Rectifier Huawei R4850G2 48V 42~58V 3000w

frank_zero said:
Has any of you put 2 Rectifier Huawei R4850G2 48V 42~58V 3000w in serie ?
can you also mont them standing up or down ? instaid of laying down ?

Yes, I made a charger with 2 in series with them standing on the sides. I think having them laying down is preferred though, less stress on the internals.
 
j bjork said:
frank_zero said:
Has any of you put 2 Rectifier Huawei R4850G2 48V 42~58V 3000w in serie ?
can you also mont them standing up or down ? instaid of laying down ?

Yes, I made a charger with 2 in series with them standing on the sides. I think having them laying down is preferred though, less stress on the internals.

j bjor what are you using them for ?
this is what my seller of the 2 huawei told me To charge 96volt you need to connect them in series.The output will be 107v dc at 50amps.The power supply must not touch each other as they need to be electrically isolated, what do you think about that ?
 
frank_zero said:
j bjork said:
frank_zero said:
Has any of you put 2 Rectifier Huawei R4850G2 48V 42~58V 3000w in serie ?
can you also mont them standing up or down ? instaid of laying down ?

Yes, I made a charger with 2 in series with them standing on the sides. I think having them laying down is preferred though, less stress on the internals.

j bjor what are you using them for ?
this is what my seller of the 2 huawei told me To charge 96volt you need to connect them in series.The output will be 107v dc at 50amps.The power supply must not touch each other as they need to be electrically isolated, what do you think about that ?

I am using them as a charger for my ktm. They dont like to go under about 40v each, the fan goes high speed and the psu might restart after a while.
I dont really understand what you mean with they cant touch each other, they are both grounded. I wouldent use them without grounding. So the outer shell is absolutely the same, and can touch or not. It dosent matter.
 
j bjork said:
I am using them as a charger for my ktm. They dont like to go under about 40v each, the fan goes high speed and the psu might restart after a while.
I dont really understand what you mean with they cant touch each other, they are both grounded. I wouldent use them without grounding. So the outer shell is absolutely the same, and can touch or not. It dosent matter.

That' because the outputs on these Rectifiers is isolated, if they weren't isolated you'd see sparks when touching the case while having them connected in series.
 
Hello I am Cha_ly (Charly in clear) and new in this forum .
I need some help.

I use a Huawei R4850G2 power supply and after reading many articles here just ordered the mentioned chinese can controler to change the rectifier output voltage.
From what I read, the controler works, but non of the owners/users has provided any information of how to use it.
External PS for the controler and the can bus connection is clear so far.

Is there an english "user manual" supplied with the ordered item?
May be someone who successfully modified some default settings of the R4850 could be so kind and provide some instructiuons of how to use it.
I only wish to change the default voltage from 53V to 50V permanently.

Thanks for the help
 
salavati05 said:
rigmo said:
I have direct contact of PCB controller builder from china.. 任星星

hi rigmo
you can help me ?
do you have cod canBUS for R4850G2.

Hi rigmo,
I'm interested too. I bought 2 of these Huawei contollers but I need them for Emerson.
 
Here a easy way to connect toe Recifier no soldering and it cost around 15 euro on ebay,
also an easy way to start it up is to connect like this.
is there a easy way to get 2 can bus controlers so i can adjust the voltage ? i need 5 volt more.
Or has some one 1 of 2 of the can bus controlers for me ?
TB2NyCIdpGWBuNjy0FbXXb4sXXa_!!0-fleamarket.jpg_728x728.jpg
 

Attachments

  • aansluiting.jpg
    aansluiting.jpg
    77.9 KB · Views: 2,256
Hi all
I still waiting for some info how to permanently adjust the voltage of my R4850G2 to 50V using the chinese controller received on the meantime.

No info at all was supplied.

I know how to connect the board - but how to permanently adjust the voltage.
I do NOT want to change the curent limit.

