PIP inverter repairs and hardware modifications

[Vissie]

Oh I found shunt resistors on an old scc board will check mine

Will it help to reload the firmware? Does it also update the scc firmware?

No it didn't help
I replaced the scc and the other one shows solar charge amps just fine
Will sit with the faulty one and compare to another working one to see if I can pick up any faulty components

[coulomb]

DC/DC IGBT Q28 blown out.

I note that when one transistor blows, its partner in the "totem pole" often is at least damaged, often also blown. So check Q27 carefully.

As always, check the gate driver components: R91, D32-D35, etc.

Please suggest equelent part number.
current part number is : FGH75T65SHDT

Best is to order the exact same part, if possible. From Mouser India: https://www.mouser.in/ProductDetail/ON- ... PEWw%3D%3D

Probably, most 650 V 75 A TO-247 IGBTs would work, but it's a slight gamble.

Q27 Q28 schematic.png (65.35 KiB) Viewed 424 times

I note that the above has two R91s; I'll alert the author to get it corrected.

[Vissie]

I need to know how it measures charge current from solar.

The Solar Charge Controller (SCC) has its own 8-bit microcontroller with its own firmware. It sends the PV current to the main DSP in the inverter via an internal serial connection. Once a second, the SCC sends a QGS command to the DSP. The first three fields of this command are the PV voltage, battery voltage as measured by the SCC, and SCC output current. Note: the current value sent and displayed by the inverter are battery-side amps, NOT PV amps. So if your panels are at 100 V and your battery is at 50 V, the sent value will be twice the PV current.

How the SCC microcontroller actually measures the current, I don't know. It's likely a current shunt and operational amplifier.

You are so clever Coulomb!
Ok I found R2(12k) resistor on top not measuring correctly. Traced it to U11 (LM358 op amp) to bottom of board. Replaced it and everything is working shananan!
Case closed...

[ Edited Coulomb: fixed quoting. ]

[friendnb2]

Hallo guys, I would ask you for some help with my inverter.I have EASUN ISolar SM5kW48V1PF 2 years so far, lead acid 300Ah battery pack,amorphous panels with 80V 2000W at peak power. All has been working well up to last cca beginning of November, when, during sunless days in Slovakia (middle of Europe), inverter going to be crazy. Switched off PV input for a seconds and then svitch it back with abnormal input power level (shows display, but really not) again for a seconds. And this two statuses repeats again and again until the dark or until disconnect PV breaker. ( I have short videos but because large amount of MB do not put it here) .
Display shows like this:

20210131_151451.jpg (1.14 MiB) Viewed 387 times

Then like this:

20210131_151445.jpg (826.46 KiB) Viewed 387 times

Really, from PV flows no current and to battery as well (have own measurement). And voltage from PV is ok, like this:

20210131_151614.jpg (1.03 MiB) Viewed 387 times

or this:

20210131_151617.jpg (1.12 MiB) Viewed 387 times

Have you any opinion what can be reason for this abnormal behavior? thanks..

[coulomb]

All has been working well up to last cca beginning of November, when, during sunless days in Slovakia (middle of Europe), inverter going... crazy.

A quick search says that Slovakia has less than 50° latitude, so I assume that you get at least some sun in winter (unlike countries with midnight sun and midday darkness). My guess is that with only 2000W of panels and poor sun, there isn't quite enough power to run the Solar Charge Controller processor, and junk values are the result.

Other readers have reported strange things happening at twilight. Do you perhaps have extended twilight there in winter? I've never experienced it.

The - 1A is normal, that's self consumption, no load, and no PV charging current. When the SCC starts up, there is about 15 seconds of doing nothing, then about 15s of low charge current (about 1A), then about another 15s of ramping up to full charge current. That's if it's not running crazy, of course. No need to rush things, it seems

Edit: another possibility is that your SCC power supply is marginal or faulty.

[friendnb2]

There are a few things to force me thinking something is wrong.First of all, my inverter is used two years and never works like this before. It worked from a few watts from early mornig up to 2kW if direct sun (if), doesnt matter if is summer or winter, let say wake up me when relay connected panels at 6AM and the power from PV slowly growing up, works whole day then panels are disconnected at late evening.
Next, for example today is sunny winter day, at 10AM, I have over 80V from panels (its near to maximum), display shows this

20210201_093330.jpg (1024 KiB) Viewed 370 times

but as we can see, let say "picture of panel" is not full like when panel connected with full power. It simply not works properly.
I assume measurement of power from panels should be wrong or SCC not works properly or what else..

[friendnb2]

I finally have red older posts an saw that other members had the same issues, anyway im not find detailed what was wrong. It looks that it can be problem of SCC board.

[friendnb2]

Hi boys, has anybody idea where can I buy SCC modul?

[coulomb]

I thought I'd post an update on what I've been doing to try and support a forum reader that needed to replace the DSP chip on his inverter's control board. The problem is that a new totally blank DSP chip needs a bootloader. A bootloader image was obtained many months ago, and I've packaged it in such a way that a new DSP chip can be reprogrammed.

This is how I run it on the bench. The 6-pin 2.54 mm spacing header socket (the orange thing) is from an electronic organ I pulled apart years ago.

