PIP inverter repairs and hardware modifications

non-EV Solar, Wind and other renewable power sources
non-EV batteries and other energy storage stuff
Forum rules
Important!
This forum is for discussion of Non-EV matters.
Alex_50174
Noobie
Posts: 4
Joined: Fri, 13 Jul 2018, 18:36
Location: Novokuibyshevsk, RU

Re: PIP inverter repairs and hardware modifications

Post by Alex_50174 »

Hi, coulomb.
coulomb wrote: Fri, 13 Jul 2018, 21:56 100 A is quite light for two inverters in parallel. It also sounds like they were only AC rated; you need DC rated fuses or breakers there. When they opened at something over their rated current, they may well have arced, pitting the contacts, and that may have interrupted the power to the inverters many times in a fraction of a second. The natural inductance of the battery cables may have caused voltage transients.
Yes, I know that C100 AC breaker is not a good idea for that invertors and I already said to the electrician all I think about he. It should be noted that I only deal with the repair of inverters and they got to me already in a faulty state. I already repaired a month ago the MPPT charger part of one of this inverters when the electrician connected about 7000Wt solar panels to one inverter...
coulomb wrote: Fri, 13 Jul 2018, 21:56 If you did not replace the original MOSFETs and capacitors, then they were already running very close to their limits. A few rapid inductive spikes could well have sent them into avalanche breakdown.
There are original 75V IRFB3307Z. It is interesting that their appearance is normal, but the leftmost and rightmost transistors are with burnt legs because of the high current. So, it's was a my guess that the voltage raised above 75V for a fraction of a second.
coulomb wrote: Fri, 13 Jul 2018, 21:56 However, it's also possible that heat caused the capacitors that protect the battery-side MOSFETs to dry and become high impedance, allowing the transients on the 50 V bus to kill those MOSFETs, and the transformer coupled the transients (plus new ones from the short on the 50 V bus) to the DC-DC IGBTs.
I think that it's not a reason. Invertors worked only 1 or 2 month, the capacitors usually do not dry so quickly. Full capacitance of 4 capacitors with unsoldered MOSFETs is 13300 (but I haven't ESR meter so I can not be sure).
Also, my inverters are bit different in some parts from described - capacitors are 3300uF 80V "success" (I guess that it is Chinese noname).
DC-DC and buck IGBTs are FGW50N60HD, inverter bridge IGBTs are FGW75N60HD.
coulomb wrote: Fri, 13 Jul 2018, 21:56 This is all very speculative; it's hard to know what happens in a breakdown situation unless everything is instrumented and logged in a laboratory situation.
Yes, of course. But users of this invertors said that they not exclude short circuit at the AC output. But I think that if it really was, it couldn't be a reason, it could cause increased load and tripping the battery breaker.
coulomb wrote: Fri, 13 Jul 2018, 21:56 Edit: you have also likely blown several gate driver parts. See the index (first post this topic) for what to look out for, should you attempt to repair the inverters.
Of course, I will check and repair gate drivers.

Thank you!
User avatar
coulomb
Site Admin
Posts: 6357
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: PIP inverter repairs and hardware modifications

Post by coulomb »

Alex_50174 wrote: Sat, 14 Jul 2018, 00:20 It should be noted that I only deal with the repair of inverters and they got to me already in a faulty state.
Good to hear that there are other repairers out there. There is such a throw away mentality these days, perhaps more so in the West.
I already repaired a month ago the MPPT charger part of one of this inverters when the electrician connected about 7000Wt solar panels to one inverter...
I would have thought that too many panels would not be such a serious problem. But I suppose the control system has to operate faster with higher power panels (in good light conditions). Good to know, thanks.
There are original 75V IRFB3307Z. It is interesting that their appearance is normal, but the leftmost and rightmost transistors are with burnt legs because of the high current.
Interesting. They seem to be using the original worst-spec MOSFET, except for an added "Z" in the main part of the part number. These "Z" parts have the same or slightly improved specs in most areas, except for pulsed drain current, where it is improved from 120 A (wire bond limited) to 510 A. My guess is that the "Z" versions use some sort of welding technique to connect the package leads to the die, and the non-"Z" versions use traditional wire bonding. So in the non-"Z" versions, you effectively have internal 120 A fuses, and in the "Z" versions, these effective fuses are absent. So the TO-220 leads become the fuse.
coulomb wrote: Fri, 13 Jul 2018, 21:56 However, it's also possible that heat caused the capacitors that protect the battery-side MOSFETs to dry ...
I think that it's not a reason. Invertors worked only 1 or 2 month, the capacitors usually do not dry so quickly.
Fair enough, but see next.
Also, my inverters are bit different in some parts from described - capacitors are 3300uF 80V "success" (I guess that it is Chinese noname).
Oh. Considering that the original Jamicons were a special order, it seems that someone somewhere has used inferior parts, with inferior ripple current ratings. Is it possible that these are actually clones, not genuine Voltronic Power manufactured?
But users of this invertors said that they not exclude short circuit at the AC output. But I think that if it really was, it couldn't be a reason, it could cause increased load and tripping the battery breaker.
I would guess that shorting the output would affect the 230 V IGBTs more than the other IGBTs and MOSFETs.

