diy solar

diy solar

Over-paneling Victron MPPT controllers.

FilterGuy

Solar Engineering Consultant - EG4 and Consumers
Joined
Nov 26, 2019
Messages
7,870
Location
Los Gatos CA
Folks,
(Sorry, this got to be a long post)
I recently wrote and posted this resource on over-paneling.

A reviewer pointed out the Victron Max Isc Spec on the Victron controllers and another forum member PM'd me about it. Since then I have been trying to get a clear understanding of all of the implications of the Max Isc spec and have made a *little* progress.

* The controller manual implies that the Isc is only an issue if the PV id hooked up in reverse polarity. (And that was my belief till recently).

* On the victron community site, one forum manager (that is also a Victron employee) had a post that implied the same thing as the manual. However, when pressed on it he indicated that there are certain uncommon events where the PV input is shorted together and therefore the Isc spec is important beyond a PV reverse polarity situation.

* I was able to contact a Victron person Via email.... but he just pointed me back to the Victron Community forum manager so that did not help clarify things.

* A Victron Comunity member posted a link to an obscure MPPT Error code document that gives a little bit of info on the 'short' situation.

Err 38, Err 39 - PV Input shutdown​

To protect the battery from over-charging the panel input is shorted.
Possible reasons for this error to occur:
  • The Battery voltage (12/24/48V) is set, or auto-detected, incorrectly. Use VictronConnect to disable auto-detect and set the Battery Voltage to a fixed voltage.
  • There is another device connected to the battery, which is configured to a higher voltage. For example a MultiPlus, configured to equalise at 17 Volts, while in the MPPT this is not configured.
  • The battery is disconnected using a manual switch. Ideally the charger should be switched off before disconnecting the battery, this avoids a voltage overshoot on the charger output. If necessary the voltage trip-level for the PV Short protection can be increased by raising the Equalization voltage set-point (note: equalization does not have to be enabled in this case).
  • The battery is disconnected using a Lithium charge relay connected to the “allow-to-charge” output of a BMS. Consider wiring this signal to the Remote terminal of the charger instead. This shuts down the charger gracefully without creating a voltage overshoot.
Error recovery:
  • Error 38: First disconnect the solar panels and disconnect the battery. Wait for 3 minutes, then reconnect the battery first and next the panels.
  • Error 39: The charger will automatically resume operation once the battery voltage drops below its maximum voltage setting (normally Equalisation or Absorption voltages), for 250V versions or float voltage for the other units. It can also takes a minute to reset the fault.
If the error persists the charge controller is probably faulty.

I pressed Victron for a more detailed explanation of why the controller 'shorts' the input but they declined. My *guess* is that if the output voltage gets above the input voltage, something in the circuit ends up enabling a series of FETs between Positive and Negative. (I do not think it is an intentional short... but it is a side effect of the design. )

Another couple of interesting points is that 1) using specs from real solar panels I can come up with configurations that are not over-paneled but violate the Isc Spec and 2) Feeding the 'bad' configurations into the Victron online MPPT calculator does not indicate a problem.

Victron has called the Max Isc spec a soft requirement, with an implication that it may not always be a problem to violate it, but since they declined any further elaboration, I have concluded the Max Isc spec must be followed closely... even if not over-paneling.

I have written the attached update to my resource and would appreciate any comments before I update the resource section. The key pages about Victron are the last two pages and I have copied them here:

==========================================================================================================

1634422776302.png
======================================================================================================
1634422904435.png
 

Attachments

  • Over-paneling.pdf
    345.5 KB · Views: 27
I would add that "over-paneling" the controller should not raise the Voc. Usually, over-paneling means just adding more panels of the same type or at least same number of cells in parallel. No more voltage at Voc.

What can happen is that during sunny times when over-paneling would over-current the controller, the (buck converter) controller reduces its output power by raising the PV voltage above the non-over-paneled Vmp and this makes the controller get hotter and work harder... (Pwr =I X E)

It could also reduce the PV voltage towards the battery to reduce current somewhat but that is not a guaranteed method. (I don't know of any commercial buck-boost controllers)

That should be a problem only if the controller is WAY over-paneled and on a nice 1000 watt per square meter day.

I don't normally worry about Isc BUT if Victron has some sort of battery over-voltage protection that can actually short the panels, then that could be a neat feature.

Usually when a MOSFET melts into slag, they will short. If the top FET of the half-bridge shorts and the bottom FET does not short and the center is connected to the battery, maybe because the series FET or relay should short, then shorting the PV array makes sense to limit the voltage.

