diy solar

diy solar

To over Watt, or not

"2 mpp trackers" Don't connect those together.
Normally that's the case however it depends on the model.

There are some where two MPPT inputs can indeed be joined together in parallel to provide a single MPPT with greater capacity, they even have provision for a joiner link between the inputs. I can't recall a specific model(s) off top of my head right now but I had this discussion elsewhere about six months ago and they do exist.
 
"2 mpp trackers" Don't connect those together.

Each tracker has terminals to wire up 2 parallel strings.

425 or 455 watt 72 cell panels should be wired in 2 strings on the separate trackers. That does require 4 conductors from the array(s) to the inverter. The Q-cells 425 watt panel has a VMP of 41.2 volts and IMP of 10.32 amps. The VOC is 49.13 volts. With a 500 volt peak input, figure we need at least 25% safety margin because you said it gets vey cold. So 375 volts at 25C to be safe. 375 / 49.13 = 7.6 means 7 panels in series. 7 panels at VMP = 288 volts x IMP = 2976 watts. If you put that on both trackers, you get just under 6,000 watts.

I found the GSL web page describing this inverter, and it does show a block diagram with the 2 + 2 solar inputs and the battery input all going to a common DC bus in front of the inverter block. I can't find any mention of specs for over paneling this unit. The fact that each MPPT tracker input is rated at 18 amps, seems to me, that is the maximum current those inputs should ever see. And here is how I came to that conclusion.

7.6 kw of inverter capacity
Maximum charge rate 190 amps, with a voltage range of 40 to 60 volts. 7,600 / 190 = 40 volts. The current probably has to ramp down as the voltage climbs to stay at 7,600 watts

18 amp MPPT input up to 425 volts. 18 x 425 = 7,650 watts in for EACH of the two trackers. That is up to 200% solar input power. It will most likely limit down to the 7,600 watts. The chosen panels are not the best match to maximize this inverter's solar inputs. Putting 2 strings of the Q-cells panels on one MPPT input would be a waste of the panels. 14 x 425 watts should be up to 5,950 watts. But even if it only goes into current limiting, and does not damage the MPPT input, it holds it down to 18 amps, and we still only have 288 volts. 18 x 288 = 5,184 so you lose almost the full output of 2 panels. If you put 7 on each MPPT input, then you get the full power of 5,950 watts to the DC bus.

Why do I get the feeling this is another Deye 8K inverter? All the specs match the Sol-Ark 8 K as well.
Other then the pv watts are lower, and the 8k has 22+22a they are a lot alike. I'm not even sure why they built 2 so close. 7.6 and 8
 
The PV input voltage, and the PV Array MPPT Voltage Range, have me a bit confused max PV max voltage 500, mppt range 425
 
Other then the pv watts are lower, and the 8k has 22+22a they are a lot alike. I'm not even sure why they built 2 so close. 7.6 and 8
Someone went over the reason for the 7.6kW inverters to exist in the first place: NEC allows 20% over busbar rating input to a panel when multiple inputs are used, apparently implemented to keep people installing solar happy. That 20% on a common 200A house panel works out to 40A, or 7.6kVA/KW.

My guess is they used the same box, same inverter with a software limit on the 7.6, and put in a bigger SCC, figuring the 8K is more likely to be used off-grid with a battery system and those who get it are more likely to want the higher input to charge their batteries when the sun is shining.

RE MPPT range vs PVmax, that probably means string should be kept under 425V, but the components are rated to handle overvoltage to 500V. It might stop producing power over 425V, but it won't risk damage until after 500V.
 
As an addendum, if someone were to open both models and found the same SCC board in both, I would be comfortable relying on the higher of the two Isc ratings, but without that level of checking I would assume they put a beefier SCC in the 22Ax2 8kW model. If someone tries to check, be warned the same PCB can be populated with different power handling parts.
 
It sure would be nice if Deye had some dos and don'ts on this subject.

wattmatters says some can be connected together. GXMnow says don't do it. I won't do it unless I saw something in writing from Deye.
And thanks for looking up the 7.6 GXM.

I also received a PM from someone who didn't want to join in on my Q&A I know very little thread.

He/she said the 9880 is the max watts allowed on the 7.6k. I won't need to worry about that. Keep below the 425v mppt range, not the max pv input 500v using the cold factor. If 2p, use a combiner box with surge protection 15a breakers would be fine for each string. Don't worry about going over a bit on the amps. The mppt is amp self limiting. (Who knows? The manual certainly doesn't state anything.)

I also went to rapidtables.com to calculate watts and volts into amps, of the various panels I can get. The only one that's over is the 455 watt longi at 9.13 amps, using rapidtables 18.26 if I use 2p. The others would be under using the calculator. I'm sure longi and Q.cell know there stuff, but I wonder why the discrepancy.

