diy solar

diy solar

Short Circuiting error with MPP 8048 MAX and SE5000 Autotransformer...

I can't figure out why bonding neutral center tap of transformer to ground would cause overload fault in inverter, unless one of the inverter Line outputs was grounded.
I don't easily find the manual for it to review; they have so many models. I only see a few details in the listing.


Have you checked that all wires have correct voltage relative to each other and to ground?
With a 230V isolated inverter, you should be able to ground either line, or the center tap of a transformer.
With a source that already has one end grounded, like 230V utility line & neutral, an auto-transformer producing 115V center tap couldn't have the center-tap grounded. In that case, an autotransformer would need and extra winding and tap, for +230V, +115V, Neutral, -115V

I found this post because I have a similar situation. I have the MPP 8048MAX inverter and a 5000VA torrid center tap transformer, which I believe is electrically similar to the SolarEdge SE5000 Autotransformer.

Hedges is almost correct in his guess that one of the Line inputs is grounded . . .

ibkickinit says
"50-Amp 2-pole breaker for L1/L2 coming from the inverter into the panel, . . ."
this is incorrect. The MPP 8048MAX is not a USA inverter. If you look at the inverter and manual on page 6 (8th page in the PDF) you will see the terminals for both AC line in and AC line out are labeled L, N, and ground. The neutral is internally connected to ground. So, with the ground bonded in the circuit breaker box the inverter would send 230 VAC to ground and the transformer center tap would send 115 VAC to ground via the neutral bond. Thus the F05 error from the 4048MAX.

The neutral on the AC input is not connected to ground on the MPP 8048MAX, but the neutral on the AC output is. This unit is designed for Europe and other parts of the world that use 220 VAC at the outlet and then what is called the second line output in the USA is a neutral as there is no 110VAC nor line 2, just line and neutral which are normally wired with brown and blue wires respectively (one can see this internally on the 8048MAX.

So measuring continuity across the terminals with the 8048MAX off, you will see that the neutral and the ground are bounded together and have continuity. With the unit on, from the terminals labeled, L to N 230VAC, L to ground 230VAC, Neutral to Ground 0 VAC.

However this still leaves me with the question of how to properly ground this setup. Currently I have the ground wire from the inverter not connected anywhere.

I believe that ibkickinit has his inverter and transformer grounds connected to the ground bus in the circuit box but his neutrals are not bonded to ground as he has the bond disconnected, which I believe leaves them grounded.

Any ideas on how to correctly wire this with proper grounding would be appreciated.
 
I would assume this would be the same as deal as the 240V Growatt with an auto transformer in that the internal bond must be removed so the neutral from the transformer and ground can be bonded.

I have been looking at this one over the spf5000 or second choice of an 8000t LF. I know watts247 sells the Mpp 8048. Did you get it there? Curious if he sells them set up for the US.
 
@Clifton I purchased the MPP 8048 MAX on ebay from the seller maximumsolar_us.

Not sure if watts247 sells them setup for the US. If you get one or contact them about it and find that they have a solution, like unbonding the neutral and ground in the unit, I would be greatful for any information.
 
You can remove the ground N bonding screw on the PCB as shown in this video....
To magically make it into a US "version" and call it by a US suffix. (Just like the 5000ES.
After your remove the bonding screw you can rename it the 8048max-US , who know one little screw had so much power?)

The 8048max has 2 x 500V mppts (total.120A)

 
Last edited:
You can remove the ground N bonding screw on the PCB as shown in this video....
To magically make it into a US "version" and call it by a US suffix. (Just like the 5000ES.
After your remove the bonding screw you can rename it the 8048max-US , who know one little screw had so much power?)

Suggest you determine if removing the screw actually provides sufficient isolation, won't inadvertently make contact or arc across due to power line transients. I think proper design would have several mm gap or suitable insulator.

Knowing what I do about PCB designs, I wouldn't assume that simply removing a screw provided isolation, or that ohm meter confirms it. If it was a single-sided PCB with traces on top side, it might do the trick (though probably not with a gap meeting "clearance" requirements.) But if a double-sided or multi-layer PCB, very likely not.
 
Will open one up 8048max remove the mainboard and video it for both the 8048max and 5 units we have waiting for spares of the 5000ES (non "US") and wait to see if the others claiming US are any different.
Till Monday evening.., should be interesting.
 
@Clifton
Signature Solar says that Growatt is not bonding the neutral to ground on the Growatt inverters they sell in the US. Not putting the bonding screw in and filling the hole with silicon.
 
You can remove the ground N bonding screw on the PCB as shown in this video....
To magically make it into a US "version" and call it by a US suffix. (Just like the 5000ES.
After your remove the bonding screw you can rename it the 8048max-US , who know one little screw had so much power?)