Thanks
Charly
 

Attachments

  • huawei SPS controler.jpg
    huawei SPS controler.jpg
    99.8 KB · Views: 2,659
What is the contact info for the unit's seller?

Maybe someone who speaks Chinese could help put a rough howto together?
 
There is a link to a Chinese forum in this thread and a ton of other posts with info, it's just a few pages back.
 
Has anyone got one of these units charging batteries? How is it?

What has been users experience with the current limit? I have set mine to 1.0A for a test (I understand it can go from 0 to 60A in 0.1A steps). The current control loop appears to be very slow.

When testing at 1A limit, the power supply appears to supply 150% overload no problems - i.e. 1.5A.
When I increase the load to 5A, it supplies the full 5A for about 10 seconds before starting to slowly ramp down the voltage.
 
I tried to control the R4850G2 supply with the CAN interface, botox software flashed on an ESP32, that is mounted on an Olimex ESP32-EVB board. Schematics can be found on their homepage => https://www.olimex.com/Products/IoT/ESP32/ESP32-EVB/open-source-hardware

There are some problems with the board, that CAN in the library uses Ports 4 and 5 for data transmission/receicing and the Olimex board has mapped pins 5 and 35 to the CAN driver interface. So I tried to bridge 35 to 4, but without success.
The problem here is that the ESP doesn't spit out any signals.
After that I tied a regular ESP32 board named ESP32 DEVKIT V1. Same here. I tested all pins for signals, but with no success.

Quick question to the compile process on the arduino IDE of the botox software: The comiler says, that the program is too big (a little bit over 100% of space). After that I deactivated the usage of spiffs and no OTA to "solve" this problem. I hope that doen't affect the program itself.

After struggling with outputting some CAN data from the ESP32 I connected a FTDI board to the pins of the CAN receiver chip on the Olimex board (after I programmed all pins to inputs to put them into high impedance state). Via HTerm I tried to communicate with the PSU, Baud was 125000, 8N1, all values in HEX.
I got spammed from the supply with a lot of data (all the same) and got no response when sending some data.
I tested the signals on the not-can-side of the CAN transmitter chip with an scope - everything looked fine.

Is it possible to send CAN packets with an general purpose terminal program?
I tried (without the spaces and the 0x):
0x108040FE 00 and 0x108040FE 000000000 0000000 for getting some info data
0x108180FE 01000000 0000DC00 for setting 55.0V for the online output

Are there any suggestions to get thes beast working? Yeah. It works, but I like to change the values :)
 
jmibk said:
Is it possible to send CAN packets with an general purpose terminal program?

I wouldn't think so. The CAN frame has addressing/arbitration fields at the start of the frame which you won't be able to replicate with a UART. You need a dedicated CAN controller, like what you have on the ESP32.

Are there any suggestions to get thes beast working? Yeah. It works, but I like to change the values

You wouldn't happen to have a CAN adapter for Linux (SocketCAN) or Raspberry PI/Beaglebone with CAN interface/Transceiver:

If so, I've written a Linux program to set the values:
https://github.com/craigpeacock/Huawei_R4850G2_CAN

It's a little easier to debug on the command line, especially with CAN tools such as candump. The first thing you can do is dump packets from the R4850 to determine if the hardware is working.
 
jmibk said:
Olimex board has mapped pins 5 and 35 to the CAN driver interface. So I tried to bridge 35 to 4, but without success.
The problem here is that the ESP doesn't spit out any signals.

I've just given BotoX's code a go and conclude the same thing.

I've put a DSO on CTX (5) and CRX (4) and there is absolutely nothing. If I load other code (compiled against esp-idf) I can get CAN working. Hence I believe my hardware setup is all O.K.

I don't think there is anything wrong with BotoX's code, but the underlying arduino-CAN library. I tried running examples/CANSender.ino but nothing either.

If baudrate/clock frequencies were incorrect, you would see something on the DSO.

