Anyone else finding their MPPT readings all over the place on cloudy days?

Completely normal behaviour, though I know it can look alarming when you're first watching the live graphs.

On overcast days your panels are essentially harvesting diffuse radiation rather than direct irradiance, and the power curve becomes much flatter and less predictable. Your MPPT algorithm is constantly hunting for the true maximum power point, and when the irradiance is shifting every few seconds — cloud edges, partial shading, reflections — it never quite settles. You end up with that characteristic spiky, erratic trace rather than the smooth bell curve you'd get on a clear day.

I've got a Victron SmartSolar 100/30 on my shepherd's hut and a separate 150/45 on the garden office, and both do exactly this. The 150/45 particularly so because that array has some morning shading from a neighbour's oak, which compounds the issue. Victron's "HEX" logging via VRM makes it painfully visible.

A few things worth checking though:

• MC4 connections — corroded or loose connectors will mimic this behaviour even on bright days
• Panel mismatch — if you've mixed panel ages or wattages in a string, the MPPT struggles more under low light
• Controller temperature — some units throttle their tracking speed when warm

If it's only happening on genuinely overcast days and your clear-day curves look sensible, I wouldn't lose sleep over it. You're probably still harvesting 70-80% of what's theoretically available from that diffuse light.

What controller are you running and how many panels in your array? That'd help narrow it down further.

@Trigger makes a solid point about diffuse vs direct radiation. Worth adding that panel temperature plays a bigger role than most realise on cloudy days — cold overcast conditions can actually push Voc above your STC-rated voltage, so if your MPPT's absorption setpoint is marginal, you'll see the controller hunting around trying to track properly.

On my Victron SmartSolar 100/30 I noticed this particularly in winter — the MPPT history in VictronConnect showed the tracking efficiency dropping to ~85% during broken cloud versus 98%+ on clear days. Completely expected behaviour.

If your readings look genuinely erratic rather than just variable, worth checking:

• Cable connections at combiner box (oxidation is nasty in damp UK weather)
• Whether partial shading is hitting any one string disproportionately
• Your controller's sweep interval settings

Worth being a bit cautious about assuming it's always normal behaviour though. I had erratic readings on my shepherd's hut setup for weeks blaming the weather before eventually finding a loose MC4 connector that was intermittently dropping out. The voltage swings looked almost identical to genuine diffuse-light harvesting.

So yes, cloudy days will give you variable output — that's physics — but if the readings seem worse than they should be even accounting for overcast conditions, don't dismiss it too quickly.

My Victron MPPT logs were genuinely useful here. Pulling the history and comparing yield on similar cloud-cover days gave me a baseline. When one day looked dramatically worse than comparable days, that pointed toward a fault rather than just weather.

Basically: normal until proven otherwise, but do rule out the mundane stuff first.

@ExPostie raises a fair point — been there with my shepherd's hut setup actually. Spent ages thinking my Victron SmartSolar was just being temperamental on grey days before I realised one of my MC4 connectors had started corroding. Readings were all over the shop.

Quick checks worth doing if it does feel worse than usual:

• Inspect your MC4 connections (moisture gets in more than you'd think)
• Check your cable runs for any chafing
• Pull up the VRM portal history if you're on Victron — patterns over a few days tell you a lot more than live graphs

Most of the time it's just diffuse light doing its thing, but a dodgy connection will absolutely mimic that behaviour and you won't notice until you're losing serious harvest.

@ExPostie and @GazAllen are onto something important there. One thing nobody's mentioned yet — partial shading hitting individual cells can cause really chaotic MPPT hunting behaviour that looks like a fault but isn't. My narrowboat setup spent weeks driving me mad with this; turned out a mooring post was clipping the corner of one panel for about 90 minutes each afternoon.

Worth logging your irradiance against output using something like VRM Portal if you're on Victron — the mismatch between what should be coming in versus what actually is will tell you pretty quickly whether you've got a genuine efficiency problem or just physics doing its thing on a grey day.

Also check your MC4 connectors haven't taken on moisture. Had a corroded pair giving intermittent resistance last winter that produced almost identical erratic readings.

@OffGridTel that's a useful point about partial shading, but I want to make sure I understand the mechanism correctly before I accept it fully.

When you say partial shading "hitting" the MPPT — are you referring to the bypass diodes activating within the panel strings, which then causes the I-V curve to develop multiple local power peaks? Because if so, I'd expect the controller to be hunting between those peaks rather than simply showing erratic voltage swings.

Is that what you're actually seeing on your graphs, or is it more of a wattage fluctuation with relatively stable voltage?

I ask because on my van setup with a single 200W Renogy panel I've noticed voltage staying fairly consistent on overcast days even when output drops sharply — which makes me wonder whether multi-panel arrays with shading behave fundamentally differently to single-panel setups.