Dynamic Current Limiting

[mike90045]

Why not have the WBJr at the battery only, and tap the inverter feed at the controller.  Wire management (ohmic loss) becomes important, and your inverter may appreciate a super cap at it's input for 60/120hz ripple suppression. 

instead of the WBJr at the output of the classic, then feeding the battery bank & inverter & loads

Classic -> Busbar :  (busbar wired to
Inverter
DC Loads
WBJr -to - battery bank
This way, the WBJr measures in and out of the battery only.
  BUT it isolates the inverter a bit from the huge capacitance that the battery bank provides for the AC peaks the inverter draws from the battery.   If your loads are not much the internal cap in the inverter will be fine, but with heavy AC loads, since the inverter is not tightly coupled to the battery anymore (WBJr , cables and bus bar isolate it) you may want some low impedance capacitance that will handle the 60/120Hz ripple on the DC wires as the inverter creates the AC waveform.

The Classic records it's daily harvest, subtract that from the WBJr and you have the direct loads consumption

AND the WBJr will record the end amps of the battery in order to properly terminate the Absorb cycle, without the loads throwing it off.   

BUT the risk is some inverter foul-up if it does not like being decoupled from the battery bank.  Way less than an ohm, but
could have an effect.

[EW Zuber]

A 500A 50ma shunt is something like .0001 ohms which is less than a lot of battery connections.
But thanks for your alternate layout. Something to consider.

[mike90045]

A 500A 50ma shunt is something like .0001 ohms which is less than a lot of battery connections.
But thanks for your alternate layout. Something to consider. 

It's the shunt, and all the extra connections, the resistance adds up, and can sometimes effect a demanding system.

[WillEert]

I plan on using solid state switching with PWM trigger pulses from the classic diversion output to switch loads off and on.



I tried this method of diversion and struggled to make it work.

If you use the AUX2 PWM then you cannot have a WBJr.

Also I was diverting on the AC side. The Classic PWM is so fast that the Magnum  inverter could not keep up with the load changes and the whole house acted like a disco - lighting flickering, etc. The place was not livable. Maybe another brand of inverter would be better ( possibly a new Midnite inverter??)  Another issue is that with the Classic you cannot put more than one Classic on an array. This means with a 10KW array you will have 3 Classics and 3 arrays - joining the array output together after the Classic. A great way to divert and make hot water is to design your PV string voltage to produce  a suitable voltage for standard elements and divert excess DC before the charge controller using suitable diodes and PWMing the DC. The problem with this concept is that with bigger systems, as I understand it, you cannot join the DC diversion wires together as that effectively joins the arrays together. This would limit DC diversion systems to the output of 1 Classic for each system. That is why I went to AC diversion.

At present I  use a PWM diversion controller that PWMs an AC SSR at 10hz. The inverter is very happy at this frequency and my electrical system is also. I lose efficiency in the inverter but I vent the heat to assist in heating my house. The limitation to an optimized offgrid system size, for me,  now becomes inverter capacity. I put all my extra power into a domestic hot water tank and send any excess hot water, via a heat exchanger, into a radiant floor for house heating. The mass in the radiant floor acts as a battery to get me through the cloudy days however a backup heating source is necessary.

Will

[boB]

Hi Will !   You know that the Classic's Aux 2 PWM diversion, at least the clipper PWM can be set
to 20 Hz or so ?  Set it to DC rather than AC in the Clipper menu.  I can't remember  now if
that also changes the other PWM modes to 20 Hz from several hundred Hz ?  20 Hz assumes that
the PWM width is set to 1.0 volt.  Higher width will lower the PWM frequency more...  Actually, for
the DC setting, this  is based on  PV input voltage and not battery voltage and the width is actually
quite a it more than 1 or 2 volts but for PV input that is high  compared to the battery voltage (usually),
it's OK to be wider.

I think that the Magnum flashes your lights because of its slow(ish) voltage regulation when
loaded and unloaded....   If it were PWMing the DC side, it would more likely be much better but
DC diversion is harder than AC diversion.

And yeah, you can't parallel the inputs of the Classics, at least right now.  Shouldn't be
too big of a deal since  you  need more PV anyway or larger copper if you were going
to parallel PV for higher power.

