In 8 months I’ve never seen more than 73 volts on three 200 watt panels even in -15F weather here in Colorado. I am running 3 200w panels that actually seem to over perform in regards to OC Voltage. I want to throw another in series which would put me right up there on voltage. Has anyone got this close in regards to voltage?
It would be very surprising if smoke comes out when slightly over 100volts. I would install the fourth panel and enjoy the extra power.
1 Like
You do not say what voltage your battery is, if it is 12V or 24V consider wiring 4 panels as 2 parallel strings of 2 panels in series.
I run 2x HiE-S390UF 49.3V OC in series on a 24V system 100/30. I have never seen over 90V on the MPPT. I have a long cable run and live in the tropics so the panels are hot. Vmp is around 40V per panel which means it will sit around 80V. Has been running for 2 years and produced just over 3000kwh. The average based on time of the year is 3-5kwh a day.
1 Like
Your panel temp will help you stay under the limit, but just to nitpick … the cable run will only drop voltage as current rises, and as soon as current rises , the panels will come down the curve from Voc towards Vmp.
I think (for both Owen and Rottieguy) your only risks are;
- a controller that happens to actually be close to Max Voc = 100v (I’d be very surprised if the spec’d max Voc is the actual max Voc, i’d expect at least 3v-10v headroom)
- a cold day
- with cloud effect to bring you closer to STC or even over STC (1000w/m^2)
- a system that has just been turned on (i.e in most cases your array will start the day way under max Voc, and be pulled down to Vmp giving you more headroom, so you would have to break that and go back to where the array is connected but not yet being drawn down)
But … remember that Rottieguy and Owen are dealing with real world measurements. In the theoretical, your nominal 200w panel might be 5% over that, which will increase the Voc by only a tiny tiny amount, but some amount. So if you got unlucky with a SCC that happened to have minimum Voc headroom, and an array that happened to be at the top end of the 5% range, and weather that happened to be exceptional such as a light snow (max reflection, and therefore also cold) followed by a clear sky with a few clouds (and therefore cloud edge effect) and having to reboot your SCC, you could be lining up the swiss-cheese model and see an issue.
This is a great justification for using the VRM - you are able to pick up those maximums over very long periods of time and therefore also covering a great range of conditions
1 Like
There’s no headroom on Victron solar controllers, the max spec is a genuine hard limit.
2 Likes
That’s interesting.
I ran a BlueSolar 75/10 or 75/15 (can’t remember now) on a 2s1p of 375w (Voc at STC would be 82.4v) to charge a battery and left it running for about 15 min before suddenly realizing that the Voc might be an issue. Tested at the terminals of the SCC, it would probably have been under 75v at Vmp, but at Voc was about 78v, so i disconnected it immediately and changed the array to 1s2p. That controller still runs fine, about 5 years later. The SCC was probably complaining but it was a bluesolar, I was in the field, and didn’t have a VE.Direct to bluetooth dongle, so i just reconfigured the array and plugged the battery back in, hoped for the best, and it charged fine. It was a good lesson i learned and it stuck with me because at the time i had a terrible feeling i had destroyed the SCC, it was the only one we had at the time, we were in the bush (forest) a long way from any other resources, and a replacement SCC would have been days away.
I wasn’t trying to say that Voc is a soft limit like current is, just that with component variability the Victron MPPT boards would not be designed/made with exactly 75v being their design limit, because that would produce boards that in around 50% of cases had a Voc limit less than 75v. Even if they used automotive or mil-spec components vs COTS spec, there will be a range.
So lets say the design Voc limit is 78v with a std. dev. of 1v, then ~99.7% of the boards produced would have an actual Voc limit of between 75v and 81v (and critically for warranty and reliability issues, almost no boards would have a Voc limit less than 75v.
So in this scenario, you would have to be a recipient of one of the 3 boards in every 1000 produced to get one with an actual Voc limit less than 75v (and 994 boards would be between 75v and 81v, and 3 would have a Voc limit above 81v) AND you would have to have an array that can hit 75v, for there to be a problem.
It also means that you have a ~97% chance that your Voc limit is 76v or above (> μ -2σ)
All this presumes that the distribution of the actual Voc limit is a normal distribution - maybe its not. But even if the MPPT had an accurate fuse link that blew at exactly 75v, this component would also have a distribution of its true value vs its nominal 75v value.
So just like with CPUs, where some are possible to overclock and some just don’t work when overclocked, there must be some headroom between design limit and actual limit.
2 Likes
The primary consideration in this matter is that when providing a five-year warranty, it is prudent to exercise caution, (such as incorporating a firewall if you will). All equipment of this nature typically has a tolerance (lets say 5 to 10%). However, it is advisable not to publicize this tolerance, as it may lead people to believe that exceeding the specified limits, such as allowing an additional 5 volts, is acceptable. Such practices could result in the equipment operating at the edge of its capabilities (component’s limit), potentially leading to premature failure or suboptimal performance.
Furthermore, it is well understood that various environments conditions (or circumstances) can impact electronic devices in diverse ways.
Therefore, sticking to a strict limit is likely to give longevity and reliability of the product.
I do have a lot of experience in this, so I fully understand the reasoning behind this.
3 Likes
12V system, 8 100ah batteries in a star parallel. 
So to make everything and the math clear, I’m running 3 Renogy RSP200D-US panels that truly seem to over perform. I’ve seen a max of 72.9V over a period of 8 months via the VRM and in my opinion, that is what matters. Of course I could split it up and go 2 x 2, but that would incur more loss over the 75’ of 10 AWG wire I’m using. Adding one more panel in series would be using my present system to it’s maximum.
The cable run is long. Much cheaper for me to use panels in series with the higher voltage and 2mm cable for 10amps vs 4mm cable for 20amps. I figured it was worth the risk and try the 100 if I blew it up use a 150. So far no problem.
I’m using 10 AWG high strand cable… It’s much more efficient to use higher voltage vs. current/amperage.
Yeah agree, worth the risk using the higher voltage.
1 Like
Technically not possible as alll used components have tolerances.
Its not wise to exceed the specs, that’s sure.
Ah, but does OC voltage, which the controller in theory will never see, make a difference? Or if all batteries are 100% charged, does the controller have to deal with a temporary high voltage?
Exactly, when the lpf’s are full, cold ore shut down due to cell over/undervoltage the pv will be in may voltage / no load.
@Ludo, I never saw the voltage exceed 72 V via the VRM metrics even when the bank of 8, 100ah batteries were fully charged and the inverter shut down. That voltage was even a rare occurrence right around New Years Day. I’m going to add the extra panel and report back in 3 months and then 6 months. Of course if I have issues, I’ll report sooner.
Ok, after running through all my recorded files, highest voltage I’ve seen is as stated. Prepping the next panel as I speak… It’s dark here so I’m not going to break my neck putting it on the roof tonight. Once it’s installed, I’ll report back on the peak voltage. So that would be 4, 200W Renogy panels in series rated @ 23 per panel. However, as I’ve said before, they have over performed and I’ve seen 634 watts when conditions were perfect on three of these panels. I believe I’ll be running the max amount of panels for the 100/50 Victron controller, but I don’t believe I’ll exceed the 100V maximum. I do plan on adding another 4 panels, but that would be with another Victron controller with the output in parallel. I’ll let you all know how it works out in the near future.
Maybe on a sunny midwinter day ?
1 Like