incandescent lamp was used as the load was as follows.
- At 1A load: 20mV
- At 1.5A load: 100m
- At 2.0A load: 100mV
- At 2.5A load: 200mV
- At 2.8A (maximum) load: 3.6V
When running more than 2.5A, the ripple increased considerably, reaching a maximum of 3.6V. The ripple noise I observed is at a frequency of 100Hz, so it is the same as the waveform of a full-wave rectified 50Hz commercial power supply.
As was the case in the past with the flickering problem of LED fluorescent lamps from Daiso (which has been fixed in the current product), there is a possibility that the input capacitor is insufficient in capacity.
So, I decided to try the Chinese stabilized power supply, DPS605U, if the capacity of the capacitor of the input is increased, whether the ripple will be reduced or not.
Modification of the power supply
Disassembly
Let's open the case. The two capacitors that can be seen in the lower left corner are the capacitors of the input stage that smoothes the 100V full-wave rectified electricity.
It is written 200V 680uF. I'm going to increase the capacitance by connecting another capacitor in parallel with this one.
The back of the board looks like this. In the upper left corner, where the capacitor is soldered, a capacitor will be added.
Buy the capacitors
I found the perfect capacitor at a reasonable price. It is an electrolytic capacitor of 200V 680uF.
I soldered the thickest cable possible to this capacitor.
Adding Capacitors
Connect the purchased capacitor to the capacitor pad on the back of the power supply board. Be careful not to mistake the polarity. With this, the capacity has been doubled.
I turned on the power supply. It started normally. Now, how is the effect?
Load experiment
As before, I used an incandescent light bulb as the load and tried to run electricity.
At 1.5A load, the ripple was 44mV.
For a 2A load, the ripple was 28mV.
At 2.5A load, the ripple became 64mV.
At the maximum current of this load, 2.8A, it was 24mV.
To summarize
- At 1.5A load: 100mV → 44mV
- At 2.0A load: 100mV → 28mV
- At 2.5A load: 200mV → 64mV
- 2.8A (maximum) load: 3.6V → 24mV
The ripple noise at 100Hz is only a few 10mV. The ripple noise at 100Hz is only a few tens of mV. There is no change in the high frequency noise, but the noise from the commercial power supply has become very small. Now we can use it without worrying about ripple noise even under high loads.
This power supply itself is very inexpensive for a power supply, so an inexpensive power supply turned out to be an inexpensive power supply with good performance. A very good purchase.
