LTC3780 ebay module malfunction

[npelov]

Hi,

I bought one of these:
https://www.ebay.com/itm/295076976184

because I needed a 12V battery to stable 12V output. SEPIC is terribly noisy and inefficient. I don't like it. I thought this would be nice. But the CC bulshit they tried to implement doesn't work. Rising the current from 0.1A to 2-3A drops the output voltage with about 1 volt, regardless of what position the CC pot is. I tried to power a NAS which has current spikes here and there and it started beeping for undervoltage.
Also the undervoltage pot doesn't cut off the voltage but is doing something strange when near the threashold - the bulb I used dropped in intensity and inductor starts producing noise. I've read some poeople killed it by adjusting this pot. That's not a problem because I have my own undervoltage cut off for the battery.

My question is: can I remove the CC feature? Maybe I can just desolder the opamp and the voltage regulator?

[barshatriplee]

Removing the CC (Constant Current) feature from the voltage regulator module you purchased is possible. But it is not recommended to do this if you do not have a schematic or PCB layout.

[Whales]

> Rising the current from 0.1A to 2-3A drops the output voltage with about 1 volt, regardless of what position the CC pot is

That's not how these should be behaving.  What resistance is the current pot?  See my last note on the list below, perhaps the installed the wrong size pot.  Also check your input voltage when this problem occurs (you might have eg bad wires).

I've used these modules for years as one of my main bench power supplies.  I replace the V and I pots with external ones, then add a voltage readout.

Problems I've had:

 - Each board has behaved differently.  There are some subtle differences between board versions even though they look mostly the same.
 - They're delicate, do something wrong and the LTC3780 blows
 - Behaviour when the CC and CV knobs are turned down.  My last one had a min output of 0.8V.  The current one shuts off to 0V out until you turn the current pot up a bit (to a few hundred mA?).
 - Current pot is nonlinear (yay) except the wrong way around (very non-yay).  It gives fine detail at higher currents but little detail at lower currents.  I bodged a resistor in to try and counter act this a bit, it helped but isn't perfect.
 - Current pot is used as a 2-terminal variable resistor, not a voltage dividing pot.  This means you can't change the pot to a different value unless you also change resistors on the board.  Requires some reverse-engineering.  The current feedback is implemented with an opamp IIRC (the LT3780 only has a pin for voltage feedback).

[npelov]

Undervoltage pot - 500k (504)
Current pot - 200k (204)
Voltage Pot - 500k (504)

[npelov]

- Current pot is nonlinear (yay) except the wrong way around (very non-yay).  It gives fine detail at higher currents but little detail at lower currents.  I bodged a resistor in to try and counter act this a bit, it helped but isn't perfect.

I think you can counter that by rotating to 180 deg. The only problem will be that rotation is reversed - you have to rotate clockwise to lower the value.

[npelov]

Btw, I'm 99% sure I don't have bad wires. I measure voltage at the terminals. Something is wrong with this thing. I should have read/watched the reviews.

Update.
I'm unable to reproduce the huge voltage drop. so maybe there was some wires/connections loose.Now the drop is 0.2V with the same bulb, short thick cables, measured at the terminals.0.2V at 12V output is still a lot, but it's probably not going to be a problem. I'm pretty sure this drop is due to CC circuit. All chinese CC circuits slowly drop voltage when approaching set current. This one starts dropping from a far. I also inspected with a thermal camera - no heating at the terminals.

The problem is however the over current protection while ramping up. It's too fast. If during the power up the current is too high it'll switch off, wait a second, then try to start again. There is no problem if I power the SMPS and then connect the bulb. I measured the cold bulb to be 0.58R which is 12/0.58=20.7W so it' makes sense.

My hole idea is to power a QNAP NAS with it. It has spinning disks. And it doesnt' survive the power up. I think I'll give up now. I'll power lower power devices with this board and I'll use 20V battery + buck to power the nas.

[Whales]

- Current pot is nonlinear (yay) except the wrong way around (very non-yay).  It gives fine detail at higher currents but little detail at lower currents.  I bodged a resistor in to try and counter act this a bit, it helped but isn't perfect.

I think you can counter that by rotating to 180 deg. The only problem will be that rotation is reversed - you have to rotate clockwise to lower the value.

Sorry I miswrote.  The pot itself is linear.  The overall control acts nonlinear.