Either that or just add a remote trip breaker to the PV input when an overvoltage might happen.

boB
 
Last edited:
What can happen is that during sunny times when over-paneling would over-current the controller, the controller reduces its output power by raising the PV voltage above the non-over-paneled Vmp and this makes the controller get hotter and work harder... (I X E) That should be a problem only if the controller is WAY over-paneled and on a nice 1000 watt per square meter day.
The controller can reduce the current which will cause the panels to raise the voltage, but the controller does not have direct control over the PV voltage.
 
Last edited:
I don't normally worry about Isc BUT if Victron has some sort of battery over-voltage protection that can actually short the panels, then that could be a neat feature.
A ~neat~ feature, indeed :ROFLMAO::ROFLMAO::ROFLMAO:

This whole process of trying to figure out how to properly operate Victron's gear has been kind of convoluted
 
Any recommendations for SCCs that may be good for overpabeling?
================
I may get this shutdown with my VIctron. The difference is, my shutdown is reset by removing and reconnecting the solar panels. It does not automatically reset off some voltage. This tripping also tends to happen when overpaneled on a full charged set of batteries, and not so much when the SOC is at 50%.

No one of these shutdowns have any error codes. No blinking lights indicating errors on the SCC. I think the Error 38 and 39 don’t appear on the lesser Victron’s they only have a light blinking. OF course my $20 BMS has no readings, not even a light.

I occasionally see shut downs on my over-paneled Victron 75/15. I’ve seen a three panels produce 12 charging amps, and hook three more up and limit the SCC to 14 amps So as not to trip the BMS. So, the sunny days I’ve had this setup, the panels are capable of 24 real charging amps and are limited by the VIctron SCC algorithm.

I had assumed it was the BMS shutting down at 15, which is the max output of the BMS, but I recently added a second battery with a 15 amp BMS, so no battery should be receiving more than 7 amps with a 14 amp SCC limit with plenty of room for slop, but it still tripped. THis only happens once.

I CAN’T SAY FOR CERTAIN THIS IS THE VICTRON, BUT ITS REALLY POINTING TO IT.

With what you’re reporting it makes it seem more and more likely this is the issue.
 
The controller can reduce the current which will cause the panels to raise the voltage, but the controller does not have direct control over the PV voltage.

Yes, the controller has almost complete control over the PV input voltage. Except when the battery side cannot take that amount of power which happens when the battery gets full. This is how the MPPT controller is able to operate at the Vmp max power point in the first place. But the controller cannot just let the PV do what it wants to.

Fully loaded, the PV array outputs higher current but lower voltage (lower power than Vmp) and unloaded the PV array is sitting at Voc.

In-between fully loaded (PV voltage equals battery voltage) and open circuit voltage is where the Vmp lies....

If the proper voltage of PV is connected that is
 
Last edited:
BTW: I sent this document to Victron for comment but have not heard back.
 
One thing I almost hate to bring up is that the wording of the Err38/Err39 kinda resurfaces the old debate about damaging an SCC by disconnecting the battery before disconnecting the PV. It does not say disconnecting the battery first will damage the controller, but it implies it will throw an error.
 
Any recommendations for SCCs that may be good for overpabeling?
================
I may get this shutdown with my VIctron. The difference is, my shutdown is reset by removing and reconnecting the solar panels. It does not automatically reset off some voltage. This tripping also tends to happen when overpaneled on a full charged set of batteries, and not so much when the SOC is at 50%.

No one of these shutdowns have any error codes. No blinking lights indicating errors on the SCC. I think the Error 38 and 39 don’t appear on the lesser Victron’s they only have a light blinking. OF course my $20 BMS has no readings, not even a light.

I occasionally see shut downs on my over-paneled Victron 75/15. I’ve seen a three panels produce 12 charging amps, and hook three more up and limit the SCC to 14 amps So as not to trip the BMS. So, the sunny days I’ve had this setup, the panels are capable of 24 real charging amps and are limited by the VIctron SCC algorithm.

I had assumed it was the BMS shutting down at 15, which is the max output of the BMS, but I recently added a second battery with a 15 amp BMS, so no battery should be receiving more than 7 amps with a 14 amp SCC limit with plenty of room for slop, but it still tripped. THis only happens once.

I CAN’T SAY FOR CERTAIN THIS IS THE VICTRON, BUT ITS REALLY POINTING TO IT.