The 8048max has 2 x 500V mppts (total.120A)

Thanks Ian Roux! I plan on unbonding the neutral in my MPP 8048 MAX using this information. I am fully off-grid with my setup, so that eliminates a lot of the issues.
 
Thanks Ian Roux! I plan on unbonding the neutral in my MPP 8048 MAX using this information. I am fully off-grid with my setup, so that eliminates a lot of the issues.

You can remove the ground N bonding screw on the PCB as shown in this video....
To magically make it into a US "version" and call it by a US suffix. (Just like the 5000ES.

Suggest you determine if removing the screw actually provides sufficient isolation, won't inadvertently make contact or arc across due to power line transients. I think proper design would have several mm gap or suitable insulator.

Knowing what I do about PCB designs, I wouldn't assume that simply removing a screw provided isolation, or that ohm meter confirms it. If it was a single-sided PCB with traces on top side, it might do the trick (though probably not with a gap meeting "clearance" requirements.) But if a double-sided or multi-layer PCB, very likely not.

Like I said, I think it is quite likely that simply removing the screw does not separate "N" (grid L2) from Ground by sufficient distance. I think design requirements are normally around 3.2mm for 120Vrms. Even if it measured open with an ohm-meter, it could arc across with normal transient voltages that occur on AC lines.

I think that insertion of a suitable insulator between PCB and standoff is necessary. Or, maybe standoff can be removed.
 
You might want/need to slip an insulating material between the PCB's removed screw and the chassis to which the screw attaches to.

With some inverters or controllers, this has to be done when hipoting the unit so that the MOVs do not conduct during that test (dielectric strength test)

You probably aren't going to do that anyway but thought I'd mention it.

boB
 
Like I said, I think it is quite likely that simply removing the screw does not separate "N" (grid L2) from Ground by sufficient distance. I think design requirements are normally around 3.2mm for 120Vrms. Even if it measured open with an ohm-meter, it could arc across with normal transient voltages that occur on AC lines.

I think that insertion of a suitable insulator between PCB and standoff is necessary. Or, maybe standoff can be removed.
I will check the gap. We are dealing with 220 VAC across the gap. Is the for 220 VAC different?
The transients that you are taking about are these from the grid or generated in a normal household?
Any thoughts on using silicon as an isulator in the hole as Growatts is doing for the units sent to Signature Solar.
 
You might want/need to slip an insulating material between the PCB's removed screw and the chassis to which the screw attaches to.

With some inverters or controllers, this has to be done when hipoting the unit so that the MOVs do not conduct during that test (dielectric strength test)

You probably aren't going to do that anyway but thought I'd mention it.

boB
I haven't heard the term hipot for years when I was doing AM transmitter maintenance. Don't have the equipment to do that on the unit.
 
@Clifton I purchased the MPP 8048 MAX on ebay from the seller maximumsolar_us.

Not sure if watts247 sells them setup for the US. If you get one or contact them about it and find that they have a solution, like unbonding the neutral and ground in the unit, I would be greatful for any information.
I don't understand why not buying a dedicated splitphase model like LV6548 or the LVX6048 ?
 
I don't understand why not buying a dedicated splitphase model like LV6548 or the LVX6048 ?
The LV6548 is only 120v. The 6048 is 2000 watts less than the 8048MAX. The 8048 uses 70w. I think the LVX6048 is around 140w.
 
I will check the gap. We are dealing with 220 VAC across the gap. Is the for 220 VAC different?
The transients that you are taking about are these from the grid or generated in a normal household?
Any thoughts on using silicon as an isulator in the hole as Growatts is doing for the units sent to Signature Solar.

220 VAC? Are you sure?
In European market, that leg would be "N", and I think is selectively grounded by a relay, or connected to N from the grid.

In American market, inverter is changed from 220 V 50 Hz to 240V 60 Hz. You (or a vendor) are then trying to isolate that leg and call it "L2".
I think you want it to be 120V, but prior to establishing a midpoint (e.g. by grounding centertap of an auto-transformer), I suppose L1 and L2 could each take on anything between zero and 240V.

When on the grid, certain transients come from various events (not even including lightning, just loads and switching.
I don't have a schematic to go by, but I get the impression this inverter has a relay it closes at about the time it disconnects what it knows as "L1" and "N" from the grid. The relay it closes would then tie the off-grid "N" to ground.
If my understand is correct, that means line voltages and transients are held off by the relay. Only when relay closes are inverter voltages held off from chassis by gap where screw was removed.

