Cheap Chinese DC-DC CC CV Buck Power Supply

[byteseven]

Hi all,

I purchased some cheap Chinese buck converters with what I believe is CV and CC control. The idea was to use it to supply 12 V @ 500 mA, stepped down from an old 19 V 3A laptop PSU. This is all for an electrolysis rig (rust and paint removal from a bicycle frame). Please see the link below for an example of the board I received.

https://www.aliexpress.com/item/Free-Shipping-Non-Isolated-Constant-Current-And-Voltage-Lithium-Charger-Power-Supply-Module-5A-LED-driver/1084552308.html?spm=2114.13010608.0.0.tCMLv1

I wanted to replace each of the 10K pots with "proper" ones I could mount to a project box. In the process of desoldering them I pulled the PCB pads off making them somewhat useless.

I decided to reverse engineer them to see how they worked. I used the X4015E1 datasheet to get an idea of how that IC is configured, broke out the multimeter and started buzzing out connections. I had to remove the 0.05 Ohm current sense resistor and the inductor to make my life easier. The resulting schematic is attached to this post. Note, the values are mostly the default from EasyEDA, so please ignore any specific values.

Now, assuming I have the schematic correct, I cannot see for the life of me how it regulates current. I can see the constant voltage section of the circuit, although it does seem odd that the pot is shorted to itself. The current potentiometer next to the op-amp seems to just set an whether the LEDs off U2.2 light up or not. There is no switching circuitry to regulate the current, besides the op-amp itself.

So, are there any clangers? Things I may have missed? I intend to buy another couple of boards because they are cheap and it annoys me that I can't see how it works.

[DBecker]

You probably have a few of the resistor values wrong.

When in CV mode, with current under the limit, the LM358 output is low.  The voltage regulation works as normal, with the pot and resistor+LED forming a voltage divider to the 0V LM358 output.

When the output current rises in CC mode, the LM358 output rises above 0V, resulting in the output voltage dropping to keep the FB pin at the reference level. 

[jeroen79]

That schematic looks odd to me.
The XL4015E1 would try to keep the FB pin 1.25V above it's GND pin.
But wouldn't the blue LED drop more than that?

[byteseven]

You probably have a few of the resistor values wrong.

Yup, didn't bother taking note of them. I've added them to my schematic now, if it helps you to cure my ignorance.

When in CV mode, with current under the limit, the LM358 output is low.  The voltage regulation works as normal, with the pot and resistor+LED forming a voltage divider to the 0V LM358 output.

When the output current rises in CC mode, the LM358 output rises above 0V, resulting in the output voltage dropping to keep the FB pin at the reference level.

Thank you for the explanation, I have a better grasp of what the function of the board is now. I had assumed a different mode of operation, ie when the current limit it reached that the voltage remains constant, as well as the current. After reading around, it seems that this is contrary to Ohm's Law.

I'm still puzzled by how and when the red/blue LED pair light up. The easiest thing would be to get a working board and try some different settings and take some measurements.

That schematic looks odd to me.
The XL4015E1 would try to keep the FB pin 1.25V above it's GND pin.
But wouldn't the blue LED drop more than that?

You may have a point there. After watching some Youtube clips of Russian's playing around with similar boards it seems the LED is red. I may have mis-remembered. I've changed the label on the schematic. I'll get some more and double check.

Since starting this thread I have realised what I really want is a buck converter than can step down 19 V to 12 V and be capable of handling the load I wish to apply. In this case ~ 8 Ohms. Rough calculations tells me at 12 V I'll need a buck converter that can handle 24 W.

Thanks again for your help.

[Kleinstein]

The LED towards the FB Pin (blue in first version, now red) should be the other way around: if the current is to high, the OPs output goes high and thus turns the LED on and lifts the FB pin higher than the critical 1.2x V.

It is interesting to see than the LEDs allready indicate current limiting of the current is only close to the limit. It looks like the divider for the level where indication starts should be more like the other way around - indication at 10% makes no sense.

[byteseven]

The LED towards the FB Pin (blue in first version, now red) should be the other way around: if the current is to high, the OPs output goes high and thus turns the LED on and lifts the FB pin higher than the critical 1.2x V.

Indeed, that makes more sense. I have updated my schematic (attached).

It is interesting to see than the LEDs allready indicate current limiting of the current is only close to the limit. It looks like the divider for the level where indication starts should be more like the other way around - indication at 10% makes no sense.

I think I see your point. I'll double check that part of the circuit.

[byteseven]

It is interesting to see than the LEDs allready indicate current limiting of the current is only close to the limit. It looks like the divider for the level where indication starts should be more like the other way around - indication at 10% makes no sense.

I double checked the resistor values and they are correct. Reading the description from the seller it makes sense that they indicate when the current drawn drops to 10%. They sell these boards as CV/CC battery charges and the battery is deemed "charged" or close to it when the current draw drops to 10% of the set point.