diy solar

diy solar

Microinverter Voltage Rise Design Issue (Enphase users beware!)

svetz

Works in theory! Practice? That's something else
Joined
Sep 20, 2019
Messages
7,201
Location
Key Largo
Have the same microinverters randomly turning off for 5 minutes every so often? It so, it might be a Voltage Rise design issue in your setup. This thread explains the problem and some quick fixes.

Symptoms
The same microinverters turn off randomly multiple times per day.

General Causes
Microinverters are set to anti-islanding, so if the voltage/phase detection goes outside of some fairly tight bounds they shut down for 5 minutes. The usual causes are:
  • Grid issues - when this occurs it's usually more than one, and typically not the same ones
  • A bad microinverter - As a microinverter starts to fail it might not fail all at once, it may start producing harmonics or be slightly out phase that can negatively impact other microinverters.
  • Voltage Rise - Typically occurs with the same inverters at the end of a cable run and is caused by resistance greater than 2% voltage drop.

Voltage Rise
Wires have resistance causing Voltage Drop. All grid-tied inverters
increase voltage to export power. Typically they only need to raise
the voltage above the grid and any wire resistance. Enphase calls
this voltage rise, or Vrise.

The total voltage rise shouldn't exceed 2% of the grid voltage,
otherwise, it can trip the microinverter.

To calculate Vrise is the same as any voltage drop, it's the length
of all the interconnecting cables, the current, and wire gauge.
1646748297514.png

It's not clear to me, but if your grid voltage is normally high, it
might need to be even less than 2%.

The table to the right is Enphase's 240V quide for IQ7+s.
For other Enphase microinverters or more details, this link
might help. For non-Enphase microinverters, I'm not sure how
you'd measure the voltage rise.

Here's an example of 3 runs of eleven IQ7+ microinverters:
1646748780403.png


1646749039761.png
From the table, the maximum current for 11 IQ7+s in a branch circuit is 13.3 amps. From that a voltage drop calculator shows that .26% is the voltage drop from 20' or 6 AWG (~40 amps), .22% is the voltage drop from the longest run of 20' of 10 AWG, and .86 is from the Enphase table for 11 microinverters. The total of 1.34% is less than 2%, so it's a good design.

Quick Fixes
Increasing the gauge of the homerun wiring reduces the voltage drop, another solution is to use a center-fed configuration:

1646749641724.png
Here, the maximum resistance from the Q-Cables is .12% from the table and the design works. Note that without the center-fed design 1.25% + .26% + .86% exceeds a 2% voltage rise.
 
Last edited:
In my sketchier days, I had a GT-inverter feeding a sub panel fed with a 50A circuit on #6's a bit over 300ft from the main panel. With a 10kw array, I by far exceeded the 20% rule, and was actually closer to 100%, because who cared if the 50 amp panel had 100A, the wire was rated at 50, all should be kosher and we shouldn't have voltage drop feeding from the main panel, because the power in reality should be coming from the solar, right? wrong.....

This was my first experience with voltage rise.

The wire sizing of the long run, that was not compensated for the losses, experienced significant difference in voltage. In this sub panel, one day I measured over 130v L2-N, and since loads were present on L1, it was normal at 120v...But still, it was very interesting learning about this phenomenon. Everything always seemed to run fine and the voltage drop across such a long run never caused much issue for what I was doing.
 
Larger gauge wire would reduce the voltage rise/drop.
A redundant wire connecting to far end of Enphase microinverter string would reduce it. Yeah, a loop - so what?

A "buck-boost" transformer can be used to add or subtract a few volts, e.g. 120V - 6V or 240V - 12V. If utility consistently feeds you voltage on the high side, and your inverters boost voltage a few more volts (causing themselves to trip offline), buck-boost could reduce the line voltage they see.
This would behave similarly to (improperly) altering the grid voltage settings in the inverter, reducing min and max voltage 5%.

 
Last edited:
This was my first experience with voltage rise.
I never really thought about it before stumbling into it. I knew current sources would up the voltage so figured who cares about voltage drop?
But, it makes sense microinverters have a tolerance that they'll shutdown if exceeded.

Given the random 5 minute outages of the same microinverters I see folks on the Enphase forums complaining about, this makes a whole lot more sense. I still have a few that do that now and then, but as it's not the same ones repeatedly. I suspect mine are just power issues (either grid or internal noise) while they happen to be sensing.
 
Could anyone elaborate on the grid-side causes for voltage spikes and microinverter drop-out?

What kinds of grid events could cause a voltage spike? It only needs to be for 50ms to trip, so it could be noise.

I've heard that service meters with Advanced Metering Infrastructure (AMI) remote monitoring can be a cause of grid instability. Has anyone heard of this?

Other than reducing voltage with a buck device, are there other simple ways of suppressing voltage spikes enough to keep microinverters happy? I'm thinking something like ferrite cores or a filter that could remove transient spikes that could be caused by interference on the grid lines (which might have nothing to do with the joule-heating voltage rise caused by wire resistance).
 
Could anyone elaborate on the grid-side causes for voltage spikes and microinverter drop-out?

What kinds of grid events could cause a voltage spike? It only needs to be for 50ms to trip, so it could be noise.

I've heard that service meters with Advanced Metering Infrastructure (AMI) remote monitoring can be a cause of grid instability. Has anyone heard of this?

Other than reducing voltage with a buck device, are there other simple ways of suppressing voltage spikes enough to keep microinverters happy? I'm thinking something like ferrite cores or a filter that could remove transient spikes that could be caused by interference on the grid lines (which might have nothing to do with the joule-heating voltage rise caused by wire resistance).
So you put an oscilloscope on the line and saw the spike? Taking a swag(sweet wild ass guess) on this one....Could it be a motor starting up like the refrigerator/AC unit with a bad capacitor or could be a bad connection on maybe on the neutral...
 
Back
Top