Possibly not as severe a transient event. I'd still recommend insulation, creepage, and clearance appropriate for both 120 and 240V operating, and for transients to about 1250V (what HyPot testing might be done in factory.)

Don't call it "silicon", that is an element, and a semiconductor. You might use silicone as an insulator. But it is soft, make sure standoff wouldn't pierce through. Simply squirting it into the hole wouldn't guarantee an insulator between PCB and standoff, which were nominally in contact.

Any other (harder) material with suitable electric ratings. I'm imagining removing the board and sticking a piece of plastic over the standoff, or removing the standoff, or adhering a piece of plastic to the PCB (maybe with a squirt of silicone). Thickness to withstand mechanical damage and withstand 1250V (look up the material.) Wide enough that electricity would have to crawl > 3mm across surface to get between PCB and standoff.
 
Will open one up 8048max remove the mainboard and video it for both the 8048max and 5 units we have waiting for spares of the 5000ES (non "US") and wait to see if the others claiming US are any different.
Till Monday evening.., should be interesting.
I see you posted a video of the board from the Growatt 5000ES on Youtube. Any chance you will post a video showing the MPP Solar PIP8048MAX mainboard too? Thanks!
 
220 VAC? Are you sure?
In European market, that leg would be "N", and I think is selectively grounded by a relay, or connected to N from the grid.

In American market, inverter is changed from 220 V 50 Hz to 240V 60 Hz. You (or a vendor) are then trying to isolate that leg and call it "L2".
I think you want it to be 120V, but prior to establishing a midpoint (e.g. by grounding centertap of an auto-transformer), I suppose L1 and L2 could each take on anything between zero and 240V.

When on the grid, certain transients come from various events (not even including lightning, just loads and switching.
I don't have a schematic to go by, but I get the impression this inverter has a relay it closes at about the time it disconnects what it knows as "L1" and "N" from the grid. The relay it closes would then tie the off-grid "N" to ground.
If my understand is correct, that means line voltages and transients are held off by the relay. Only when relay closes are inverter voltages held off from chassis by gap where screw was removed.

Possibly not as severe a transient event. I'd still recommend insulation, creepage, and clearance appropriate for both 120 and 240V operating, and for transients to about 1250V (what HyPot testing might be done in factory.)

Don't call it "silicon", that is an element, and a semiconductor. You might use silicone as an insulator. But it is soft, make sure standoff wouldn't pierce through. Simply squirting it into the hole wouldn't guarantee an insulator between PCB and standoff, which were nominally in contact.

Any other (harder) material with suitable electric ratings. I'm imagining removing the board and sticking a piece of plastic over the standoff, or removing the standoff, or adhering a piece of plastic to the PCB (maybe with a squirt of silicone). Thickness to withstand mechanical damage and withstand 1250V (look up the material.) Wide enough that electricity would have to crawl > 3mm across surface to get between PCB and standoff.
The inverter output voltage can be set to 220, 230 or 240 VAC. 230 is the default. 50 or 60 Hz is also a setting.

On the AC output side the neutral is bounded to ground by a screw even when the unit is off see the video that MPP Solar USA - Ian Roux posted on January 29.

I checked the unit with a continuity meter, with it off and no wires connected to the AC input or output terminals. On the input side the neutral and ground were not connected and on the output side the neutral and ground are bounded together as the unit is shipped.
 
Last edited:
On the AC output side the neutral is bounded to ground by a screw even when the unit is off see the video that MPP Solar USA - Ian Roux posted on January 29.

I check the unit with a continuity meter with it off and no wires connected to the AC input or output terminals. On the input side the netural and ground were not connected and on the output side the neutral and ground are bounded together as the unit is shipped.

It may differ between models, but I saw a relay near the screw in still photos. Link below says at least some models use a relay to do the bonding.
This could be a NC relay which bonds N-G offgrid but isolates it when on-grid. That would be desired if grid connection already has neutral bonded to ground; otherwise, neutral currents would split at flow partially in ground wire.



Since you observed neutral on output is bonded to ground with power off but input is isolated, a HyPot test on input would see relay gap (I'm not sure what voltage that can hold off). Testing output would of course see the short. With bonding screw removed, testing output would apply voltage across whatever gap there might be between PCB conductor and case; that is where we think it is likely there is very little gap, not a reliable open circuit, so HyPot is appropriate to prove it is well isolated.
 
Even in his video, Ian say's he does not like removing the screw.
However he is supposedly going to be removing the PCB and seeing if removing the screw is a valid method of "Americanazing" the inverter.
When I opened up a chinese ripoff of an MPP unit, there was a wire link that was part of the grounding/capacitor area we are talking about.
 
Back
Top