RuiDeng Riden RD6006 DC power supply

[ptluis]

Some updates...
https://mega.nz/file/9lMVha7A#KJNN_WOBqJyCZf8KIIIZF8nv3Ys86f4i8T4Eu240hYE

RD60062_V1.36.1e.bin (RD6006)
RD60065_V1.40.1e.bin (RD6006P)
RD60121_V1.34.1e.bin (RD6012)
RD60181_V1.36.1e.bin (RD6018)

Unisoft, nice addons 

I have a suggestion 

Since the procedure to calibrate requires that (0.1v and 30V) and (0.1A and 3A) to be entered I think it's more intuitive to do it like I show you in the picture below

The labels on PC software are not clear so I modify my PC software to show as follow (for those who don't undertand the labels on the original software):




///--------------------------------------------------------------------\\\

My sugestion for RD unit:




Explanation:

A. Voltage / Current values to enter for calibration procedure
B. Red box to call atention this is a calibration process that mess with precision of the unit
C. "Back" should be called "Display" (makes more sense)
D. Line splitting "V-BAT" and "Input" from the Other settings since (0.1V/A and 30V-3A) are not related to this calibration
E. Group Voltage adjustments together (output & Display)
F. Group Current adjustments together (output & Display)

Note: "REAL Output" adjustments require a precision multimeter connected to the output of RD60XX and Output turn ON, for those who haven't read the instructions.

That's all folks! 

[dougg]

Here is a note that users may want to keep in mind, especially if they use desktop power supplies (like laptops use) that are >= 100 Watts to drive a Riden PS. Such power supplies will typically have a three pin AC plug and some have an IEC connector on the AC side of the power supply itself. As far as I can tell these PS will have the protective ground wired to the negative DC output. [Evidently 48 Volts supplies used in POTS (plain old telephone systems) have positive earths!]. Looking at the indications printed on a desktop PS the one to look for is the "double insulated" one which is a box within a box. If you see that symbol, then the DC side will be floating and the PS has a 2 pin DC connector and thus no protective ground connection. There are probably exceptions, a multimeter (in ohms or continuity mode) can quickly determine if either DC output terminal is connected to the AC's protective ground.

Why does this matter? It matters because the Riden series of DC/DC switchers are _not_ ideal in the sense of having a common (usually negative) terminal on both the input and the output. According to the RD6006P schematic there is a MOSFET and 10 Amp fuse in series between the input negative terminal and output negative terminal. So if one uses a non-double_insulated DC power source for the Riden and then connect the Riden output to a circuit that a desktop oscilloscope is connected to, then you have a major issue. Desktop 'scopes (and some signal generators such as the Siglent SDG2042X) have their ground connected to the AC's protective ground.

<A few days later> While doing a firmware update (to Unisoft "1e") I realized the USB port on a RD60xx wired to a desktop PC could be another "fuse blowing" scenario. Even small laptops (e.g. my Lenovo X270) with a floating PS can be connected to protective ground! In my case this happens via my Dell U2515 screen which is connected to the X270 when I'm not on the road. I'm guessing Riden have run into this problem and it explains why a relatively cheap DC switcher has added a few dollars to its cost with an _isolated_ USB port. IOW the ground on the USB Micro B receptacle on the RD60xx PSs is floating with respect to all other inputs and outputs on the RD60xx (apart from D+, D1 and Vbus on that receptacle, of course).

[Pukker]

Here is a note that users may want to keep in mind, especially if they use desktop power supplies (like laptops use) that are >= 100 Watts to drive a Riden PS. ..............

Thanks, didn't realize that.
Checked it and some have earth connection to minus and some not.

[UniSoft]

I realized the USB port on a RD60xx wired to a desktop PC could be another "fuse blowing" scenario. Even small laptops (e.g. my Lenovo X270) with a floating PS can be connected to protective ground! In my case this happens via my Dell U2515 screen which is connected to the X270 when I'm not on the road. I'm guessing Riden have run into this problem and it explains why a relatively cheap DC switcher has added a few dollars to its cost with an _isolated_ USB port. IOW the ground on the USB Micro B receptacle on the RD60xx PSs is floating with respect to all other inputs and outputs on the RD60xx (apart from D+, D1 and Vbus on that receptacle, of course).

