1st off-grid adventure — need charge controller advice

Hedges said:

If you want to switch both PV+ and PV-, remove the red jumper in middle, bottom of picture.

Click to expand...

Then it would be rated to 300v? Correct?

In that picture, how would a person wire that correctly given it has positive and negative clearly marked?
I know that dc is directional and having the arc snuffing bits going in the right direction is critical, but how is this intuitive?

It can still open a 600V string, because it has 4x 150V poles in series.
However, if the PV string had a short to ground near the PV- wire as well as a short to ground on the PV+ output of breaker, it would be trying to open 600V with 2x 150V = 300V and would fail. I guess that is a case of not protecting for a fault before the breaker.

The proper use of polarized breakers isn't very clear once we have sources on both sides, like PV and battery.
Most positive source (PV+) should go to terminal labeled "++", most negative (PV- if breaking both poles, SCC positive input if only breaking one) goes to terminal labeled "-".

Those four poles aren't all labeled, but must alternate between ++ at bottom of picture & -- at top, with ++ at top and -- at bottom. Internally I guess that means they have magnets flipped one way or the other.

We'd rather have non-polarized breakers. There are supposed to be some, but hard to find.

Hedges said:

It can still open a 600V string, because it has 4x 150V poles in series.
However, if the PV string had a short to ground near the PV- wire as well as a short to ground on the PV+ output of breaker, it would be trying to open 600V with 2x 150V = 300V and would fail. I guess that is a case of not protecting for a fault before the breaker.

The proper use of polarized breakers isn't very clear once we have sources on both sides, like PV and battery.
Most positive source (PV+) should go to terminal labeled "++", most negative (PV- if breaking both poles, SCC positive input if only breaking one) goes to terminal labeled "-".

Those four poles aren't all labeled, but must alternate between ++ at bottom of picture & -- at top, with ++ at top and -- at bottom. Internally I guess that means they have magnets flipped one way or the other.

We'd rather have non-polarized breakers. There are supposed to be some, but hard to find.

Click to expand...

Thank you very much for explaining this.

Zwy said:

You need to define string voltage, amperage and series/parallel.

That decides what IMO is needed. You read the guide? https://downloads.imopc.com/solar-isolator-selection-guide.pdf

Click to expand...

Yes I read that. I understand what's needed, but the diagrams and labels in the picture I posted above, which come from this very document, didn't make any sense. The two I notated as #1 & #2 say "4 Poles" but only show two connections. While I'm still a bit fuzzy, I now understand that they do indeed have 4 poles, but 2 poles are used for each polarity to handle the labeled "Higher Current/Amperage"

Zwy said:

Are you combining strings before the IMO, in the IMO or after the IMO?

Click to expand...

3S2P, Y combined at the panels, ran via 10 AWG. For my panels, this will be ~150V ~22A. Once 6000XP is ordered and received, it will go to 6S which will be ~300V ~11A. I would like to wire it now for the end result of ~300V ~11A, even though it will potentially be ~150V ~22A for a number of months.

Zwy said:

That makes a difference. Before the IMO, you need a 2 pole IMO, just a negative and a positive. In the IMO, you need a 4 pole with jumpers in place so both negatives get combined and both positives. After the IMO, you need a 4 pole that keeps both strings separate, 2 negatives and 2 positives.

Click to expand...

I'm a little confused, I would like to keep the disconnect at a minimum, a safe minimum. While I am not treating my project's voltage and amperage with disrespect, it seems like a rather simple setup from my time reading this forum, reddit, and other places over the past couple of months. Do I really need two layers of IMOs? I was gonna start with this:
the 1000V/30A version which is DC-only and lists "Internal Spatial: Bigger" as one of its differentiators from the lower rated and combo AC/DC version.

The IMO was just an option someone mentioned (and it looks fantastic). I was just not understanding their diagrams and logic for the "4 Poles" versions which only handle "1 String Max."