I would call Deye but they wouldn't understand what I'm saying anyway. Lol

It looks like I have to run wire, and tell the wife no kitchen till next year. Ha

Or be very low on panel production.
 
Regarding ganging SCCs/MPPT, the output can always be ganged, up to the point where you exceed the acceptable charge rate for your batteries - and even that is a soft limit if your SCCs can talk to each other and coordinate total charge current.

It's the PV input where bad things can happen if the units aren't explicitly designed to be tied together. Unless you see a manual line stating you can tie two different MPPT inputs to one array, act under the assumption that doing so risks letting magic smoke out.
 
Regarding ganging SCCs/MPPT, the output can always be ganged, up to the point where you exceed the acceptable charge rate for your batteries - and even that is a soft limit if your SCCs can talk to each other and coordinate total charge current.

It's the PV input where bad things can happen if the units aren't explicitly designed to be tied together. Unless you see a manual line stating you can tie two different MPPT inputs to one array, act under the assumption that doing so risks letting magic smoke out.
I'm not going there with splitting PV input through 2 conductors, into 2 mppt. I like my smoke at the camp fire, while sipping on a cold beer.
1-2 conductor into 1 mppt till next year. Now the problem to resolve is getting the max I can while keeping volts down somewhat do to cold.
 
RE MPPT range vs PVmax, that probably means string should be kept under 425V, but the components are rated to handle overvoltage to 500V. It might stop producing power over 425V, but it won't risk damage until after 500V.

I think between 425Voc and 500Voc, it will still turn on and begin operating. But if the array's Vmp is > 425V, it will sit at 425V and not go higher.
If your Vmp is above 425V, probably your Voc is going to exceed 500V in cold weather, unless your location doesn't get cold.

Some inverters, Midnight Classic, have "Hyper Voc" where they tolerate a certain amount above max voltage spec but don't turn on until the panel warms up in the sun and drops below max spec.
 
Last edited:
I think between 425Voc and 500Voc, it will still turn on and begin operating. But if the array's Vmp is > 425V, it will sit at 425V and not go higher.
If your Vmp is above 425V, probably your Voc is going to exceed 500V in cold weather, unless your location doesn't get cold.

Some inverters, Midnight Classic, have "Hyper Voc" where they tolerate a certain amount above max voltage spec but don't turn on until the panel warms up in the sun and drops below max spec.
Based on the 425 Voc above if I keep the volts between 300, and 350, max in cold weather, and between 250 and 300 in warm weather, without exceeding the watts per mppt, will I be running reasonably efficient?

I'm going to try to ask Deye a few questions.

With 4 awg 2 conductor in the ground, and only 1 of the 2 mppt in the inverter at my disposal I either have to exceed the amps, or start looking for lower watt, lower amp panels, or use a max of 7 panels, at 11+- amps. So far volts is the easy number to stay in line with.

This -40 and lower weather here does present a few interesting challenges, vs you lucky people who live in fair weather. When I calculate at -50c vs +20c 6s 455 watt longi, there is a whopping 60v difference. The others in the 400w + range are close, and none below 11a Isc except q.cell which is just under.

My thanks again for the help and obvious patience a lot of you are giving me.
 
Based on the 425 Voc above if I keep the volts between 300, and 350, max in cold weather, and between 250 and 300 in warm weather, without exceeding the watts per mppt, will I be running reasonably efficient?

Does inverter documentation quote efficiency vs. PV Vmp? I've always used as high a voltage as I could, but for some SCC specs show higher efficiency when PV voltage is closer to battery (or AC voltage, if inverter.) For a long run, higher voltage reduces IR loss. It also saves on copper; silicon costs less than copper these days.

With 4 awg 2 conductor in the ground, and only 1 of the 2 mppt in the inverter at my disposal I either have to exceed the amps, or start looking for lower watt, lower amp panels, or use a max of 7 panels, at 11+- amps. So far volts is the easy number to stay in line with.

Can't deliver as much power using just one MPPT, so plan to pull wires to use both. If you bury conduit, you can add wires later.
I was going to say "Copper expensive, PVC cheap" but I discovered 2" PVC conduit is $3/foot, so almost as much as two lengths of 6 awg.
Not sure 6 awg is needed though. What is Vmp and Isc of array? What is the distance?

Oh, 4 awg? Even more expensive and harder to pull.
I would try to use 12 awg, maybe 10 awg. If PV array has two orientations, I'd like to parallel two string of different orientations and run current through one pair of wires, better utilization of copper that way.

It is perfectly acceptable to have higher percentage voltage drop in PV wires. Unlike AC wires or battery cables where voltage drop needs to be limited to just a few percent, even if PV voltage loses 10% or 25% at times of peak current, it is only money, watts, etc. Just obey ampacity limits, wire ampacity must be at least 1.56x Isc. And, have enough voltage headroom to maintain minimum MPPT voltage on a hot day.