USB part of RD is complete isolated, including ground...

[chadodesu]

Hello folks,

First post here, wonderful resources we have here, and wonderful work being done by the community.
My question comes in surrounding automatically adjusting the power curve as the voltage increases.

I understand that this new firmware has a over power (watts) protection, and I understand that the PC app has staged charges with time and or voltage triggers but I am wondering if there is any type of automatic curve.

I'll explain my scenario a little more in depth, as I am sure this may be a more common then expected setup.

My target load: A 24v EVE 280ah 8 cell LifePo4 running through a overkill solar clone BMS.
My primary power supply that feeds the Riden: Rigol DP832 with channels 1 and 2 in series to output 64 volts.
The last power supply is the Riden RD6018W with the latest Unisoft Firmware 1.36.1e

Here is my problem:
In order to limit my charge time, I try to charge my LifePo4 pack at the highest current I can.
Each channel of my Rigol caps out at 3.2 amps. if you go over, it OCP's and kills that channel.
The max I can put out via the Rigol is 3.2 x 2 or 6.4 amps at 64 volts.

Now if I have a dead pack and I start my charge at say 26v at the highest amps I can before I trigger my Rigol's OCP it would be say 10 amps.
Because I only know how to set the current to a static setting, as the voltage of the pack increases, the current stays the same, so I start at 200 watts for 26 volts, but as soon as it hits 27 volts or 211 watts, it OCP's the Rigol as it requires more current then the Rigol can put out, and I have to adjust it in timed steps all the way down to 7 amps when the pack is near max 3.5/3.6 volts.

Finally the question:
Is there a way to set a target voltage of 29.2v, but automatically adjust the current in such a way that it goes to 200 watts, but never goes over and ramps the current down according to the voltage curve?

If there isn't, can this be implemented?
Thanks again everyone!

[Maxie]

Hi chad, welcome.

Your calculation is wrong.

Rigol DP832 with channels 1 and 2 in series to output 64 volts.
The max I can put out via the Rigol is 3.2 x 2 or 6.4 amps at 64 volts.

When you series your Rigol, you don't get more Amps, only the voltage doubles.
You can get 30V x 3A = 90W power per channel, so 180W in total.
There is no "curve" charging in the RIDEN firmware, but you can (should) set "Max Power".
From the manual:
  "Limits output power to set value.
    - Protects the primary power supply if not capable of providing full power to the RD60xx.
      Manufacturer recommended setting:
      No higher than 95% of the rated power for the Primary Power Supply."

So for 180W power IN, set the Riden to 171W power OUT. This can be delivered to your batteries.
This is the best case but buck efficiency might be worse than 95% when regulating 64V down to 24V.

A fully charged LiFePO4 cell is 3.6V so 8 in series are 28.8V.
This is the V out you have to set your RIDEN to.
BTW, you can parallel your Rigol channels to get 32V 6.4A out. It is more complicated, you need to make sure that one output doesn't feed back to the other, but you might get better buck efficiency on the RIDEN side. It doesn't change the power calculations though.

I don't know much about LiFePO4 but Li-Ion must be charged CC-CV (Constant Current then Constant Voltage).
With the above values, current limit your Riden to 171W / 28.8V = 5.93A.

As you charge, the voltage will slowly increase - keeping the current at 5.93A - until it reaches the preset 28.8V value.
At this point the charging will change to Constant Voltage and the current will start to decrease.
With such a large battery pack I would stop at around 300-400mA.
Using this method you only reach peak 171W output power at the end of the CC phase, when the voltage reaches 28.8V and the current is at max.
Before this point the voltage is lower, after this point the current is lower so both Input and Output power will be lower. (not much)
I know this is not the possible fastest way, but it doesn't require someone sitting there the whole time manually tweaking values with the tongue at the right angle.   

However, I think you are on the wrong way charging like this. I don't recommend it. Here is why.

You have a 280ah pack.
Charging at a constant 6A will take 46.6 hours, not counting losses and the short CV phase at the end.
Thats 2 days!

Meanwhile you are running the Rigol at 32V 3.2A output, which is 114% of its rated power.
I am almost 99.9% sure the manufacturer didn't design that equipment to survive such load on the long run.