Zwy said:

For example, running 2 strings of 10AWG 400 feet to an IMO that will combine those strings into parallel will be cheaper than running 1 string of 8AWG 400 feet.

Click to expand...

Great tip, thanks!

Hedges said:

Of course - that configuration works as a SPST switch, good for 600V. Switch PV+ only.
If you want to switch both PV+ and PV-, remove the red jumper in middle, bottom of picture.

Click to expand...

So in the picture as shown, it is for a single wire coming in, and you'd use an identical switch for the other polarity? One entire switch as pictured for positive, and a separate entire unit for negative? And if you didn't want to go that entire distance, you simply remove the middle jumper and get two switches? The outputs would then be right next to eachother, if removing the red jumper in the middle which is on the bottom is the way you'd accomplish having the ability to handle both polarities coming in from your single string. This doesn't seem like the greatest idea.

I will have 3S2P with roughly 150V & 22A, and probably once the 6000XP is ordered and delivered, 6S for 300V 11A. While I'm all for safety, it seems like an extreme to get such a large switch for each polarity — but I will if that's what's needed.

Hedges said:

Good news is, IF you connect polarity correctly, it is load-break rated.
(If not, be sure to make a video and post it under Up In Smoke!)

Click to expand...

So you can throw it while it's under load? How would you normally do it with a non-load-break rated switch, try to turn the load off first? What if it's an emergency vs a planned disconnect, are you SOL with a non-load-break rated switch?

AC/DC said:

are you SOL with a non-load-break rated switch?

Click to expand...

No, you start praying and then grab your insulated gloves, goggles, and a large scissors like object before you switch the breaker.

AC/DC said:

So in the picture as shown, it is for a single wire coming in, and you'd use an identical switch for the other polarity? One entire switch as pictured for positive, and a separate entire unit for negative? And if you didn't want to go that entire distance, you simply remove the middle jumper and get two switches? The outputs would then be right next to eachother, if removing the red jumper in the middle which is on the bottom is the way you'd accomplish having the ability to handle both polarities coming in from your single string. This doesn't seem like the greatest idea.

Click to expand...

No, single unit with all poles ganged to disconnect PV. This one is good up to 600Voc string.

Another guy needed > 600V for a higher voltage string feeding Fronius GT PV inverters, and I was able to point him to a deal on SMA 1000V fused string combiner/disconnects.

Removing the middle jumper was to have two, 300V disconnects. Those can be used in series for PV+/- of one string rated up to 600V. Or could be used for two 300V strings, opening just one end.

Side by side is OK, there is insulation between them, and "creepage and clearance" distance.

AC/DC said:

I will have 3S2P with roughly 150V & 22A, and probably once the 6000XP is ordered and delivered, 6S for 300V 11A. While I'm all for safety, it seems like an extreme to get such a large switch for each polarity — but I will if that's what's needed.

Click to expand...

For 3s2p, no need for OCP. Can use either breaker or switch.
22A x 1.56 = 34A per NEC for thermal/magnetic OCP.
But those magnetic-hydraulic breakers are rated 100%, don't need to be derated to 80% to avoid nuisance trips.
So I think 22A x 1.25 = 27.5A is sufficient.
Assuming 22A is Isc.

AC/DC said:

So you can throw it while it's under load? How would you normally do it with a non-load-break rated switch, try to turn the load off first? What if it's an emergency vs a planned disconnect, are you SOL with a non-load-break rated switch?

Click to expand...

You can throw these breakers, and suitably rated switches, under load. Although, it is still recommended to turn off the circuit first. For GT PV inverters, that means turning off AC breaker. For SCC or hybrids, there may be other ways to shut down.

Non load break rated includes MC4 connectors and touch-safe fuse holders.
It is a safety hazard, so better to just leave current flowing until sundown unless it is starting a fire. Covering PV array should drop current low enough. Otherwise, you do what you gotta do and try not to get hurt.

Hedges said:

No, single unit with all poles ganged to disconnect PV. This one is good up to 600Voc string.

