Fangpusun Charge Controllers
I discovered Fangpusun and a look at their website and found that they sell copies of well known brands from Outback, Xantrex, Steca et.al.
In this page, we will have a look at:
Fangpusun Flexmax FM60 (A ripoff of the Outback FM series MPPT controllers)
modification - LCD contrast mod of the above.
PR245 - PWM controller (ripoff of the Steca PR series).
Fangpusun Flexmax FM80 Flexmax 80A version clone.
Blue 100/50 ripoff of the Victron BlueSolar MPPT 100/50
Outback Mate3 - Tested the FM60 clone on the Outback Mate3.
At first glance, It looks and feels like the real thing. Even the web advert from their site and online aution sites show that the description and specs are basically plagiarized copies from the Outback Flexmax series with the word "Outback" replaced with "Fangpusun"! Check out the pictures of one I took one apart.
Sticker on the box outside.
I could not remember if they are the same with the original I worked with a few months back.
It is well packaged with foam ends and wrapped in plastic.
Looks very much like the original with a slight difference.
The cover plate is now branded Fangpusun.
Side sticker is just like the original. The earlier version is a silver sticker with black text.
Cover piece. Die casting is not as nice as the real thing but is ok.
Yeah, looks similar to the original.
The large bulk capacitor is a no-name UUcap brand.
Earlier versions I have seen used an Elna LP5 capacitor, as did the genuine Flexmax controllers.
The wire loops on the right of the big cap are the output current shunts.
No, it wasn't me. Came like that out of the box.
Holes do not line up.
Must be why the screw head was stripped.
Fan is a Delta brand.
Decent enough and found on countless other products.
Main chassis out of the back case.
Back side of the PCB.
Back has some flux residues from hand soldered components but otherwise clean.
Power devices are clamped with a plastic bar. I believe the original used a metal bar.
I'm not sure how well it will hold up when it gets hot.
There are some damage on the PCB edges.
Some copper got exposed but there is still continuity.
Under the terminal blocks. The white part is a cardboard insulator that extends between the terminal blocks.
Main board separated from heatsink/chassis.
It is pretty well packed with parts.
Back of the heatsink area.
They have a glue gun and are not afraid to use it!!!
All the heavy parts on the main board are gunked all over and stuck down as well. Does not look pretty but does the job.
They used a lot of silicone grease on the back of the silpads which is not necessary and could increase thermal resistance. But lots of manufacturers do it!
Control board overview
Back is insulated with... Masking tape! I think they could have done better with, say, kapton tape or fishpaper.
Cap is a Rubycon. Not sure how legit it is though.
The white box behind the electrolytic cap is a 110degC thermal switch if I was not mistaken reading the faint numbers. But I think that may be incorrect as it is too high. I was not able to see other thermistors to sense internal temp.
Main uC section.
Main uC is an Atmega644P
Interface? section of the control board.
There are connections for ICSP or Jtag under the sticker.
SOIC8 chip is an LM393.
The 14pin SOICs are 74 series logic.
And now, we move on to the main power board.
Main power devices are Fairchild FDP2532 150V 79A mosfets.
There are 13 in all.
six are used per inductor in sychronous rectified buck topology.
The last one is used for the auxiliary supply.
Electrolytics for the auxiliary supply.
Two are Samwha brand, the other two are Rubycon.
Two 30A relays are used.
These are used to disconnect the output of the buck converter from the battery when the controller is in sleep mode.
More of the large UUcap branded bulk capacitors.
There are four in all in this controller. All are 2200uF 200V rated.
Auxiliary power supply sub board.
It is based on a UC3843 driver chip.
Power stage control circuitry are mostly through hole parts.
Nice 1% metal film resistors.
DIP8 ICs are HCPL3120 2.5A IGBT gate drive optocouplers.
AA battery for scale. Those inductors are huge!
I'm not sure if they run parallel or interleaved phase. But they certainly should share the battery charge current at 30A each without saturating.
PV and battery terminals are beefy. These could take up to 4ga wire. No flimsy plastic parts here.
Connectors for data, temp sensor and auxiliary output.
I was surprised to see a mil spec Dale RN50 resistor in this.
One is for the battery voltage sense.
Rough adjustement with the pots and fine tuning on software.
