EEVblog > EEVblog Specific

EEVblog 1620 - Deye Solar Hybrid Inverter EXTREME TEARDOWN

(1/4) > >>

EEVblog:
Extreme teardown of the Deye SUN-5K-SG04LP1 5kW hybrid solar inverter.
Will it contain the quality components claimed?
Also a topology reverse engineering of all the stages.

00:00 - A generous and coincidental donation from a viewer
00:37 - We have the brochure, let's check the claims!
02:44 - Front panel PCB
03:46 - The gigantic heatsink
04:44 - Nippon Chemi-Con Capacitors? REALLY?
05:54 - Heatsunk and potted inductors and transformer teardown
10:19 - Main PCB's and the confusing layout
12:53 - The fused battery input
13:37 - Are the MOSFET's and IGBT's as advertised?
15:10 - SMD high current links for PnP benefits
16:14 - Detailed circuit topology schematic reverse engineering
18:29 - Earth connection testing
19:07 - MOSFET vs IGBT
19:51 - Grid side IGBT switching and tri-level switching
20:46 - Solar Panels are capacitors to earth and that screws the CMRR
21:34 - HERIC switching topology and patent lawsuits!
23:50 - Grid and Load connections
25:07 - The battery bidirectional DC-DC converter
28:14 - H-Bridge and parallelism
30:14 - Switching drive stages
31:25 - Main processor PCB and a TI DSP ripoff?
32:51 - Redundant backup DSP controller?
34:40 - Genrator and Load conection relay switching
35:49 - Panasonic Relays? WHERE?
37:02 - User Connection PCB
38:25 - Conclusion



Datasheets:
https://www.fet.discoveree.io/datasheet.php?view=pdf&file=crmicro/crsq027n10n.pdf
https://www.magnachip.com/wp-content/uploads/2020/10/MBQ40T65QES-1.pdf
https://www.lcsc.com/datasheet/lcsc_datasheet_2209081730_Wuxi-NCE-Power-Semiconductor-NCE40TS120VTP_C5160411.pdf
https://wmsc.lcsc.com/wmsc/upload/file/pdf/v2/lcsc/2211151030_Vishay-Intertech-VS-EPX3007L-N3_C5263741.pdf
ADVchip DSP: https://file.elecfans.com/web2/M00/6F/21/poYBAGNEwS-AKpNeAIfcFmyDrAQ938.pdf
AVP32F335 DSP: http://advancechip.com/mobile/32fd/61.html
Fraunhofer ISE Successful in Patent Infringement Process for HERIC Inverter Technology:
https://www.ise.fraunhofer.de/en/press-media/press-releases/2022/fraunhofer-ise-successful-in-patent-infringement-process-for-inverter-technology.html

Kleinstein:
The supply crisis during Covid may could be a reason for changing the suppliers.  They may have also updated there documentations too.

Phoenix:
Looks good!

DRM = Demand Response Modes. It's an Australian thing that never really took off. It was to allow your smart meter or other utility device to control things like your aircon, hot water, solar, battery etc. It can read a weird switch matrix with resistance levels, hence the odd looking circuit to monitor it.

BMK:
I have the 8kW version of this.

-Main MCU/DSP is TI (TMS320) Also another small MCU beside it. 2 separate JTAGS, assuming one for each.
-Power relays all Panasonic.
-Cannot see main AC or DC caps, but I don't see Chemicon aywhere.
-Display MCU is Gigadevice ARM thingy.

@Dave. I have a falulty LCD on mine. Would you sell it?

Coyote:
Great review, I do appreciate also the block diagram of the user @Phoenix. It would be nice to have a source code of the Firmware or some reverse engineering around it, since it has many flaws. It would be nice to see also the high voltage model teardown and inverters with two batteries. When two batteries are connected, then inverter stops charging when one of the batteries reaches charge SOC setpoint, even if the other is half empty. But there is even worse: if you have multiple inverters and multiple batteries, inverters will stop charging batteries if one battery of the entire bank becomes charged.

The inverter itself isn't such bad quality, but the FW is a pain. DEYE could do a favor to itself by leaking the FW source code and then collect all, when users would make a stable version of it. 

Navigation

[0] Message Index

[#] Next page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod