diy solar

diy solar

Two MPPTs on one string through diodes?

fafrd

Solar Wizard
Joined
Aug 11, 2020
Messages
4,188
I’m looking at adding another string to my DC-coupled 24V system and I’m mulling the idea of having the new string primary-connected to a new 1kW GTIL inverter and secondary-connected to my existing SCC, both connections being made through forward-biased diodes.

The new string has slightly higher Voc and Vmp than the older string and the SCC needs a minimum of Vbat + 2V to draw any power (while the GTIL can operate all he way down to Vbat -2V).

So by connecting the new string through diodes to both MPPTs, I’m hoping the GTL inverter will have first dibs and can draw as much of the new strings power as it wants (by dipping below Vbat+2V where the SCC cannot follow) and then once the GTIL says ‘no mas’ and has more power than it can use, it can raise MPPT voltage above the voltage of the SCC to choke off input power and redirect it to the SCC.

If/when the SCC also has more input power than it can use (battery nearing full) and raises string voltage above Vmp, the MPPT of the GTIL inverter will need to track and once both SCC and GTIL reach Voc, input power from the new PV string will be completely choked off.

Either GTIL or SCC can drop voltage back under Voc once it has need of more power.

Placing a diode in the path means that I’ll be losing ~0.7V / Vmp = ~1.75% of my string power (or less if the diode drop is only 0.5V), but I believe it’s going to allow the GTIL inverter to draw as much power as it can use before redirecting any excess over to the SCC.

Because the MPPT of the SCC is also serviced directly by the existing string, it will never be able the ‘take’ any power from the new string and will be limited to ‘accepting’ whatever power the GTIL decides it cannot use.

If the SCC tries to hog power from the shared string by dropping voltage to it’s minimum of Vbat +2V, the GTIL inverter will drop below that to get all of the current/power from the new string. Once it’s only getting current/power from the old string at Vbat +2V, the SCC will increase input power by increasing votage to Vmp, at which point the GTIL inverter can also increase input power by increasing voltage to just under Vmp (meaning the string voltage of the new string will increase to just under Vmp + Vdiode).

This situation will be stable and will mean all current/power from the new string is going to the GTIL inverter while all current/power from the old string is going to the SCC.

This situation will continue until the GTIL wants to taper down input current/power by increasing voltage all the way to Vmp+Vdiode a which point excess string current/power is redirected to the SCC (including redirecting all current/power to the SCC by increasing voltage sufficiently above Vmp + Vdiode).

Would appreciate any comments on this logic as especially elements I may be missing. And of course, if anyone has tried anything similar, I’m all eyes…
 
My strings are 30A, so I’ll be protecting them with 40A fuses and will need diodes rated for a minimum of 50A.

A diode dissipating 15-21W will obviously need a heatsink, but I found these that look like they would more than fit the bill:

 
What is the goal of this system?
Are all your AC loads offset during daytime. It sounds complicated on the DC side versus managing AC loads to harvest maximum solar.
 
Last edited:
What is the goal of this system?
I’m thinking about using the new 1kW array with a new GTIL to max out my combined production at the 3.5kW max export rating of my existing NEM agreement.

I have a 4kW grid-tied array under NEM with a 3.5W max export power agreement but it never exceeds 3kW of output and is under 2.5kW for most of the day (morning and afternoon).

So I’m thinking about connecting the new GTIL to ~1kW of panels and wiring the sensor to have it limit output to ‘3.5kW minus grid-tied production’.

This would allow the new inverter to output 100% of energy coming from the additional 1kW of panels until combined output of 4kW array + new 1kW array exceeds 3.5kW, at which point the GTIL will begin to throttle-back at that level and diodes will redirect excess energy from the new 1kW array to instead assist my existing 1.1kW DC-coupled array in charging 24V LiFePO4 through SCC.

I’m already getting all the daily charge I need into my battery for purely time-shifting (covering peak consumption after the sun has gone down), so I’d like to find a way to push as much energy from a new 1kW array as I can to the grid without exceeding my 3.5kW export maximum while also not throwing away the excess energy available during the daily ‘plateau’ surrounding mid-day (so use excess energy during the daily plateau to add to stored energy in the battery instead).

I’ve been mulling various schemes to do this based on use of CT sensors controlling switches / relays, but since these GTIL inverters already do a very good job throttling output based on the use of a CT sensor, a solution based on a shared string through diodes would be simpler, cheaper, and more effective (unless there is something I am missing).

2 MPPTs on one string can ‘fight’ but 2 MPPTs fed from one string through diodes should allow one primary / dominant MPPT to hog power until it decides to start redirecting excess power over to the second (higher-minimum-voltage) subordinate MPPT.
 
I don't know about how much flexibility you have with orientation of more panels but have you considered some panels facing east and west to give you more production when your existing system is not maxed out?
 
I don't know about how much flexibility you have with orientation of more panels but have you considered some panels facing east and west to give you more production when your existing system is not maxed out?
The 4kW grid-tie array is facing SW.

The existing 1.1kW DC-coupled array is facing SE (so much more power in the morning).

The additional 1kW I’m planning to add can only go on the SE roof (which is why I’m interested to use tan new 1kW to complement / add to the grid-tied output until the 4kW array nears peak output around ~1pm).

So yeah, it’s exactly that ‘different-orientation’ I’m trying to exploit (while also hoping I can find a way to capture the excess energy for battery charging when the the combined output exceeds my 3.5kW max output level).
 
Hook two things together and they will always figure out something. I don't think it will work as you imagined. Each controller doesn't know about the other. Dropping 2V would mean nothing as the controller would just figure out the array power point was 2V lower. There are people who have throttled a second controller, it is not simple.
 
Hook two things together and they will always figure out something. I don't think it will work as you imagined. Each controller doesn't know about the other.
Yes, I know that. What each controller ‘knows’ is whether dropping voltage results in increased power input or decreased power input.
Dropping 2V would mean nothing as the controller would just figure out the array power point was 2V lower.
Whichever controller has the ability to drop to the lowest voltage is in the position to hog all of the current. Since that current will be at a voltage far below Vmp, input power level will be possible-poor, but the dominant / lower-voltage MPPT can starve the higher-voltage MPPT (while the opposite is not true).

The situation you want to avoid is the two MPPTs ‘fighting’ for the same current, which results is all that current going to the lower-voltage MPPT but at greatly reduced power.

If the ‘starved’ MPPT has another dedicated string of similar Vmp, that will pretty much ‘clamp’ the higher-voltage MPPT near Vmp, allowing the dominant / lower-voltage MPPT to increase voltage to whatever maximum allows it to still hog all current from the shared string, but at much higher voltage & power.
There are people who have throttled a second controller, it is not simple.
If by ‘throttled a second controller’ you mean connected one string to two MPPTs, I’m interested. Do you have any threads to share and / or search terms to try?
 
Back
Top