The "Meg" Build

[CarlCruzin]

I used it so I could bypass the inverter and out the grid directly to the RV. I stopped using it because I didn't see the need for me. It worked good. I actually have it for sale on the forum.


Yeah my pump was on an 80a breaker, and only draws around 60a, but the spike was enough to pull voltage down below 6v.

I wonder if this would be a good solution..... https://www.safiery.com/scotty-1.5kw-12-48v-canbus-bi-directional-dc-to-dc

Quite expensive, but it also does double duty by allowing the alternator to charge a 48v battery pack.

 

[Lt.Dan]

I wonder if this would be a good solution..... https://www.safiery.com/scotty-1.5kw-12-48v-canbus-bi-directional-dc-to-dc

Quite expensive, but it also does double duty by allowing the alternator to charge a 48v battery pack.

That is a very neat tool. And 96% efficient! If it was a little cheaper I would really consider it.

 

[CarlCruzin]

That is a very neat tool. And 96% efficient! If it was a little cheaper I would really consider it.

I am finalizing the wiring design on my system before installing. I wonder if the 1200w converter doesn't work, maybe I can run two in parallel?

Edit: I had a light bulb go off, If the converter isn't enough, maybe I can run a switch to the engine's starter battery.

 

[Lt.Dan]

I am finalizing the wiring design on my system before installing. I wonder if the 1200w converter doesn't work, maybe I can run two in parallel?

Might have to check with the manufacturer first. Some are a big no-no.

 

[Lt.Dan]

I think I have another clue to my RV giving earth voltage. I noticed today that inverter 1 had way more power coming in than going out (like a super bad efficiency). I just checked and I had an 1800w load, but 2700 watts were coming out of the batteries! If I turned off the PV disconnect, then the current would drop to about 1950w, and the inverter got a good bit quieter. With the PV switch on, it has a high pitch wine.

Looks like tomorrow i am tracing more wiring and see whats going on....

EDIT: I sure hope its not internal to the AIO. This is the same unit that I already replaced the main board on. I guess I can test by hooking up the roof panels and flipping the PV switch on.

EDIT2: Confirmed not the AIO. I hooked up the roof panels and flipped on the switch, no current draw, and no noise. Plug in the container panels, and immediate current draw and noise.

 

[CamoGreg]

Isolated to the container array. Narrowing it down.

I thought about your dilemma a couple days ago when a firefighter friend of mine texted me this:

Just ran a call for a 13 year old cardiac arrest. He was walking home from 7-Eleven with his 11 year old brother doing kid things in the fresh snow and rain. Apparently he started shaking and dropped to the ground next to a sign that was lit by a ground light. The ambulance crew said they got shocked when they picked him up, so possible electrocution. Initial rhythm was V-tach, one shock, and asystole after that. Dead.

Power company found no stray power in the area, but it's possible a breaker tripped before they arrived.

 

[12VoltInstalls]

firefighter friend of mine texted me

Well that sucks.
Powerco looking for stray voltage: I wouldn’t trust that as conclusive imho

 

[Lt.Dan]

Found my problem. Poor seal on some connectors caused some burn through. I never caught it because it was wrapped in split loom and wasn't showing through until today. I have since corrected this, and replaced it properly.

I took this opportunity to dig a shallow trench and bury the wires in a much shorter run, instead of 250ft, its now about 170ft. I have confirmed no voltage potential between the RV and the Earth now. Totally shock free! That's what I get for not doing a good enough job with my connections and not inspecting my wires good enough when trying to diagnose my problems.

Now that there is no voltage potential between the rv and earth, I check the solar panels and the shipping container to earth, and I found 90VAC and 38~VDC between the ground mount and the earth. The ground mount is lifted off the ground with concrete risers and wood, so im unsure how there is voltage potential there, but I ran out of daylight and will check tomorrow.

 

[12VoltInstalls]

found 90VAC and 38~VDC between the ground mount and the earth

I’m glad you found the half-wave leak but it’s so anticlimactic after all that.

Soooo have you determined how and corrected the fault that made the camper ‘hot?’ I’m assuming there’s something wrong? The camper shouldn’t be in a circuit- fault or not.