In other words, you'll F up your primary PS. Which seems to be a pretty nice unit.
If you are lucky, only 1 channel will die and when you get it fixed the other one will die in a few days or weeks.

So if you are rich and money doesn't count, buy a proper charger for your pack.
One would think a nice charger that speaks 7 languages can be had for 2-3 hundred bucks.
If you are not rich, save your Rigol and buy (or build) an economy charger that was designed to do this job.
In any case, with a dedicated charger you can charge this pack in 6-10 hours, max a day.

[Maxie]

Oh, I almost forgot. When charging at 200W, use the external temp probe.
You probably know your batteries, how much they get warmer during charging.
I'd probably set a cutoff temp ambient +10 Celsius.

[myf]

Hello,
Thanks Dougg !

Here is a note that users may want to keep in mind, especially if they use desktop power supplies (like laptops use) that are >= 100 Watts to drive a Riden PS. Such power supplies will typically have a three pin AC plug and some have an IEC connector on the AC side of the power supply itself. As far as I can tell these PS will have the protective ground wired to the negative DC output. [Evidently 48 Volts supplies used in POTS (plain old telephone systems) have positive earths!]. Looking at the indications printed on a desktop PS the one to look for is the "double insulated" one which is a box within a box. If you see that symbol, then the DC side will be floating and the PS has a 2 pin DC connector and thus no protective ground connection. There are probably exceptions, a multimeter (in ohms or continuity mode) can quickly determine if either DC output terminal is connected to the AC's protective ground...

So there is a connection (A) between the DC(-) input and DC(-) output inside the Riden rd60xxx power supply.

I am measuring full isolation (with a multimeter) between the AC ground and the two DC-(+) and DC-(-) outputs of a generic Chinese "brick"  power supply (500W-12V). I also assume that the two AC-hands input wires are fully isolated with the two DC outputs but I haven't tested it online (I'm afraid of an electric shock...).

So can I really mount 2 bricks-PSU with 2-riden-rd60xxx in series or derivative?
Of course, I'll add a diode in series to protect each converter against reverse current...

Does a reverse diode between the two DC outputs (-) and (+) also prevent a rd60xxx converter from burning out?
What is the best way to do it ?

I can't mount a single PSU on two Riden rd60xxx on the same circuit, unless I connect the 2 DC-(-) outputs of the rd60xxx together.

I also have to be careful with a 100W-36V laptop PSU and the previous small Rideng dps3005 about the ground...  I will measure.

Please feel free to correct if I have said any mistakes ! 

F. from France.

[jesserockz]

Hi UniSoft,

Not sure if I missed a setting, or if its just not available (yet), but what would be nice is to be able to enable the wifi power when in TTL mode for when using custom firmware (ESPHome) on the ESP module.

Currently it requires modifying the circuit to disconnect the EN pin so the wifi board gets power all the time (currently using a cable so I did not have to actually modify).

Thanks for the great work so far!

Jesse

[Maxie]

Hi Jesse,

Do I understand well, that you have a custom FW on the Riden WiFi module and it is still usable in the Riden?
Could you please share more info on that? I would like to use the WiFi module to play around with for other purposes, to be able to upload my own sketch, but I'd also like to keep the option to pop it in the Riden if I want to.

[bateau020]

For custom software, you might find interesting stuff here:

https://community.home-assistant.io/t/riden-rd6006-dc-power-supply-ha-support-wifi/163849
https://cuttlefishblacknet.wordpress.com/2020/03/01/riden-rd6006-wifi/

[jesserockz]

Hi Maxie,

Yes I have flashed ESPHome to the wifi module and when the PSU is in TTL mode, ESPHome can read and write the registers via uart/modbus to get the current values, turn on/off and change set points etc. See docs here for generic modbus implementation: https://esphome.io/components/sensor/modbus_controller.html

Here is my current configuration: https://gist.github.com/jesserockz/664c1e3ac505757d6229f2473a4aa442

Jesse

[Mr.B]

Hi Maxie,

Yes I have flashed ESPHome to the wifi module and when the PSU is in TTL mode, ESPHome can read and write the registers via uart/modbus to get the current values, turn on/off and change set points etc. See docs here for generic modbus implementation: https://esphome.io/components/sensor/modbus_controller.html

Here is my current configuration: https://gist.github.com/jesserockz/664c1e3ac505757d6229f2473a4aa442

Jesse

Very cool.
I will have to try that with mine tonight.