[ Edit: added "to connect the package leads to the die". ]
[ Edit: added "with inferior ripple current ratings" .]
MG ZS EV 2021 April 2021. Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
Patching PIP-4048/5048 inverter-chargers.
If you appreciate my work, you can buy me a coffee.
Alex_50174
Noobie
Posts: 4
Joined: Fri, 13 Jul 2018, 18:36
Location: Novokuibyshevsk, RU

Re: PIP inverter repairs and hardware modifications

Post by Alex_50174 »

coulomb wrote: Sat, 14 Jul 2018, 08:20 Good to hear that there are other repairers out there. There is such a throw away mentality these days, perhaps more so in the West.
I think that it is all over the World tendention.
coulomb wrote: Sat, 14 Jul 2018, 08:20 I would have thought that too many panels would not be such a serious problem. But I suppose the control system has to operate faster with higher power panels (in good light conditions). Good to know, thanks.
Maybe it was a manufacturing defect. It burned after first turning on. I have changed a logic chip before gate driver PWM сhip to stop burning transistors.
coulomb wrote: Sat, 14 Jul 2018, 08:20 Interesting. They seem to be using the original worst-spec MOSFET, except for an added "Z" in the main part of the part number. These "Z" parts have the same or slightly improved specs in most areas, except for pulsed drain current, where it is improved from 120 A (wire bond limited) to 510 A.
As I can see in service manual from first page, there were IRFB3307Z in original inverters too (in older versions). And pulsed drain current for Z - 512A, for non-Z - 510A (from datasheets), so differences are not so big.
coulomb wrote: Sat, 14 Jul 2018, 08:20 Oh. Considering that the original Jamicons were a special order, it seems that someone somewhere has used inferior parts, with inferior ripple current ratings. Is it possible that these are actually clones, not genuine Voltronic Power manufactured?
Maybe there are a clones. I can see only model name "Inverter 5kVa-48VDC" and there is no a firm name. Inverters were bought at the Aliexpress.
I'll try to measure ESR with PC's audio card. If it will be not big that it will mean good quality of capacitors. If ripple current not enough, ESR will be big I think.
Alex_50174
Noobie
Posts: 4
Joined: Fri, 13 Jul 2018, 18:36
Location: Novokuibyshevsk, RU

Re: PIP inverter repairs and hardware modifications

Post by Alex_50174 »

Going further with burned inverters. Maybe it will be interesting for someone.
1) Capacitors "success". Tried to measure ESR by audio card at 16MHz and it's about 10-20mOhm. I think this capacitors are good but I can't find a place in the internet where to buy them.
2) Tried to measure optocouplers U1/U2/U3/U4/U12 (PIN8 TO PIN5: 2k PIN7 TO PIN5: 2k in the service manual) - all the values on orders of magnitude greater. It is normal or will be better to change all of them, and, maybe, inv IGBTs?
3) Strange symptom - in the one of inverters the compound holding the toroidal inductor has cracked - maybe there was a heavy vibration.