This -40 and lower weather here does present a few interesting challenges, vs you lucky people who live in fair weather. When I calculate at -50c vs +20c 6s 455 watt longi, there is a whopping 60v difference. The others in the 400w + range are close, and none below 11a Isc except q.cell which is just under.

You do have to design array so max Voc at your record cold doesn't kill inverter.
 
Locally, 2-2-4 Triplex is $1.05' and does not require conduit. Double it for $2.05'. Cheap run from the SCC to the panels.
 
Does inverter documentation quote efficiency vs. PV Vmp? I've always used as high a voltage as I could, but for some SCC specs show higher efficiency when PV voltage is closer to battery (or AC voltage, if inverter.) For a long run, higher voltage reduces IR loss. It also saves on copper; silicon costs less than copper these days.



Can't deliver as much power using just one MPPT, so plan to pull wires to use both. If you bury conduit, you can add wires later.
I was going to say "Copper expensive, PVC cheap" but I discovered 2" PVC conduit is $3/foot, so almost as much as two lengths of 6 awg.
Not sure 6 awg is needed though. What is Vmp and Isc of array? What is the distance?

Oh, 4 awg? Even more expensive and harder to pull.
I would try to use 12 awg, maybe 10 awg. If PV array has two orientations, I'd like to parallel two string of different orientations and run current through one pair of wires, better utilization of copper that way.

It is perfectly acceptable to have higher percentage voltage drop in PV wires. Unlike AC wires or battery cables where voltage drop needs to be limited to just a few percent, even if PV voltage loses 10% or 25% at times of peak current, it is only money, watts, etc. Just obey ampacity limits, wire ampacity must be at least 1.56x Isc. And, have enough voltage headroom to maintain minimum MPPT voltage on a hot day.



You do have to design array so max Voc at your record cold doesn't kill inverter.
The length of run is 250'+-. I know I'm going to need to run more wire. Unfortunately right now the money doesn't allow it, with the other construction we have to do. The property is part time use right now, then full time in about 2 years when we sell the house. For now I need to deal with the 2 4awg conductors in the ground and 1 mppt. As many have said DO NOT use a busbar and split the power into both Mppt.
 
Where is locally?
$1.05 would be at Menards in the lower MidSouth. Currently, copper is cheapest at Lowes Home Improvement; $1.51 for Cu 4awg thhn/thwn (x2) is $3.02'. $.65 for Cu 6awg thhn/thwn (x2) is $1.30'. Using one of the online voltage drop calcs yields the 2awg Al and 4awg Cu at nearly the same resistance over 250'. I would assume, possibly wrongly, that Al triplex might also be cost effective up there, plus it doesn't require a run of conduit @$20 per 10' of 1.25".
 
$1.05 would be at Menards in the lower MidSouth. Currently, copper is cheapest at Lowes Home Improvement; $1.51 for Cu 4awg thhn/thwn (x2) is $3.02'. $.65 for Cu 6awg thhn/thwn (x2) is $1.30'. Using one of the online voltage drop calcs yields the 2awg Al and 4awg Cu at nearly the same resistance over 250'. I would assume, possibly wrongly, that Al triplex might also be cost effective up there, plus it doesn't require a run of conduit @$20 per 10' of 1.25".
Pricier up here in Canada.
 
Pricier up here in Canada.
Is the Al triplex more affordable than thhn/thwn up there? If you have to bury more conduit, direct buriel Al triplex may still be a viable economical solution. I plan on using two runs of the 2-2-4 triplex from scc to pv array, and din rail mounted terminals at both ends to go from the 2awg triplex to the smaller 4awg input of the scc, and 6awg input of the combiner boxes.

Just throwing that out there.
 
$1.05 would be at Menards in the lower MidSouth. Currently, copper is cheapest at Lowes Home Improvement; $1.51 for Cu 4awg thhn/thwn (x2) is $3.02'. $.65 for Cu 6awg thhn/thwn (x2) is $1.30'. Using one of the online voltage drop calcs yields the 2awg Al and 4awg Cu at nearly the same resistance over 250'. I would assume, possibly wrongly, that Al triplex might also be cost effective up there, plus it doesn't require a run of conduit @$20 per 10' of 1.25".
Hmm, you get a better price than we do in WY.

Menards 224 AL.PNG
 
I haven't checked lately, but add our exchange rate then 30-50% higher on top is the norm
 
Myself. I don't trust aluminum. A buddy had a house that was done in it years ago. For a short period of time they were wiring with it. He had a bad outcome. I'm probably paranoid.
 
Back
Top