According to the documentation, you can change the pins with CAN.setPins(rx, tx). I've tried, but no luck.
 
I've got BotoX's code working (or sortoff) :D.

--- STATUS ----
Input Voltage: 240.75 V ~ 50.01 Hz
Input Current: 0.00 A
Input Power: 0.00 W
Input Temperature: 30.00 °C
PSU Efficiency: 0.00 %
Output Voltage: 47.96 V
Output Current: 0.00 A / 37.70 A
Output Power: 0.00 W
Output Temperature: 29.00 °C
Coulomb Counter: 0.00 Ah
--- STATUS ----
description
--- HUAWEI R48XX DESCRIPTION ---
/$[ArchivesInfo Version]
/$ArchivesInfoVersion=3.0


[Board Properties]
BoardType=EN1MRC5G1A1
BarCode=2102310FFALUH9011684
Item=02310FFA
Description=Function Module,R4850G2,EN1MRC5G1A1,1U 3000W Rectifier,Power Only
Manufactured=2017-09-27
VendorName=Huawei
IssueNumber=00
CLEICode=
BOM=

I found the baudrate was half of what was programmed, and this was causing a bus error resulting in no activity on the bus.

I'm not sure why the baudrate is half and where that clock rate is derived - i.e. should there be a setting in PlatformIO somewhere, and if you can that, do you break the UART baudrate?

So to hack it, I simply doubled the CAN baud rate and used:

if(!CAN.begin(250E3))

CAN.setPins(rx,tx); also seem to work, so for the Olimex board you can change the IO. Just make sure this is before CAN.begin.
 
I am having some problems with CAN. The ESP32 seems to crash whenever commands are sent, I have tried bauld at 125E and 250E. I hooked up a ssd1306 screen with a running count so I can see when it crashes.

With debug turned on I get some data out

17:59:31.770 -> Starting CAN 125E3 success!
17:59:31.770 -> 1001107E: 30 02 6C 68 84 64 2D E1
17:59:31.823 -> 1001107E: 30 02 6C 68 84 64 2D E1
17:59:31.823 -> 1001107E: 30 02 6C 68 84 64 2D E1
17:59:31.870 -> 1001107E: 30 02 6C 68 84 64 2D E1
17:59:31.870 -> 1001107E: 30 02 6C 68 84 64 2D 88

18:04:18.325 -> Starting CAN 125E3 success!
18:04:22.684 -> 1E000020: 01 00 00 00 00 88 80 08 36 00 80
18:04:22.737 -> 1E000020: 01 00 00 00 00 88 80 08 35 00 80
18:04:22.737 -> 1E000020: 01 00 00 00 00 88 80 08 34 00 80
18:04:22.784 -> 1E000020: 01 00 00 00 00 88 80 08 33 00 80
18:04:22.784 -> 1E000020: 01 00 00 00 00 80 08 8B 32 00 80

I cant decipher what either of those mean.

Logic analyser hooked up to CTX, CRX, canH and canL, can_H looks bad? it should be an inverted copy of can_L?
Can.png
 
kram said:
17:59:31.770 -> 1001107E: 30 02 6C 68 84 64 2D E1

This data looks correct, so you have the right baud rate.

kram said:
Logic analyser hooked up to CTX, CRX, canH and canL, can_H looks bad? it should be an inverted copy of can_L?
Can.png

Let me guess, you have a SN65HVD230 from Aliexpress or ebay?

I have seen counterfeit parts behaving like above. What does the IC look like - does it have a poor molding and engraving like below?

counterfeitSN65HVD230.jpg
 
It could be but the package and etching look good. I poked it with the meter, its showing 30k R to ground on pin 8, which puts it in slope control mode. I couldnt measure any voltage on pin 5 Vref, it should be half Vcc.

It was this
www.ebay.co.uk/itm/SN65HVD230-CAN-bus-transceiver-communication-module-For-Arduino/273769696964

I have ordered a TJA1050 board instead, should be here tomorrow. Thanks.
 