Another guy needed > 600V for a higher voltage string feeding Fronius GT PV inverters, and I was able to point him to a deal on SMA 1000V fused string combiner/disconnects.

Removing the middle jumper was to have two, 300V disconnects. Those can be used in series for PV+/- of one string rated up to 600V. Or could be used for two 300V strings, opening just one end.

Click to expand...

Soooo, you can open just one end? Duh, why didn't I consider that

Hedges said:

For 3s2p, no need for OCP. Can use either breaker or switch.
22A x 1.56 = 34A per NEC for thermal/magnetic OCP.
But those magnetic-hydraulic breakers are rated 100%, don't need to be derated to 80% to avoid nuisance trips.
So I think 22A x 1.25 = 27.5A is sufficient.
Assuming 22A is Isc.

Click to expand...

Yes, it is Isc, my panels are 45.2Voc & 11.02 Isc. I will be using this device which is rated 1000V/30A and is DC-only. Do you see any issue with doing so? It does come with MC4s, but I'm well under 30A for the connectors.

Will I have to change anything when it's switched from 3S2P to 6S and 150Voc/11A Isc becomes 300Voc/22A Isc?

Hedges said:

You can throw these breakers, and suitably rated switches, under load. Although, it is still recommended to turn off the circuit first. For GT PV inverters, that means turning off AC breaker. For SCC or hybrids, there may be other ways to shut down.

Click to expand...

Very good info. The device I will be using is able to be thrown under load correct? I will of course avoid that as much as possible (read: only not stop load first in case of emergency).

Hedges said:

Non load break rated includes MC4 connectors and touch-safe fuse holders.
It is a safety hazard, so better to just leave current flowing until sundown unless it is starting a fire. Covering PV array should drop current low enough. Otherwise, you do what you gotta do and try not to get hurt.

Click to expand...

Great to know. But if I do have the device above, I would just throw it first, right?

The Dihool looks from its label ratings like it is good. 1000V, probably means 500V/pole. Non-polarized the listing says. Never trust what retailer writes in ad, get spec sheet. But I think this is from the Dihool store.

Should be fine to use for load-break. But the suggestion is to always shut down equipment with commands or AC breakers if possible.

The Midnight breakers are also rated for load-break.

150V, 300V, 600V and 11A, 22A all fit within ratings of these devices.

Opening one end stops the current flow, but I want both ends of PV string disconnected before I grab wires. Also want capacitors discharged, and power sources to equipment disconnected. Confirm zero AC volts as well as zero DC, some superimpose AC on PV +/-.

I got shocked from an inverter chassis after I disconnected the battery. With no load capacitors kept AC output running for a while. Thrice bitten, quite shy.


(That time was only twice, read on for more.)

AC/DC said:

3S2P, Y combined at the panels, ran via 10 AWG. For my panels, this will be ~150V ~22A. Once 6000XP is ordered and received, it will go to 6S which will be ~300V ~11A. I would like to wire it now for the end result of ~300V ~11A, even though it will potentially be ~150V ~22A for a number of months.

Click to expand...

Distance? The best way to wire any PV string is higher voltage/less amps. SCC should be chosen based upon this concept. I run 420 feet one way distance with higher string voltage.

AC/DC said:

I'm a little confused, I would like to keep the disconnect at a minimum, a safe minimum. While I am not treating my project's voltage and amperage with disrespect, it seems like a rather simple setup from my time reading this forum, reddit, and other places over the past couple of months. Do I really need two layers of IMOs? I was gonna start with this:

https://www.amazon.com/dp/B0CFCZX9N4

the 1000V/30A version which is DC-only and lists "Internal Spatial: Bigger" as one of its differentiators from the lower rated and combo AC/DC version.

The IMO was just an option someone mentioned (and it looks fantastic). I was just not understanding their diagrams and logic for the "4 Poles" versions which only handle "1 String Max."

Click to expand...

I mentioned it.