 

[Lt.Dan]

I’m glad you found the half-wave leak but it’s so anticlimactic after all that.

Soooo have you determined how and corrected the fault that made the camper ‘hot?’ I’m assuming there’s something wrong? The camper shouldn’t be in a circuit- fault or not.

It was suggested by someone (I think @Hedges) that the MPPT was backfeeding the PV lines with AC power, which 100% explains where this extra power was coming from (or going to, rather):


I mean almost 1200 watts going no where was pretty substantial.

I think the MPPT was backfeeding the PV circuit, which was energizing the earth, and completing circuit with the RV.

I hope this is the reason, because it makes half sense to me in my head lol

 

[Hedges]

"Today at 7:21 PM" Fault and solution found

Now that there is no voltage potential between the rv and earth, I check the solar panels and the shipping container to earth, and I found 90VAC and 38~VDC between the ground mount and the earth. The ground mount is lifted off the ground with concrete risers and wood, so im unsure how there is voltage potential there, but I ran out of daylight and will check tomorrow.

So you still have a problem (at the container/array), but it no longer affects your dog trying to climb RV steps?

Then I ordered 24x 250watt Snail Trail Panels from San Tan solar and 12x 240watt Sunpower panels from them as well. Which those showed up today.
View attachment 43178

The 12x 240watt panels I will be mounting on the roof of the trailer, and the 24x 250 watt Snail Trails I will be mounting on a metal building that we park next to on the property. These will most likely be getting mounted/wired up this coming weekend, or maybe next, we'll see.

Could those SanTan/Trina panels have leakage to frame? Trina had a problem, 2011 ... 2014 by one reference.
(Without electrocuting yourself), maybe you could isolate one panel frame, and see if it has potential.
HiPot with 1 mA trip is the normal way to test isolation, but with a live PV string you can accomplish the same with lethal current behind the voltage.
Maybe a smaller setup, just a couple PV panels in series, and measure voltage from each end to each frame.

 

[12VoltInstalls]

So you still have a problem (at the container/array), but it no longer affects your dog trying to climb RV steps?

That was how I read it.

I’m glad he found the cable “leak” even though the fascination level was disappointing with such a banal solution. But finding the ‘other’ problem still intrigues me to know its source.

 

[Lt.Dan]

"Today at 7:21 PM" Fault and solution found


So you still have a problem (at the container/array), but it no longer affects your dog trying to climb RV steps?

Yes, you are correct, but the problem at the container was unknown until today. I had no idea it was doing the same thing, so it might have been doing it all along, it might have just started. No more zapping when going in the trailer though. Granted I have left the trailer plugged in this whole time to prevent any shocking up until now.

Could those SanTan/Trina panels have leakage to frame? Trina had a problem, 2011 ... 2014 by one reference.
(Without electrocuting yourself), maybe you could isolate one panel frame, and see if it has potential.
HiPot with 1 mA trip is the normal way to test isolation, but with a live PV string you can accomplish the same with lethal current behind the voltage.
Maybe a smaller setup, just a couple PV panels in series, and measure voltage from each end to each frame.

I believe that is my next step, voltage leaking to the frame. I just did a quick search for a Hipot test, and that is very cool!

I could easily break apart each individual panel and check for voltage potential between the Pos lead and the frame though, correct? Or is the frame always attached to the Negative lead?

Thanks guys!

 

[Hedges]

Frame is (supposed to be) floating from the leads. A fault could be leakage from any trace or PV cell to frame.
First check voltage. If zero you can check resistance (will be near infinite.)
Perhaps nothing interesting happens except at higher voltage. Perhaps only with some degree of humidity and/or dirt.

HiPot is how devices are tested to ensure chassis has kV isolation from power wires. It takes a large airgap or shorter distance through intact insulator to hold off the voltage, so damaged insulation can be observed.

Without unbolting frames, you could just shut off & disconnect the circuit, measure for voltage between MC ends and frame. Divide and conquer.

We still don't know what is driving AC onto the PV wires.

