diy solar

diy solar

How about a solar generator repair?

That would take all the fun out of it. Just please be careful ;-)
Who knows, if you can get them cheap enough on eBay perhaps you'll soon have a nice sideline of refurbs?


You can probably recognize the BMS portion of the PCB by the cell wires and the thick(er) wires for the battery +/- to the BMS (or possibly thicker traces on the PCB).


Sweet!
No chance of a sideline... this is the only one I have found.
 
The BMS is on the vertical card that is labeled "AGND, B1, B2, B3". I think the voltage sense leads come in through J1 and you can see the traces in photo *441.
 
I was figuring it would be the BMS, but I didn’t see any mosfets or high amperage leads feeding it...
 
I didn’t see any mosfets or high amperage leads feeding it...
I think it doesn't need to pass/block the battery current in this design since the main board already has huge mosfets which can be controlled to do that (the ones mounted to the heatsinks). So in this case it's more of a monitor + control board and doesn't deal with high currents.
 
I haven’t opened the heatshrink on the battery pack, but I would guess it is about 6P3S
 
We'd need to see what all is on that board. It has a ton of header pins and we'd need to figure out where they all go.
They are all labeled. I think I posted a picture. I found the 12.6V pin has voltage on it, and the b1,b2,b3 pins have corresponding voltages.
 
Does the on/off switch disconnect the battery? If not, where do the wires connect to the circuit board? It'd be funny if it was something simple like the switch.

As those fets attached to the aluminum fins in the upper left of photo 728 are close to the big red wire (assuming battery +) they're probably the power out of the battery controlled by the BMS, so I'd check those fets first. If you have the battery hooked up, rather than do as the video recommends, I'd test the voltage at the gate to negative and the drain to negative.

If the gate doesn't have voltage, it'll most likely be something in the on/off switch or a BMS sensor that failed (e.g., low voltage cutoff stuck in the off position). If the source and gate both have voltage and the drain doesn't, then bad FETs.

You can probably "jumper" past to test the rest of the system by shorting the source to the drain (just be gentle and don't pull a lot of amps ... essentially you want to see if any of the display lights up).

Update: Before inserting probes into a live device you should probably yell "SMOKE TEST" or "Here, hold my beer" (it's tradition after all).
 
Last edited:
probably the power out of the battery controlled by the BMS, so I'd check those fets first.
They would have their drain connected to one side of the red fuse, I believe, if they are N-channel FETs, or their source if P-channel. However be careful as there may also be some switching FETs connected to that same net and you wouldn’t want to tie those ones always on! I think this design might have an H-bridge driving the transformer primary so if you short too many transistors on, you’ll get “crowbar” current through the transformer or through the H-bridge (or both) and that would blow something up, maybe you! You should at least test with a current limited power supply imitating the battery and not the real battery connected! This would allow you to limit the current in the event of some disaster, and not light your batteries on fire.

An H-bridge would be an identical set of 4 transistors (probably; it could also be two P and two N channel). Source and drain of two connect to one side of the transformer primary, and source and drain of another two connect to the other side of the transformer primary. However since transformer is just a wire your multimeter will show one side of all 4 are shorter together in DC. I believe there is also a different set of 2 on the primary side which is what svetz is talking about, probably.

If their drain is connected to the battery, you can also just “diode connect” them by connecting the gate to the drain. This would still conduct the current through the drain/source of the transistor and not through your shunt wire.
 
Last edited by a moderator:
Does the on/off switch disconnect the battery? If not, where do the wires connect to the circuit board? It'd be funny if it was something simple like the switch.

As those fets attached to the aluminum fins in the upper left of photo 728 are close to the big red wire (assuming battery +) they're probably the power out of the battery controlled by the BMS, so I'd check those fets first. If you have the battery hooked up, rather than do as the video recommends, I'd test the voltage at the gate to negative and the drain to negative.

If the gate doesn't have voltage, it'll most likely be something in the on/off switch or a BMS sensor that failed (e.g., low voltage cutoff stuck in the off position). If the source and gate both have voltage and the drain doesn't, then bad FETs.

You can probably "jumper" past to test the rest of the system by shorting the source to the drain (just be gentle and don't pull a lot of amps ... essentially you want to see if any of the display lights up).

Update: Before inserting probes into a live device you should probably yell "SMOKE TEST" or "Here, hold my beer" (it's tradition after all).
The switch is good, it connects to the main board... likely controls the fets to the inverter circuit.
But since he battery isn’t going into the board from the BMS, there is no charging or inverter output.
Following all the rest of the instructions you guys are posting is out of my league at the moment.
 
Ok, I picked up a dead rockpals 250W SG...
LOOKS COMPLETELY NEW, ZERO WEAR ON ANY SURFACE...
I mean no wear on anything.
I open it up, and the battery is sitting at 10.6 so, shipping voltage.
I dig into the bms, and it is an all in one board. Bms, usb, and inverter all on one board.
Nothing burned or bulged...
I start examining everything closely and find this...
It looks like poor soldering on the little processors.
Anybody know if that is normal?
I figure I will soak some Flux in there, suck out the blobs and see if it works...
I'm tempted to buy another one just to open it up and see if they are supposed to be blobs there.
 
Back
Top