BTW, How's that modbus controller coming ?  That will take care of not enough Aux 2 outputs.

boB

[WillEert]

I think I am hijacking this topic but...

The modbus controller is doing very well. With the help of Dgd it is up and running and squeezing every watt out the array that it can. The code is posted on the arduino portion of the forum. I also posted code for the dynamic amp load control that controls amps to the batteries by manipulating amps out of the Classic. This lets the system "siphon off" any spare power from the array while protecting the batteries from too high a charge rate. I am waiting to see the Midnite solution to dynamic battery amp load control with some interest.

My problem now is implementing some kind of follow me system for multiple Classics while I am already using the 3 RJ12 jacks for 2 MNGPs and the modbus diversion control. Working on getting my arduinos talking to each other in order to force floats when needed.

Will

[Hein]

So I am about to release a firmware update (BETA) that has a feature to current limit on the WBjr amps as well as the classic amps. It also has a 16bit timer changed to a 32bit timer and has stopped ALL of my network crashes on the new code. Give me a few days next week to test it after it is all compiled and I will post it up for everyone

Hello Halfcrazy,

Has this function been incorporated into firmware? What version?

[Halfcrazy]

Yes we have added that to the production firmware, I have also documented how to change it as well in this white paper:

http://kb1uas.com/MidNite/WhitePapers/Setting%20the%20WBjr%20current%20limit%20on%20the%20Classic.pdf

[Hein]

Got it, thanks Halfcrazy.


Now my question for the experts - how do I limit current to the batteries when using three Classics?

Currently I have the WBJR sitting right before the batteries, so it measures only what actually goes into the batteries from the three classics - after power has been deducted to the inverter.

I have tried to work it out in my head using three WBJRs, but I cannot get a solution. What I need is to limit the combined output of three classics to my battery bank, and to leave the rest of the juice coming from the solar array to go to the inverter for heavy loads during daytime.

Advice welcome!

[Hein]

This is the best I could come up with. If I divide the desired max Amps equally between the three WBJR's, would this work?

[Halfcrazy]

Well the first thing to determine is HOW low do you need to limit too. You say you have 3 Classics but what is the limit into the battery you are trying to stay at or below? In my case I have 4 Classics but My battery can handle 200 amps so I only need one classic to back down to be safe

[Hein]

I currently only have 6 x 220Ah's hooked up in 24V, so when I limit charging to 10% of 660Ah it leaves me with 66A max charging current, or 22A per classic. Each Classic is connected to 1,800W PV, which means it can theoretically deliver 75A each, or 225A combined. Actual amps obviously being much lower.

I use the extra amps to run a 9,000BTU a/c unit to help cool the batteries during daytime and I draw quite a bit of power in my workshop during the day. So I need all the power from the PV available but without frying my battery bank.

Currently I limit each Classic in the charge settings to around 35A each, which leaves around 60A for the batteries while the a/c and some other loads are drawing, but this can spike or drop as the loads switch on/off. If I activate too many loads it starts drawing from the bank while I have tons of pv available, but limited in the classics' settings.

The sketch in my previous post - is that possible or am I being silly to think that could work?

[EW Zuber]

Thanks so much for this additional function. have  Wbjr here to hook up. Can't wait to try it out.

[WillEert]

I am curious how this Classic/ WBJr method of current limiting is working. I am helping a person with an off grid system and he has purchased so many panels that he is going to have current limit problems for his budgeted AHr battery bank. It would be best if he could use this method of current limiting using the Classic, WBJR and correct version of software.

Does the current limit setpoint limit the Classic to the WBJr current setpoint or does the Classic limit change dynamically with system load to control maximum current to the batteries while allowing the array to produce it's maximum power then limited by the Classic input or output setpoint?

I think by now some people would have experiance with this method. Any feedback? Comments?

Thanks

Will

[Halfcrazy]

It is basically constantly watching the shunt and making sure it does not exceed the WBjr current limit (IF Set) if it does start to exceed that it will limit current to maintain the setpoint into the battery, Basically loads first then battery charging so if the loads are not great enough and battery current exceeds your setting it will limit