The other two are labeled "OS" and "G" but I'm not sure of those.
The auxiliary supply block.
Another view of the power stage.
They use buss bars internally and not rely on the thin PCB copper which is good.
It has been installed in my home RE system and the UI appears to be exactly the same as the original Outback FM series. We shall see its performance soon.Back to top of page
05 Sept 2015:
Here is another charge controller from Fangpusun.
Here is the box as recieved.
This one is model PR245 rated at 12/24V 45A
This looks surprisingly similar to Steca PR series but those only go up to 30A.
And, not surprisingly, Fangpusun also has the lower rated models with the same model numbers!
The front panel silk screening has a smudged part in this area.
The heatsink is nicely machined. Anodizing is smooth and glossy.
External connections use the plastic terminal blocks. The one used here appear to be decent quality.
Plastic case is decent. Moulding is not like the expensive stuff but doesn't feel cheap either.
The main board is smaller than expected but is packed with SMD parts.
There are busbars and connectors that slip into the terminal block. The busbars are nickel plated copper as magnets do not stick and scraping a part reveals copper. That's nice.
These appear to be test points for ICSP and batch testing.
This side has a LM324 quad op amp and a TL431 voltage reference IC.
The pads seen nearby aren't used but appear to be for a connector for external temp sensor, perhaps?
Nice LCD spacer. The LCD is the opaque type so, no easy way of hacking a backlight into the display.
The uC and support components are underneath.
The power devices are attached to the heatsink via clips.
Heatsink is well finished. I don't have complaints on that.
Back of the PCB. It is very clean and little uncleaned flux residue.
There are lots of test points probably for bed of nails testing during production.
All power devices are Fairchild FDP027N08B mosfets.
These are rated at 80V 223A. Very hefty devices.
There is a 47V MOV for input protection.
The uC is an ST micro brand.
I could not find the datasheet on a quick search but it is marked MLT22717.
With the device powered up. We get the main screen.
The LCD contrast is rather poor. It is not readable in certain angles but is fine directly in front.
It can be toggled to display battery voltage
I'm not really sure how it measures amps as there are no current sensors or shunts that can be seen.
Some traces look like they use the PCB copper as a crude shunt. Not very accurate but that is the only way I could see this doing it.
Battery charge current
Load discharge current
Accumulated charge amphours
And accumulated discharge amp hours.
Scrolling through the manual, it appears to be quite versatile with lots of features.Back to top of page
16 Apr 2016:
I have been using the Fangpusun FM60 in the home system for a while now. I don't have much complaints in terms of performance but a couple of gripes. The fan is a little noisy, something I could live with but the LCD contrast is bad. Researching online, there are no contrast adjustment option even on the original Outback controllers.
This is what the original LCD looks like.
The black squares are showing pretty bad.
I traced pin3 of the LCD module which controls the contrast and it goes to the main uC.
Contacting both Outback and Fangpusun was a dead end. They said there are no software options to control contrast even if it was connected to the uC.
I was able to purchase this standard 20x4 LCD module and it had a space for an onboard PCB and a jumper to disconnect pin3 on the external connection.
I added an SMD trimmer for on-board contrast adjustment.
I did not have 10K available but used 20K instead which should work just fine.
This solder jumper connects the contrast adjustment line to the pin3 connection.
Removing it makes the pin3 not connected and leaves the onboard pot for the contrast adjustment.
With the FM60 disassembled, the LCD modules are checked and are perfectly compatible. Exactly the same size, mounting and connection locations.
Removed lots of hot-snot and the boards came out.
Front panel PCB has some circuitry on it.
I cleaned it with IPA as there was some flux residue on it.
A perfect reason to test drive my new toy.
Hakko FR-300 desoldering tool.
Makes removing through hole parts a joy without destroying PCBs!
The old LCD removed.
The keypad PCB is reinstalled without the LCD module to fix it in the right location.
Then the new LCD module is soldered and glued in place.
Testing it while disassembled to adjust the LCD contrast before reassembly.
While I was at it, I also replaced the masking tape insulation on the busbars and replaced with Kapton tape.
The finished product, a much cleaner LCD display!Back to top of page
I'm quite happy on how it turned out.