PS: 1 of 8 capacitors (from 2 inverters) has no capacitance at all. So, quality is not good :)
User avatar
coulomb
Site Admin
Posts: 6357
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: PIP inverter repairs and hardware modifications

Post by coulomb »

Kimo-23 wrote: Wed, 15 Aug 2018, 01:58 i scrolled through the display settings, the battery voltage say 119v :-(
My apologies. It seems that battery voltages up to 999 V can be displayed. My only suggestion is that the battery sense resistors or some other part of the battery voltage sense circuit have been damaged, presumably during the SCC replacement. Possibly one of these 1 MΩ resistors:

Possible battery sense resistors.jpg
Possible battery sense resistors.jpg (133.55 KiB) Viewed 9239 times
One megohm is a high value resistor, so some sort of slightly conductive gunk around one or more of them could change the battery reading dramatically. Higher conductivity of these high-side sense resistors will lead to a higher than actual battery voltage reading. They will be under some form of coating, which would normally protect them, so this may have been scratched or otherwise damaged. I'm not 100% certain that these are the battery sense resistors, but the in-line formation and being so close to the battery fuse (marked 200 in the photo; yours may look quite different) suggests that they are.

The rest of the battery sense circuitry might be found by following the traces under the white silk screen cover (this seems to be an additional protection, over and above the usual green solder resist, to protect sensitive PCB traces). But the rest of the circuit could possibly be as simple as a single resistor and capacitor to battery negative.
[ Edit: however, with the two strings of resistors visible, this suggests an operational amplifier to respond to the difference between two voltage dividers. ]
[ Edit 2: So my suggestion would be to measure the value of those eight resistors; in-circuit should be OK. ]
[ Edit by Weber: See the op-amp schematic here. Scroll down past the index. ]
MG ZS EV 2021 April 2021. Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
Patching PIP-4048/5048 inverter-chargers.
If you appreciate my work, you can buy me a coffee.
User avatar
Vissie
Groupie
Posts: 50
Joined: Sun, 02 Sep 2018, 07:24
Location: South Africa

Re: PIP inverter repairs and hardware modifications

Post by Vissie »

Inverter/charger blowing Dc Dc converter and buck IGBTs
Good day
My name is Vissie and I am new here
I read through most of the posts and think some of you guys would be able to help me with this one
Model: SOL-I-AX-5M
This is a new type. Manufactured Nov 2017
They use a new type of IGBT so they claim a power factor of 1
Rated power 5000VA/5000W and it does not have the heat sink on the backside

The IGBTS on the dc-dc converter and the only one on buck converter were blown and blows again after replacing.
We used this inverter as a demo on a show and there was a problem with the grid supply and the inverter blew up after a short while
BUCK MOS Q31
IGBT Q27/Q28/Q29/Q30
One of the small caps (C142 to C145) 101j between gate and C of one of the IGBTs was also faulty
I tried to scope the signal from U9 Driver IC UC3525 by removing the 2 small drive transformers TX10 and 11 but by doing so I get no pulses on the primary side of transformer. I think the control circuit does not allow the ic to generate signals or the Ic is faulty. Everything on the Ic measures ok according to the service manual. All the driver resistors and transistors between IC and pulse transformers measures ok
How can I check the pulses? Should I remove all the IGBTs and then check on the gates after switch on?
I want to try and build it together without the casing on the workbench to enable me to measure the bus voltage and see if all is ok there
This unit also has a complete new board for solar charger that fits on top of the main board
I found the post from Coulomb on how to check the drive signal on the battery side mosfets and want to know if there is any method to test the signal on the dc dc side
Any help will be apreciated
Attachments
pulse transformers
pulse transformers
IMG_2478.JPG (109.53 KiB) Viewed 8927 times
U9
U9
IMG_2479.JPG (101.44 KiB) Viewed 8927 times
User avatar
coulomb
Site Admin
Posts: 6357
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: PIP inverter repairs and hardware modifications

Post by coulomb »