No luck with the TJA1050, the esp32 reads nothing from it while atleast the vp230 can recieve.
I plugged the VP230 back in and programmed the CAN Reciever example, which comes with the lib. With logic analyser I can see the r4850g2 is sending constant packets, but the esp32 will only pick up the first 500ms worth after a reset.

Have I got bad esp32's?
 
The datasheet I read did say 3.3v compatible, https://www.nxp.com/docs/en/data-sheet/TJA1050.pdf "Input levels compatible with 3.3 V and 5 V devices". I think it may be either, no CAN is 3.3v compatable, or just none of these Chinese transcievers.

Anyway I have some progress, I put the SetVoltage(42,false) call into the main loop. I put the VP230 back in, and tried the logic without the psu connected. Now it was trying to transmit and TX looks correct, but the CAN side is wrong.
Can2.png

I wanted to try 5v but didnt want to kill this chip (even if the datasheet says 6v max) so I put the TJA1050 in, I put a 1k/2k R divider on RX and gave the CAN 5v Vcc, which gave me this working output.
Can3.png


edit: The output is now correct, my changes must have messed something up.
22:32:38.211 -> --- HUAWEI R48XX INFO ---
22:32:38.211 -> Rated Current: 52
22:32:48.598 -> --- STATUS ----
22:32:48.598 -> Input Voltage: 246.91 V ~ 49.95 Hz
22:32:48.632 -> Input Current: 0.45 A
22:32:48.632 -> Input Power: 111.26 W
22:32:48.632 -> Input Temperature: 22.00 °C
22:32:48.632 -> PSU Efficiency: 91.02 %
22:32:48.632 -> Output Voltage: 48.86 V
22:32:48.632 -> Output Current: 2.07 A / 40.37 A
22:32:48.632 -> Output Power: 101.26 W
22:32:48.632 -> Output Temperature: 15.00 °C
22:32:48.632 -> Coulomb Counter: 0.01 Ah
22:32:48.632 -> --- STATUS ----
22:33:06.551 -> --- HUAWEI R48XX DESCRIPTION ---
22:33:06.551 -> /$[ArchivesInfo Version]
22:33:06.586 -> /$ArchivesInfoVersion=3.0
22:33:06.620 ->
22:33:06.654 ->
22:33:06.654 -> [Board Properties]
22:33:06.689 -> BoardType=EN1MRC5G1A1
22:33:06.723 -> BarCode=2102310FFALUH9011745
22:33:06.792 -> Item=02310FFA
22:33:06.827 -> Description=Function Module,R4850G2,EN1MRC5G1A1,1U 3000W Rectifier,Power Only
22:33:06.965 -> Manufactured=2017-09-27
22:33:07.033 -> VendorName=Huawei
22:33:07.067 -> IssueNumber=00
22:33:07.102 -> CLEICode=
22:33:07.136 -> BOM=
22:33:07.136 ->

Tried the VP230 again, no different at 5v.
 
Can you please tell me if I can use 3 modules 0-140 volts for serial connection? I need about 400 volts to charge the 96S battery.. I am worried that the diode bridge or output transistor may burn out :bolt:
 
has anyone understand out how to raise the maximum voltage? I need 75v
this is my electric 106 project.
https://photos.app.goo.gl/7Ph1hQFYyCydSPNZ9
I want to make a three-phase charger to connect to the 22kw columns. I want to use 6 huawei rectifier. I would like to change them to 75v to put 3 in series and make 225v. And three more in parallel to have double the current. When I find a 11kw column with orion2 bms with the canbus it lowers the current. I just have to understand the modification to get 75v.
Or I could connect 6 , 41v rectifier in series and do the 246v and with some diodes in series I lower the voltage due to the drop of 0.7v for each diode, there are 12, I put 24. what do you recommend?
 
Back
Top