Control card powered on bench sm.jpg (276.29 KiB) Viewed 157 times

It needs 5 V at several hundred milliamps, and 12 V at less than 1 mA. The 12 V is only needed to keep the reset chip happy; without 12 V applied, the DSP stays in reset. I use a bench dual power supply.

For a JTAG programmer, I use the below, which TI used to sell as a "docking station". It would take a DSP processor on a small daughter board that plugged into the DIMM socket; the DSP on the daughter board was one of those ball grid array ones with 1.2 squllion pins, which is not very experimenter friendly. So the daughter board allowed for experimenting. They put an XDS100 V1 JTAG programmer on the board so you can program it and debug it; it's basically that FTDI chip you can see, and a few buffers and power supply chips. A 14-way ribbon cable connects the JTAG board to the inverter control board. I ended up with one of these; they cost about $150 ten years ago. You can probably buy something similar these days. in case you're curious, a schematic diagram for the docking station / JTAG programmer-debugger is here (first one I found; I don't know what the crossing out is about).

JTAG programmer with FTDI chip crop.jpg (383.74 KiB) Viewed 157 times

It has a USB-B port (bottom of photo); the other end (USB-A) plugs into a Windows computer. I use TI's Code Composer Studio, which is free if you use an XDS100 class JTAG debugger. I was able to demonstrate programming a bootstrap loader into this control board. Alas, the forum reader that I did all this for has so far been unable to replicate this. [ Edit: He's succeeded now. ] But others may want to try, or at least take comfort from the fact that it can be done.

[ Edit 9/Feb/2021: The forum reader succeeded in programming the bootloader. He was using a "programming board" that somehow wasn't set up correctly. When he programmed directly to the inverter's control board, as I did, he succeeded. ]

[moerph]

Hello,


It's my first post although I'm reading this forum for quite some time. I think about buying a new inverter and Pip's are still part of my considerations. I didn't find any information if the "original" Voltronic units differ from MppSolar and copycats units in case of quality ? Does anyone know if Voltronics uses better components in their "retail" builds ? I guess not but sometimes ODM do. If they don't I probably will buy a cheaper one as they have to be upgraded anyway.

Thanks
moerph

[coulomb]

I didn't find any information if the "original" Voltronic units differ from MppSolar and copycats units in case of quality ?

MPPSolar inverter-chargers genuine Voltronic Power manufactured; they were possibly the first reseller. In this post, I admit that I thought that MPPSolar were actually the manufacturers.

Quality of clones varies. Obviously, it's not possible to say what the quality of all of them is. Some are so bad that you can't update the firmware at all. Some seem to have evolved beyond the original design, offering more features than the genuine product (e.g. exporting to the grid from battery and/or solar). Generally, I would not touch any clone. There are components in there that if replaced by the most common (cheapest) part, the life will be short (wild guess 1-2 years), and the failure mode will generally be a loud bank with burning epoxy smell. For example, Voltronic supply (or they did; I haven't checked lately) 2000 h 105°C capacitors to protect the MOSFETs; Weber and I recommend 10 000 h and 105°C parts. A clone maker might use an ordinary part, which is generally rated at 85°C and 1000 h (8 times less life than the Voltronic supplied part, and 40 times less than what we recommend).

Does anyone know if Voltronics uses better components in their "retail" builds ?

I don't know what you mean by "retail builds" in this context. Voltronic Power are strictly Original Design Manufacturers; you can't buy a Voltronic Power branded inverter-charger. It has to be rebranded by a reseller, who then are responsible for customer support. Only when absolutely necessary does a problem escalate to Voltronic, and then only via the reseller. Voltronic don't seem to want to deal with end users at all.

As far as I know, all Voltronic Power genuine inverter-chargers are the same quality, and are identical internally (for one particular model, obviously; there are dozens of models). So a PIP-5048MS from MPPSolar will be the same as an Axpert MKS from Mecer (give or take details like 64 V versus 58.4 V, which occasionally are not distinguished by the reseller's model number). There may be an exception with Inverex and perhaps one or two other resellers. They seem to get slightly different models, often with slightly higher specifications (e.g. 5.2 kW when the main model os 5.0 kW). I have not seen inside one of these, so I don't know if the quality is the same or not. My guess is that the parts would be identical, apart from a few critical components needed to achieve the slightly higher specifications.

If they don't I probably will buy a cheaper one as they have to be upgraded anyway.

My suggestion is to try to avoid a clone, which is not always easy, especially when attempting to buy at the lowest possible price.

[ Edit: added "from MPPSolar" ]

[moerph]

Thanks for the detailed reply.

So i guess if i consider buying a PIP i'd go with the MPP so far. As far as i noticed, no matter if it's a MPP or a rebranded unit they all seem to suffer from the same bugs and errors. Still checking on other options for a 5K inverter that's worth it's price, somehow reliable or at least somewhat easy to repair. If i choose a Pip I have to upgrade it in adavance for sure as this unit will go to country where there will be no service or warranty at all.

Thanks
moerph

[coulomb]

So i guess if i consider buying a PIP i'd go with the MPP so far.

The only problem with MPPSolar is (or was, I haven't checked lately) is that they dont sell a 58.4V model when an equivalent 64V model exists. I'm not a bug fan of the 64V models, though I realise that sometimes they are needed.