diy solar

diy solar

DIY 'Chargenectifier'


That should be helpful.

Thanks!

Ok, should be easy enough.
 
A different option if you don't own a generator. Easily watchable at 1.5x.

Spoiler it works. I wouldn't share it if it failed...
 
Ok, one of them arrived (the one I ordered from that Aliexpress seller). I've not tested it extensively, but it seems to work fine. Voltage output is correct as per the selected one on the store. I'll hook it up to a battery soon-ish, and start playing with the CAN interface.
 
Ok, one of them arrived (the one I ordered from that Aliexpress seller). I've not tested it extensively, but it seems to work fine. Voltage output is correct as per the selected one on the store. I'll hook it up to a battery soon-ish, and start playing with the CAN interface.
That's good to know, I'll probably give that seller a go since I just want a set voltage and to forget about it. I'd prefer something a little over 3kw but that seems to be the popular PSU power, at least within one unit, and curious to see the actual AC current draw from 240v.
 
Ok, so most of what is written in this post is correct (except for a calculation error):


To convert floats to hex, use this:


The pin-out on the connector:

20230411_203308.jpg

I use something like this computer side:


Software I use:

 
Ok, so most of what is written in this post is correct (except for a calculation error):


To convert floats to hex, use this:


The pin-out on the connector:

View attachment 144246

I use something like this computer side:


Software I use:

Are you able the charge charge power through CANBus commands (most easily through constant current setting)?

I’m interested to know whether one of these rectifiers can be controlled through CANBus to absorb excess AC-coupled solar export to charge a battery instead…
 
Another potential item if you are 120v only would be a EPS1500 (ICX6400-EPS1500). The put out 12v and 54v. I picked one up to test with the rackmount switches I have ,they run ~30bucks on ebay and should in theory do up to 20A of 54V which could make a nice budget DIY option for smaller gensets, with a bonus of running 12v accessories (lights, modem) directly.

The connectors are standard 20 pin ATX but theres some level of sense before output is enabled. It LOOKS like the 12v and 54v outputs are also fully isolated.
 
Another potential item if you are 120v only would be a EPS1500 (ICX6400-EPS1500). The put out 12v and 54v. I picked one up to test with the rackmount switches I have ,they run ~30bucks on ebay and should in theory do up to 20A of 54V which could make a nice budget DIY option for smaller gensets, with a bonus of running 12v accessories (lights, modem) directly.

The connectors are standard 20 pin ATX but theres some level of sense before output is enabled. It LOOKS like the 12v and 54v outputs are also fully isolated.
Wow! Thank you. Contacted a seller on ebay to inquire about output pinout. At over 90 years MTBF they seem like a solid option. 2 in parallel would work nicely with my smaller inverter generator.
 
Wow! Thank you. Contacted a seller on ebay to inquire about output pinout. At over 90 years MTBF they seem like a solid option. 2 in parallel would work nicely with my smaller inverter generator.
I've got a partial pinout. I have NOT tried putting the 3 outputs of a single unit in parallel (but since they are isolated it feels like it should work). I still don't have a way to enable output without the switch it powers, but I also have not opened up a unit. Whichever of us finds answers should post to the thread :)

Ok, looking at my notes:
Viewed from the switch or psu side. Retaining clip at top.

______________===______________
|11|12|13|14|15|16|17|18|19|20|
|01|02|03|04|05|06|07|08|09|10|

2, 3, 4 are +54V and their ground is 17, 19, 20.
7, 9, 10 are +12V and their ground is 7, 9, 10.
6, 16, seem to be NC based on the cable I have.
18 is 5 volts even when not enabled so I'm guessing it powers something in the switch which answers back in some way.

each plug's 54v output is rated about 6.8A ,
3 plugs on the back of the unit and each one is enabled independantly. I don't yet know how
 
Last edited:

Attachments

  • Screenshot_20230417-211815.jpg
    Screenshot_20230417-211815.jpg
    185.4 KB · Views: 12
Are you able the charge charge power through CANBus commands (most easily through constant current setting)?

You can limit the output current, so yes. The output current is set in percent to the rated value of the rectifier from 10% to 121%.
These settings (voltage and current) can either be permanent, or while the command is repeated in a temporary fashion. That is, you can send a command, it will keep the settings of that command while you keep sending that command every couple of seconds or so. When you stop sending it, the original setting will be used after 30 seconds.
 
You can limit the output current, so yes. The output current is set in percent to the rated value of the rectifier from 10% to 121%.
These settings (voltage and current) can either be permanent, or while the command is repeated in a temporary fashion. That is, you can send a command, it will keep the settings of that command while you keep sending that command every couple of seconds or so. When you stop sending it, the original setting will be used after 30 seconds.
Thanks. So a unit with 50A max charge rate corresponding to ‘121%’ would have ‘100%’ charge current of 41.3A and minimum (‘10%’) charge current of 4.13A?

4.13A charging a 48V LiFePO battery would be minimum charge power of ~238W, so achieving an average charge power below that level would require duty cycling the charger.

Do you know if the minimum increment above 10% is +1% or is it greater than that?
 
Oh looks like it is auto ranging up to 240v, I was wrong.
Darn good find!

assuming that pinout is correct (since the document is marked as a draft).

