Solar Charge Controllers

In this page, we will have a look at:
ST-G1220 20A PWM controller
ST-W1240 40A PWM controller

       Here is a newly purchased 20A one locally. It was cheap enough that I tore it apart to see what is inside.

       I have already installed a couple in the province and this one is also destined to be put in a remote location for lighting and it seems to work fine.
Manufacturer: Suoer from china, model: ST-G1220

Charge controller exterior.

Sticker on the back.

Front membrane detail.

Front membrane peeled off.
Why? because I can.

Removed four screws from the back aluminum cover/heatsink.

Power board screwed to back plate.

TM1628 LED driver chip

A dual comparator and voltage regulator.

A dual op amp and MCU.

RU40190R output switch mosfet.
40V/190A,RDS(ON)=2.5mΩ(Typ.)@VGS=10V

HY3506 PV input switch mosfets.
60V/190A RDS(ON)=mΩ(typ.)@VGS=10V

Back to top of page

       08 Nov 2014:

       Here is another Suoer branded charge controller model ST-W1240. It is rated at 40A, PWM mode controller.

This is how it arrived.

The controller comes wih 12pcs of uninsulated ring crimp terminals, tox and self tapping screws along with the usual warranty card and manual.

Controller exterior.

Access to the terminals is through a cover.
The cover is screwed in with self tapping screws.
There is a risk of stripping the threads when you take the cover off a lot of times.

It uses PCB mounted screw terminals.

I actually prefer these than the usual terminal blocks for very high current but are a pain to use with bare wires. It would not be a problem with the included crimp terminals though.

Removing four screws behind the heatsink, we take the top cover off.
We can see that it uses a lot of solid brass jumpers to be able to handle the current.

These are the output current shunts.
Input current is not measured.

Input protection.
A single MOV in parallel with the PV.

Calibration pots for voltage and current displays.

The LCD is not soldered flat to the board!

RU4099R PV input switch.
40V/200A RDS(ON)=2.8mΩ(Typ.)@ VGS=10V

RU40190R output switch mosfet.
40V/190A,RDS(ON)=2.5mΩ(Typ.)@VGS=10V

The buttons are surface mount but are manually soldered in place with blobs of solder.

Solder side is not perfect but I've seen worse.
The MOSFETs are manually soldered and there are a few solder blobs here and there.

I desoldered the LCD to see what is underneath.

SH79F166AF Microcontroller,
LM2931 5V regulator,
unknown SOT89 regulator (marked 7550-1)
and a bunch of discrete devices to drive the mosfets.

XL4001 DC-DC converter for USB5V output,
LM358 dual op amp for shunt voltage amplification.

The sides have a clear access to the inside.

Take this into consideration in places where bugs or lizards could crawl in and make this controller their home.

With a 1A load, the current reading registers 0.0A.

The minimum value displayed is 2.0A (load current at 2.4A).
And the current does not count up in regular increments. 2.0 → 2.2 → 2.3 → 2.5 and so on...

With a 5A load.

With a 10A load.

(My electronic load goes up to 10A only)

       I was not able to test the charge algorithm but the load overdischarge disconnect cuts off at 10.4V and reconnects at 12.4V. The PV input voltage also shows a residual voltage without anything connected on the PV input. Even connecting a resistor in parallel still shows a small voltage (~2V) on the display. When I get this installed in another setup, we shall see how it performs.

       I was also able to take a peek on the 60A model (sorry no pics) and there seems to be no difference in the power devices or PCB layout. Only thing I can think of is that the overcurrent limit for the load is different via software. Still is unfair for the higher amperage model to cost more though but that is something the marketing department does.


       23 Dec 2014:

       An update on the charge controller model ST-G1220 (white one). It has been installed for a few months already and working fine, then following several days of cloudy days causing the battery to be undercharged. The controller would turn the light on at night but turn it off exactly 1 hour after it has turned on. This would happen even if it was set to a time delay of 12Hours. Following tests, it was confirmed that it wasn't the battery discharging sooner but it was the controller just shutting it off. Setting it to mode 17 (debug mode without time delay) and it stayed lit up to dawn. I had this confirmed with another unit that also had complaints of it shutting off earlier than what the delay should be. It seems to be software related but being both units are installed in different places in the province, I cannot do extensive tests to replicate the cause but it seems only the time delay mode is affected. Mode 00 (output permanently off) Mode 16 (manual control) and mode 17 (debug mode, light sensing, no delay) are working fine.

Page created and copyright R.Quan © 09 Aug 2014.