Normally we would want inverter system to have its chassis grounded to a ground rod, a wire connecting chassis to PV racking, maybe a ground rod there too.
Since your PV rack has DC & AC superimposed but RV does not, I guess there is still no ground between them. If there was, might drive RV to voltage, or might pull rack to ground.
There shouldn't be much of any current driving those things. Few mA maybe, from things like EMI filter capacitors. But if neutral bonded to ground and ground terminal wired to ground rod, shouldn't affect anything.

 

[12VoltInstalls]

We still don't know what is driving AC onto the PV wires.

I think I saw his lot/clearing on a trailer ad for an alien crash landing coverup documentary? The antigravity drive is buried at the site….

 

[Lt.Dan]

Frame is (supposed to be) floating from the leads. A fault could be leakage from any trace or PV cell to frame.
First check voltage. If zero you can check resistance (will be near infinite.)
Perhaps nothing interesting happens except at higher voltage. Perhaps only with some degree of humidity and/or dirt.

I will break the panels apart and check for voltage potential and resistance between positive and frame. If nothing, I will begin to wire them back into series one at a time until I begin to see voltage potential or resistance.

HiPot is how devices are tested to ensure chassis has kV isolation from power wires. It takes a large airgap or shorter distance through intact insulator to hold off the voltage, so damaged insulation can be observed.

Without unbolting frames, you could just shut off & disconnect the circuit, measure for voltage between MC ends and frame. Divide and conquer.

This is my plan. When I said break apart, I meant disconnect the wiring.

We still don't know what is driving AC onto the PV wires.

Normally we would want inverter system to have its chassis grounded to a ground rod, a wire connecting chassis to PV racking, maybe a ground rod there too.
Since your PV rack has DC & AC superimposed but RV does not, I guess there is still no ground between them. If there was, might drive RV to voltage, or might pull rack to ground.

I still have not gotten a ground wire between the PV rack and the RV, it is on my to do list.

I think I saw his lot/clearing on a trailer ad for an alien crash landing coverup documentary? The antigravity drive is buried at the site….

I thought I hit something hard when I was digging....

 

[Hedges]

Are you familiar with "Binary Search"?
Instead of an order(N) problem to locate a single shorting panel, reduces it to an order(sqrt(N))

Measure whole string for voltage to frame.
(note that due to capacitance, open circuit will give a reading but decay back to zero).

If faulty, split in middle and check each half.
Divide and conquer.

The theorems for fastest possible search algorithm assume binary test in each case.
By performing other operations (e.g. math on the spelling of a search word), it can be done faster. That is where "hash tables" come in.

Assuming a single fault with leakage from PV panel to frame, the voltage you measure from each end will indicate how far in the string to where fault is. (could be multiple faults, though.)

 

[Lt.Dan]

Pretty sure I have another BMS failure. I'm getting a little tired of working on this thing now, and cheating out on a few things bit me in the ass. Debating on getting rid of the BMS and just running an active balancer like @DanF And set my inverter cutoffs conservative. Any reason not to? The simple reasons are taken care of...

 

[cinergi]

Pretty sure I have another BMS failure. I'm getting a little tired of working on this thing now, and cheating out on a few things bit me in the ass. Debating on getting rid of the BMS and just running an active balancer like @DanF And set my inverter cutoffs conservative. Any reason not to? The simple reasons are taken care of...

I wouldn't. If something gets out of balance, no active balancer will be able to handle the voltage differential quickly enough to keep a runaway cell from going too high. I forget your system voltage but if it's 48v and you're charging to just 54.4 (typical float - not really high enough), you could have 15 cells at 3.38 but one at 3.7.
What happens if the active balancer fails (do you get notified)?

 

[Sanwizard]

Are you familiar with "Binary Search"?
Instead of an order(N) problem to locate a single shorting panel, reduces it to an order(sqrt(N))

Measure whole string for voltage to frame.
(note that due to capacitance, open circuit will give a reading but decay back to zero).

If faulty, split in middle and check each half.
Divide and conquer.

The theorems for fastest possible search algorithm assume binary test in each case.
By performing other operations (e.g. math on the spelling of a search word), it can be done faster. That is where "hash tables" come in.

Assuming a single fault with leakage from PV panel to frame, the voltage you measure from each end will indicate how far in the string to where fault is. (could be multiple faults, though.)

I think you just talked about finding faults without using a "time domain reflectometer"... those bring back memories from my technical days....