Looks like they number "bottom" to "top" and right to left:
Viewed from the switch or psu side. Retaining clip at top.
BROCADE'S numbers:

______________===______________
|20|19|18|17|16|15|14|13|12|11|
|10|09|08|07|06|05|04|03|02|01|

renumbering my own notes for pins:

7, 8, 9 are +54V and their ground is 17, 18, 19.
1, 2, 4 are +12V and their ground is 11, 12, 14.
5, 15, seem to be NC based on the cable I have. (this seems to be wrong?)
13 is 5 volts even when not enabled so I'm guessing it powers something in the switch which answers back in some way.

1, 2, 4 look to line up (12v)
3 is a "power good" signal of some kind
5 "please fill in"
6, 16, 20 listed as not connected
7, 8, 9 look to line up (54v)
10 is listed as ground for 54v but tha doesn't seem to match my unit.
11,12 listed as general ground (might only be 12v ground on my unit?)
13 is listed as "EPS_on" but I think that means "the eps is on"
14 lines up
15 "EPS_PST" (eps present?)
17, 18, 19 return (ground) for 54v

Ok, so if the pinout claims 11 and 12 are simply "ground" that might help looking at what the potential signal pins are doing while running: 5, 13, 15

If we are lucky its something simple like "pull pin X low"
 
As far as I have tested, it accepts floating points just like the voltage. You could set 10.5% for example.
I will try to make some test set-up to get more details...
If I’m correct about ‘10%’ being 4.13A / 238W of charging power, control at the level of 1% / 0.413A / 23.8W is more than sufficient to track and limit export. 0.5% / 12W control is icing on the cake.

It’s a bit of a pity the minimum charge power can’t be dropped below 10% - is there a CANBus command to turn the charger off of put it in standby so charge current is 0A?
 
Some Python code for those who want to play along:

I got steered back to this thread by your pointer from the other new thread but thought this would be the better place to reply:
I saw your comment about one of these rectifiers essentially being a cheaper Chargeverter but with Modbus control, so just want to clarify a couple things:

- Chargeverter can be used to maintain a battery overnight at just above empty (~10%SOC or ~20% SOC) by setting voltage limits appropriately - can one of these rectifiers have voltage limits set to maintain a 48V LiFePO4 battery no lower than 49.6VDC, for example?

-we‘VE already discussed about the possibility of controlling charge current to absorb excess solar export but in this case, charge voltage would need to be increased as well.

So MODBUS control would allow voltage levels to be increased closer to 57.6VDC and then charge current to be controlled to absorb excess AC export in a ‘Daytime Mode’ before switching back to lower voltage levels for a ‘Nighttime Mode’ acting as a battery keeper?

If you confirm that you’d have no problem supporting both this Chargeverter-like nighttime mode and this Chargeverter-like daytime / dumpload mode using your Python / CANBUS control, I’m going to have to pick up some kit and plunge into the world of playing along…
 
- Chargeverter can be used to maintain a battery overnight at just above empty (~10%SOC or ~20% SOC) by setting voltage limits appropriately - can one of these rectifiers have voltage limits set to maintain a 48V LiFePO4 battery no lower than 49.6VDC, for example?

Yes - see the code on how to set the voltage through CAN. Minimum voltage is 41V.

-we‘VE already discussed about the possibility of controlling charge current to absorb excess solar export but in this case, charge voltage would need to be increased as well.

Max voltage on this one is 58.5V, slightly over 3.65V per cell for a 16s.

So MODBUS control would allow voltage levels to be increased closer to 57.6VDC and then charge current to be controlled to absorb excess AC export in a ‘Daytime Mode’ before switching back to lower voltage levels for a ‘Nighttime Mode’ acting as a battery keeper?

CAN, not MODBUS. But yes, you can set the voltage, and then set the max output current as a percentage of the capabilities between 10% - 121% (rated current in the datasheet = 121%), and you can limit the AC input as well, giving the possibility to reduce the overall power of the rectifier. There is a function to control voltage, current, AC limit, etc. in the Python code. All you need to do is figure out what algorithm you want based on current solar generation, etc. and you can control the rectifier to do whatever you need it to.
 
Yes - see the code on how to set the voltage through CAN. Minimum voltage is 41V.



Max voltage on this one is 58.5V, slightly over 3.65V per cell for a 16s.



CAN, not MODBUS. But yes, you can set the voltage, and then set the max output current as a percentage of the capabilities between 10% - 121% (rated current in the datasheet = 121%), and you can limit the AC input as well, giving the possibility to reduce the overall power of the rectifier. There is a function to control voltage, current, AC limit, etc. in the Python code. All you need to do is figure out what algorithm you want based on current solar generation, etc. and you can control the rectifier to do whatever you need it to.
Apologies for typing MODBUS where I meant to type CANBUS.

So through CANBUS, I could set charge voltage to 49.6V and charge current to max and there will be no charge current as long as battery SOC exceeds ~10% / 49.6V and once it dips below that level, up to 3kW / 60A of charge current will be supplied to offset load.

The during the day, if/when it becomes necessary to absorb some AC export, charge voltage can be increased to as high as 58.5V and charge current can be controlled to absorb 238W to 3kW of export by charging the battery (as long as it is not close to full).

My PLC communicates MODBUS so I’m going to have rig up a MODBUS to CANBUS converter but if all this is correct, it will be worth it.

And I’ve seen the link you’ve posted in the other thread, but for the avoidance of doubt, would you mind redposting here the AliExpress link to the rectifier you are successfully controlling for both voltage and current using CANBUS?
 

diy solar

diy solar
Back
Top