Sunway "MPPT" Solar Charge Controllers
A friend sent this to me to look into as there are uncertainties of whether this charge controller is a true MPPT or simply marketed as one but really is only PWM mode.
Here are the published specs:
Model SSCP-48-60A Battery Voltage 48V Rated charging current 60A Rated load current 60A Over load protection 75A Short-circuit Protection ≥100A No-load current ≤30mA Charge circuit voltage drop ≤0.4V Discharge circuit voltage drop ≤0.4V Over-voltage protection 66V(Adjustable) Over-voltage recovery voltage 60V(Adjustable) Over-discharge protection 42V(Adjustable) Over-discharge recovery voltage 50.4V/(Adjustable) Balance charge protection 57.6V(Hold time: 2H)(Adjustable) Floating charge protection 55.2V(Adjustable) Temp. compensation -5mA/°c/2V Operating temp. Industrial Grade: -25°c ~ 55°c Charging method PWM Operating altitude ≤4500m Dimension 175.5*155.5*48.2mm Net Weight 1.2KG
Upon initial inspection, it lacks the large magnetics needed in true MPPT charge controllers so it is not looking good. Let's look at the internals and see what it really is.
The charge controller.
Popping the hood.
There is an absolute lack of high power magnetics so this is indeed a PWM controller.
The front panel LEDs and buttons use the membrane type switches.
I'm not a big fan of these as they are not reliable in the long term. And if they do fail, you cannot just replace them with tact switches.
Three current shunts. Two for the PV input and one for the load output.
60A charge and 60A discharge? Not looking good.
PCB is marked as SW-SSP-1224-V02
Looks like it is originally designed for 12 and 24V use?
Overview of the control / logic section.
Left area has two op amps for current sense amplification.
Middle part holds the uC, LCD and support circuitry.
Right area has the 5V buck supply and 15V boost supply.
Main uC is an Atmega128.
TC4427 dual gate driver IC and optocoupler.
two dual op amp ICs.
Main buck regulator IC (LM5010) high voltage buck regulator IC to derive the 5V from the main battery voltage.
It appears to have been reworked.
The input protection is done by a UF5408 (overkill but OK), a large self resetting fuse (OK), a cap and 150V MOV (not OK - the buck IC is rated 75V MAX so it should be lower than that), bulk filtering (the electrolytic cap) is 250V 1uF - wayyy too small. It should be 100's of uF to be useful to prevent damage when the battery connection is intermittent or during accidental battery disconnection.
If the connected PV array has an open circuit voltage higher than 75V you risk blowing the 5V buck supply if the battery is accidentally disconnected with the PV still connected.
Closer look at the IC.
The corner pads of the IC footprint are peeled off from the sloppy rework.
Bottom side of PCB. Not much to see here.
The four in the corners are mosfets and the one in the middle is a dual diode wired in parallel.
The mosfet part numbers are erased.
Yes, all four of them.
The dual diode is an STPS80150 150V 40A schottky diode.
Interestingly, the mosfets are labeled which ones are used for charge and discharge. These two are for the PV switching.
The other two, one is also for the PV and the other for the load switching.
All are low side switched so the positive is commoned except for the PV+ which is isolated by the blocking diode above.
Clearly there has been a misunderstanding between the designers and marketing department. Three parallel mosfets for PV switching and only one for the load and they claim 60A charge and 60A discharge rating?!?
More like 60A charge and 20A discharge.
The heatsink is cleanly machined but they attempted to use threadlocking compound on the standoffs which clearly did not dry up as the material is incompatible with the metal.
The problem in the unit I got was that the 5V buck IC failed.
The buck IC is supposed to be heatsinked to the PCB through the thermal pad under the IC package but the PCB is clearly not laid out for that.
From what I can gather, the failure started when the load output is fully loaded to the advertised 60A causing the single switching mosfet to fail short. The gate is then shorted to source causing the gate driver to draw excessive current from the internal supply blowing the poorly cooled buck IC.
It was an easily preventable disaster if only the software was programmed correctly to disconnect the load on an overcurrent condition without exceeding the mosfet rating but the marketing department decided to set it to the (wish)full 60A spec.
Supplying 5V to the uC allowed it to power up and initialize.
But the voltage sense does not detect a battery so I cannot test other functionalities.
I could rig up an external buck converter but it may be too much work. Let's see if I can find one I can scavenge from to repair this.
30 Aug 2015:
Update. I have received another unit but has been rebranded. This unit is supposed to be a 12/24/48V 50A version.
I'm curious to see if there are any differences and what was the failure mode.
Here are the two units side by side.
With the cover off.
Same PCB part number as the above unit.
Side by side, I can't see any difference.
The LM5010 chip has also been reworked but this one is cleaner.
Same ATmega128 chip
The op amps have a touch of rework done to them too.
The heatsink anodize is a little different.
The anodize is thinner as the black color is lighter but same machining quality and there is no threadlocking fluid mess.
Back of the PCB.
Same STPS80150 diode.
Mosfet nmbers are also erased.
Yes, all four of them also.
I tried powering it up but the circuit draws a lot of current.
Used the FLIR and found the TC4427 mosfet driver IC heating up so I desoldered it.
The circuit powered up afterwards. The 5V buck IC is also blown.
Also has a teeny input bulk capacitor. This one used 400V instead of 250V.
I tried using a 7805 in place for the buck IC and it seems to work.
I might be able to scavenge parts from one unit to fix the other. We shall see when I get some free time.
Here is the sticker with the extremely minimalistic "specs". Atleast this one had some basic info.
The model is: HTW-122448-50A
The only sticker I found on the other unit. Here is a comparison on both of them.
The previous unit is model: SSCW-2448-60A-TH
01 Sept 2015:
Rigging up a 7805 regulator to operate the logic circuitry worked.
Here is the jury rigged 7805 regulator. With this configuration, I can only run this controller on a 12V system.
Seems to have worked fine. But the buttons still fail to respond.
This will do for all my intents and purposes.
Hooking up the PV array from my grid tie inverter (2x 160W panels in series, 18Vmp each)
This is the battery bank voltage and battery current.
And here is the PV input voltage and PV current. It clearly shows that the PV input voltage is about 1 diode drop plus the battery voltage. The difference in currents can be ignored as it is very small.
This PV array operates at 30V when connected to my grid tie inverter which has MPPT.
Here is the PV open circuit voltage at the time.
And this is when the PV array is connected to the controller.
The slightly higher voltage shown than before is because the probes are attached to the alligator clips which is a couple meters away from the charge controller.
I plugged the 2x 160W PV array into my main Epsolar Tracer MPPT controller and this is what you are supposed to see with a real MPPT controller.
Vmp seems a little low due to clouds rolling in but the point is, true MPPT controllers trade voltage for current, and this Sunway controller does not.
I can tell 100% from the start that it is not an MPPT controller but there are some that wanted more proof. This is for them.
Page created and copyright R.Quan ©28 Aug 2015.