I use an IMO at each end so PV power can be cut away from the SCC's or at the SCC's. IMO also has a lockout feature. If you do not have line of sight to a disconnect, then you either employ a lockout lock or use a pair of switches- one at each end.

Much easier to design with 2 before construction than to add another later.

AC/DC said:

Great tip, thanks!

Click to expand...

AC/DC said:

So in the picture as shown, it is for a single wire coming in, and you'd use an identical switch for the other polarity? One entire switch as pictured for positive, and a separate entire unit for negative? And if you didn't want to go that entire distance, you simply remove the middle jumper and get two switches? The outputs would then be right next to eachother, if removing the red jumper in the middle which is on the bottom is the way you'd accomplish having the ability to handle both polarities coming in from your single string. This doesn't seem like the greatest idea.

I will have 3S2P with roughly 150V & 22A, and probably once the 6000XP is ordered and delivered, 6S for 300V 11A. While I'm all for safety, it seems like an extreme to get such a large switch for each polarity — but I will if that's what's needed.

Click to expand...

Most likely you will add more PV at a later date, planning now allows more to be added later at a minimum expense and effort.

AC/DC said:

So you can throw it while it's under load? How would you normally do it with a non-load-break rated switch, try to turn the load off first? What if it's an emergency vs a planned disconnect, are you SOL with a non-load-break rated switch?

Click to expand...

I recommend using a proper load rated switch, otherwise keep a pair of long handled wire cutters handy so you can cut the wire when the switch fails. DC is not easy to break, especially at higher voltages.

What Class-T fuse do you guys recommend? I'm going by a basic formula I saw on the forum (thanks @John Frum!), starting with inverter and working backwards:

3000W max continuous / .85 conversion factor / 20V low cutoff = 176.5A service amps
176.5A / .8 fuse headroom = 220.6A fault amps

So looking for a 250A. Interestingly, when I will move to 48V + 6000XP, calculation will look exactly the same as I double the Watts but also double the low voltage cutoff:
6000W max continuous / .85 conversion factor / 40V low cutoff / .8 = 220.6A fault amps

So this battery should be good with a 250A Class-T — this battery is becoming 51.2V 150Ah with the addition of two more units, it is currently 25.6V 150Ah battery comprised of two 12.8V 150Ah units.

I'm having a tough time location one on Amazon, some say they meet the specs (AC/DC) but the pictures fuse only lists AC — is this okay in a Class-T?

I would multiply by an additional 1.12x for "Ripple Factor".
Current coming from battery isn't the DC average power, it is rectified 60 Hz ripple.
Mean average current at 20V or 24V delivers power to inverter, but RMS average current heats the fuse and wires, and that is up to 12% higher.

The Bussman and other name brand fuses we use do have DC ratings, and AIC is 20kA vs. 200kA for AC.
Blue Sea and others resell brand names.

Hedges said:

I would multiply by an additional 1.12x for "Ripple Factor".
Current coming from battery isn't the DC average power, it is rectified 60 Hz ripple.
Mean average current at 20V or 24V delivers power to inverter, but RMS average current heats the fuse and wires, and that is up to 12% higher.

Click to expand...

Got it. 250A seems to still cover my use case with the additional +12%, which would put me at ~247A.

Hedges said:

The Bussman and other name brand fuses we use do have DC ratings, and AIC is 20kA vs. 200kA for AC.
Blue Sea and others resell brand names.

Click to expand...

Would something like this work?
Do I need a holder? Holders seem to be more expensive than the fuse.

What about something like this to kill 2 birds with one stone?
It lists an AIC of 50k @ 125VDC — not sure about reaction time?

Yes, that looks like the right fuse for your application. (Be sure to use a battery cable with at least 250A ampacity.)

Two reasons for fuse holders. One, avoid wire whipping and shorting to something while disconnected. Careful location and length should minimize risk. Another is in case fuse comes apart (but don't think it will within specs. Also easier to work a wrench, without shorting anything, if mounted.