20 Apr 2016:
Now here's the 80A version. Another clone of the Outback FM80.
The unit as I got it out of the box.
Sticker came off and got stuck in the plastic bag.
In a quick glance, it looks the same.
Now they're using Teapo brand caps.
And these are now only 160V rated, previous revisions used 200V caps.
Wait, where is the screw?
There it is!
The thread on the case was stripped so any movement of the PCB and the screw jumps off.
External fan assembly on top. Same as the original FM80.
No name branded fan.
I'm not sure how long this one would last.
There is a bodge.
Tracing the circuitry, this appears to have something to do with the high side gate drive of the first mosfet but could not really figure out.
It was recommended by Fangpusun so I also applied it to my FM60 unit.
Same mosfets as the 60A version.
The fan port of the 60A version is blocked and is now using a bigger fan at the top externally mounted.
Main board looks the same as far as I can tell besides the brands of the electrolytic caps.
More Teapo caps!
Copper windings of the large inductors are the same in the 60A and 80A version. Both use 15 strands of 1mm wire.
CPU board also appears to be the same except for the sticker saying it is 80A.
Looks like a legit Elna cap.
Auxiliary powersupply now used Samwha and Suncon caps.
LCD contrast appears to be better than my 60A unit.Back to top of page
23 Apr 2016:
Fangpusun sent a blank PCB and it is decent quality.
4 layers, 2oz copper and 2mm thick. dimensions are 135mm x 281mm
Back to top of page
20 Jul 2016:
Here it is. This is a clone of the Victron BlueSolar MPPT 100/50, 50A 100V input MPPT controller.
The unit has already been opened since the supplier was impatient and broke it apart to see what is inside. The top cover was broken apart since the inductors were potted in it for mechanical support.
Right from the top we can see nice heavy busbars to carry the high currents.
Tops of the inductors partially potted in the top cover for mechanical support.
A large HRC fuse for the battery +. This is nice and definitely overkill.
Main power devices are IRFP4110 by IR.
These are beefy devices rated at 100V 180A.
The topology is a pair of synchronous buck converters, possibly interleaved phase.
Heatsink is nicely machined but they used silicone goop on silpads.
You use silpads so you don't have to use goop. But they still always use goop!
Back of the PCB.
Parts are well laid out, maybe because they copied it from Victron?
PCB is also conformal coated
Some shavings from the plastic case.
Main input caps are 3x 160V 330uF Teapo brand.
Main uC is a Freescale MC9S08AC60.
Output filter caps are 2x Samwha 35V 1000uF caps.
The orange wires go to the internal temperature sensor....
....Which is siliconed near the main switching devices.
This MOSFET connects the PV+ to the inputs of the DC-DC converters.
It is an IRFP3306 which is rated at 60V 160A
This MOSFET shorts the PV+ to ground and is the same device as the main switchers (IRFP4110).
There is a drop of solder possibly from hand soldering the black jumper wire.
There are two of these blocks consisting of a dual op amp and shunts which measure current after the large inductors.
This section is the auxiliary supply for the plug in remote controller.
The small drum core inductor is for the auxiliary supply.
The heatsink is grounded via a 1nF cap to negative through this standoff.
Terminal blocks are nice rising clamp types.Back to top of page
The topology is simple and can be quite efficient.
From the parts, I think this can do its nameplate rating but not sure about longevity due to the main supply caps are of not the best available. It is a good value for its price though, just change the caps out for better ones before they blow up if this will be used near or at its full rating. But wait, the unit is potted so no way to do that.
30 Aug 2016:
So an install in the province uses two FM80 controllers, one a genuine Outback FM80 and the other, a Fangpusun FM80. The owner wants data logging, graphs and online viewing of energy data so I recommended the Outback Mate3. We got one and was kind of hesitant if it would work with the clones.
I connected it to my FM60 in the home system to see what would happen. Fingers crossed!
And it actually worked! We also got a HUB4 so that we could link both controllers to the Mate3.
It also included an RTS (remote temperature sensor) for reading the battery temperature and I wanted to add one to mine. Probed the phone plug type connector and it is just a 10K thermistor. I made my DIY version so now I have one too.
And here it is half assed with what I had on hand but it works.
Page created and copyright R.Quan ©03 Sept 2015.