[ShayBox]

How'd you get the ESP to register an IP to the riden so it keeps it powered past the wifi config screen

[Maxie]

ShayBox, if you mean the factory WiFi firmware, check out the documentation by sunkmail
https://www.eevblog.com/forum/testgear/custom-firmware-reference-document-for-riden-rd60xx-power-supplies/

If you refer to Jesse's post, he said:

... flashed ESPHome to the wifi module and the PSU in TTL mode...

[luudee]

Hello Guys,

In the endless pursuit of low noise, I decided to get a P version RD6012 and toroidal transformer.

Unfortunately, I did not pay attention to the Max Input Voltage limitation of the RD6012P. I am
slightly above what the unit can handle (73V DC).

What's an easy way to drop 3V (preferably without generating a lot of heat) ?

Thanks,
rudi

[diogoc]

What's an easy way to drop 3V (preferably without generating a lot of heat) ?

I think the better way is to remove some windings from the transformer.
In a toroidal transformer this is easily done if it is not resined in the center

[Pukker]

Hello Guys,

In the endless pursuit of low noise, I decided to get a P version RD6012 and toroidal transformer.

Thanks,
rudi

10 x 15000uF? What about the inrush-current?
Or a very strong and slow fuse?
but also in my opinion is the old way, a torodial with big caps is the finest solution.
Your selfmade diodebridge looks nice.

[luudee]

I am trying to verify/reproduce the 20mV ripple they are claiming (in the data sheet of the RD6012P).

I set my RD6012P to 5V, 12A, attached an inductive load (spool of wire) that draws roughly 8.4A.

I also attached 100nF capacitor to the output terminals, and the scope probe as close as possible.

Attached are pictures.

First is with the scope (SDS5104) full bandwidth, then 200Mhz, and finally 20MHz bandwidth.

And yes, when limiting to 20MHz, I get close to the claimed 20mV, but why does the noise above 20MHz not matter?  It's still noise ...

What do you guys think?

Thanks,
rudi




 

[luudee]

10 x 15000uF? What about the inrush-current?
Or a very strong and slow fuse?
but also in my opinion is the old way, a torodial with big caps is the finest solution.
Your selfmade diodebridge looks nice.

I figure the rectifier is way over specification (4 x 40cpq080), so the transformer will bend a bit at power on ;-)

Fuses ? Those or for guys who don't like excitement in their lives ... ok ok I need to add one


Cheers,
rudi

[Meter Deter]

just want to start off by saying big thanks to Unisoft.
i purchased my RD6006 with out knowing about , Unisoft fantastic Firmware.. i found this website by chance.. i google the loud sound the fan makes with first start up. pointed me to this site. where i found the Firmware. very happy with PS epically with this customer firmware .

[UniSoft]

announced RD6024

[Cliff Matthews]

What's an easy way to drop 3V (preferably without generating a lot of heat) ?

Try 6 diode drops using 3 series connected 25-amp bridges (ie: AC tabs unused) between bridge and filter caps.

[luudee]

Try 6 diode drops using 3 series connected 25-amp bridges (ie: AC tabs unused) between bridge and filter caps.

Thanks Cliff, but I think that will only work with a load, without a load, I will still have the full 73V.
I'll give it a try next week, and will report !

rudi

[Cliff Matthews]

At 70 ma, the 6 diode junctions should drop >2.5v with a 10W 1000 ohm capacitor bleeder. It's simple and bridges are cheap and easy to mount. One drawback is that at ~10 amps each bridge will need to dissipate >20 W. You have to balance this fact out with how often you'll need to have >500W on the output. There is also a software maximum power parameter that can help.

Barring this, if you don't want to touch the existing windings or deal with bridge and bleeder heat, you could always extend the primary winding neatly outside (with Teflon wire) by 10 to 40 turns to change the total primary to secondary turns ratio enough to arrive at the same goal and the transformer should also run a bit cooler. Guessing what direction to wind the primary extension will require some testing though. Maybe you could post the primary winding extension question in a projects/design thread?