With what you’re reporting it makes it seem more and more likely this is the issue.
I found this statement in the err38/err39 description very interesting:

The battery is disconnected using a manual switch. Ideally the charger should be switched off before disconnecting the battery, this avoids a voltage overshoot on the charger output. If necessary the voltage trip-level for the PV Short protection can be increased by raising the Equalization voltage set-point (note: equalization does not have to be enabled in this case).
It looks like they provide a back door into reducing the sensitivity to the error condition even if the equalization cycle is not enabled. You might try raising the equalization voltage without enabling equalization. (In the past, if I played with the equalization voltage on a controller I would make it very low because LiFePO4 should not be equalized. However, now I am not sure that was a good idea)
 
Most of my systems get the greatest benefit from long hours of charging rather than high peak current - so more panel (above SCC rated current) gives the best performance.
As I'm running epsolar SCC's, this is a quote from the manual:

1634427495536.png


The caveat about OC-V doesn't make sense to me; The OC-V doesn't change between having one or an infinite number of (identical) panels when connected in parallel.


 
Last edited:
Yes, I have overpaneled Victron 75/15s. At the time I did not even look at Isc but the system worked fine.
Well, if its working for you and not me, could easily be my BMS.

This is kind of frustrating because I set this up before work and come home after work to check on it and I always wonder if I get this shut down and I will return home to find my BMS shut down because of low voltage. I cook with this on a crockpot which draws 8.3 amps at 24 volts.
I found this statement in the err38/err39 description very interesting:

<<Error 39>>
It looks like they provide a back door into reducing the sensitivity to the error condition
If I keep getting this mystery shutdown, I may try increasing the equalization voltage, but ensure it is disabled.
 
There are sooo many variables that it is hard to make any kind of conclusions.
Seems to be better with the twice batter capacity and not tripping, but if I come home once and my food is spoiled because the batteries never started charging and they died four hours after turning the crock pit on to cook, my answer will probably be spend $25 on a 40 amp charging BMS, and parallel the battery cells into 2P8S into one battery. I may even get a 100/30 Victron or Schneider or other SCC.

I don't see going more than 600 watts of panels which based off performance should be able to deliver at least 24 charging amps, even though the math works out to 11 amps maxed out charging a battery at 27 volts. These Lion Energy portable panels are actually operating near specs or exceeding specs for four hours a day.
 
Most of my systems get the greatest benefit from long hours of charging rather the high peak current - so more panel (above SCC rated current) gives the best performance.
As I'm running epsolar SCC's, this is a quote from the manual:

View attachment 69117


The caveat about OC-V doesn't make sense to me; The OC-V doesn't change between having one or an infinite number of (identical) panels when connected in parallel.



If you dig into them, a lot of the MPPT controllers have limits like this. They are typically 30-50% more than the rated charging power.
It turns out that Victron has a limit to....but the limit is hidden in the Max Isc spec.

I agree that the Voc comment is kinda weird. If the array can exceed the voltage limit, it will be a problem regardless of whether it is over-paneled or not. I guess I should get used to weird explanations for over-panel limits. I could not get a sensible answer out of Victron either.
 
MidNite told me in an email I could essentially Overpanel a classic 150 all I wanted, just don't allow the array to exceed VOC. The controller will only use what it can use and the rest is wasted. And That the excess would not dissipate as heat, it just simply wouldn't be used. But now reading @SpongeboB Sinewave 's comments I'm second guessing my decision.

24v system with Present array: 2565w 3s3p (VOC 116.4 / Isc 29.13). Plan is to double that with same panels to a 3s6p (VOC 116.4 / ISC 58.26). I realize it would be more effective with a second scc, but my understanding from midnite was it won't hurt the classic 150 as long as I don't exceed VOC .
 
My *guess* is that if the output voltage gets above the input voltage, something in the circuit ends up enabling a series of FETs between Positive and Negative. (I do not think it is an intentional short... but it is a side effect of the design. )

Could be intentional. That is called a crowbar, and is common in power supplies (some manuals warn to use a diode when charging batteries for that reason.)
Transistors would be damaged if voltage went too high, and this would prevent it.

It is common to have a reverse-polarity diode on PV input.
SCC buck architecture would have a pullup FET from inductor to PV input capacitor. It would have a diode between inductor and ground for freewheeling. But diode has 0.7V drop, and FET can reduce that considerably. So there are two FETs that could be enabled for this function (discharging input capacitor in the process.)

 
Back
Top