Here's a nice fuse holder with clips to secure cover (I got snap-on cover, difficult to get of.) Costs as much as some small inverters.


Yes, the Midnight/Carling breakers are good too. And they serve as a disconnect.
Fuses are the ultimate in reliability, but especially with its higher AIC rating the breaker should be fine.


Lithium battery? Consider precharge circuit across terminals of breaker.
Figure out how you will mount it. Meant to be panel mount with two screws. Could bolt to an insulator after attaching cables, or across a pair of busbars.

Got it, thanks.

Hedges said:

Lithium battery? Consider precharge circuit across terminals of breaker.

Click to expand...

Where can I get more info on building something like this? I'm using a resistor as seen in Will's videos, but I see what you mean, if you manually trip the breaker, then turn it back on, you can experience the dreaded inrush.

At this pace I don't think I'll ever get my system up, there's so much to do And I'll probably run out of $ soon, too

[Search] for "Precharge" in Resources:


Can also search titles, and entire forum.

This can be a fancy device, or just a resistor or light bulb with wire.
Any time inverter has been disconnected or switched off by breaker or switch, it would need precharge. If just shut down by control panel commands then it wouldn't.

People report blowing class T fuse and welding relays due to inrush. And we figure it is stressful to capacitors.

Hedges said:

[Search] for "Precharge" in Resources:

Search results for query: precharge

diysolarforum.com

Can also search titles, and entire forum.

This can be a fancy device, or just a resistor or light bulb with wire.
Any time inverter has been disconnected or switched off by breaker or switch, it would need precharge. If just shut down by control panel commands then it wouldn't.

People report blowing class T fuse and welding relays due to inrush. And we figure it is stressful to capacitors.

Click to expand...

Thanks, I use search exhaustively but for some reason I didn’t run into this. The pushbutton design looks good. This is a problem I did not anticipate, do most people not have a disconnect on the inverter? Or is everybody running prechsrge circuits?

Thank you for all your help, and for pointing these things out in an expert fashion — as soon as I started talking about the breaker for the inverter you right away pointed out the problem I would. It have realized was occurring, despite the fact that I’m currently using a resistor to charge the capacitors manually. It’s not lost on me, and I really appreciate it!

It may be that the UL listed/recommended batteries paired with tier-1 inverters take care of this internally.
Some of the economy brands have been discussed in 1000 post threads on the forum, never got working, but the lead-acid battery it had replaced had no such issue.
Lots of people here were making DIY batteries when it was the only reasonable way to go and worked through these issues.

I'm still using AGM, haven't touched lithium yet (except for some UL listed equipment I acquired), so only experiencing it vicariously. Your fuse would have had the problem too, maybe blown out. Maybe not at 24V, but at 48V.

It's hit or miss whether one of us who jumps on a thread happens to foresee a given problem. Seems like a good application for expert systems, create a tree with reading lists customized for each inquirer.

AC/DC said:

Got it. 250A seems to still cover my use case with the additional +12%, which would put me at ~247A.

Would something like this work?

https://www.amazon.com/Blue-Sea-Systems-5118-Class/dp/B000MMC7SY/

Do I need a holder? Holders seem to be more expensive than the fuse.

Click to expand...

I purchase my Class T's from Don Rowe. https://www.donrowe.com/SearchResults.asp?Search=class+T As for precharge, I started out with resistors but moving to 48V forklift light bulbs and the socket.

Zwy said:

I purchase my Class T's from Don Rowe. https://www.donrowe.com/SearchResults.asp?Search=class+T As for precharge, I started out with resistors but moving to 48V forklift light bulbs and the socket.

Click to expand...

Very cool, I like the visual indication of when the caps are filled. Endgame would be to have some sort of automated circuit to do the lightbulb until the measured voltage across that circuit dropped to near 0V then switch off it, and reset the system every time the breaker is tripped.

Discharging the caps is also something worth doing sometimes (like when taking the system apart for transportation), having a resistor on hand for that less common occurrence is still useful.