Vissie wrote: Wed, 05 Sep 2018, 22:33 My name is Vissie and I am new here
Hi, Vissie, and welcome.
This is a new type. Manufactured Nov 2017
They use a new type of IGBT so they claim a power factor of 1
Rated power 5000VA/5000W
I note that upgrading from power factor 0.8 to 1.0 does not affect the inverter IGBTs. Are you saying that the high-side full bridge now uses higher power IGBTs? I guess that's part of how they get the extra power.
and it does not have the heat sink on the backside
Do you mean it doesn't have the black heatsink on the top of the unit, for the SCC? The top heatsink has been gone for years.
We used this inverter as a demo on a show and there was a problem with the grid supply and the inverter blew up after a short while
Not a great demonstration, then :(
One of the small caps (C142 to C145) 101j between gate and C of one of the IGBTs was also faulty
Do you mean between emitter and collector of one of the IGBTs? Collector to gate sounds unusual, but I may have traced the schematic incorrectly.
I tried to scope the signal from U9 Driver IC UC3525 by removing the 2 small drive transformers TX10 and 11 but by doing so I get no pulses on the primary side of transformer. I think the control circuit does not allow the ic to generate signals or the Ic is faulty.

I presume you've enabled the chip by shorting the output of opto U19, as described in a post.
This unit also has a complete new board for solar charger that fits on top of the main board
Is it significantly different to the one pictured in this post?
I found the post from Coulomb on how to check the drive signal on the battery side mosfets and want to know if there is any method to test the signal on the dc dc side
At least on the PF0.8 models, and I assume it would be the same on the PF1.0 models, U9 drives both the battery-side and the high-side (bus-side) full bridge gates. So the same technique we described for the battery-side MOSFETs applies to the high-side IGBTs.

As a point of interest, are there more than 16 MOSFETs on the battery side in the PF1.0 model you have? If there are still 16 of them, are they higher current models?
Last edited by rhills on Sat, 08 Sep 2018, 10:28, edited 1 time in total.
MG ZS EV 2021 April 2021. Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
Patching PIP-4048/5048 inverter-chargers.
If you appreciate my work, you can buy me a coffee.
User avatar
Vissie
Groupie
Posts: 50
Joined: Sun, 02 Sep 2018, 07:24
Location: South Africa

Re: PIP inverter repairs and hardware modifications

Post by Vissie »

coulomb wrote: Thu, 06 Sep 2018, 07:01
Vissie wrote: Wed, 05 Sep 2018, 22:33 My name is Vissie and I am new here
Hi, Vissie, and welcome.

Thank you
This is a new type. Manufactured Nov 2017
They use a new type of IGBT so they claim a power factor of 1
Rated power 5000VA/5000W
I note that upgrading from power factor 0.8 to 1.0 does not affect the inverter IGBTs. Are you saying that the high-side full bridge now uses higher power IGBTs? I guess that's part of how they get the extra power.

They use a new IGBT STGW80H65DFB. https://www.st.com/resource/en/datashee ... h65dfb.pdf
These devices are IGBTs developed using an advanced proprietary trench gate field-stop structure

and it does not have the heat sink on the backside
Do you mean it doesn't have the black heatsink on the top of the unit, for the SCC? The top heatsink has been gone for years.

There is no external heatsink on the back as in older units
We used this inverter as a demo on a show and there was a problem with the grid supply and the inverter blew up after a short while
Not a great demonstration, then :(

Luckily we had other repaired ones to use
One of the small caps (C142 to C145) 101j between gate and C of one of the IGBTs was also faulty
Do you mean between emitter and collector of one of the IGBTs? Collector to gate sounds unusual, but I may have traced the schematic incorrectly.

Correction. Yes cap between C and E
I tried to scope the signal from U9 Driver IC UC3525 by removing the 2 small drive transformers TX10 and 11 but by doing so I get no pulses on the primary side of transformer. I think the control circuit does not allow the ic to generate signals or the Ic is faulty.

I presume you've enabled the chip by shorting the output of opto U19, as described in a post.

I did not but will. I thought it was only for the battery side. Im not to sure where to connect the dc voltage. Must I remove all that IGBTs before I do it?
I ordered new Driver IC UC3525
This unit also has a complete new board for solar charger that fits on top of the main board
Is it significantly different to the one pictured in this post?

It stretches over the whole length of the inverter and sits on top of the main board
For some reason I cannot paste a photo of it
Last edited by rhills on Sat, 08 Sep 2018, 10:41, edited 1 time in total.
User avatar
Vissie
Groupie
Posts: 50
Joined: Sun, 02 Sep 2018, 07:24
Location: South Africa

Re: PIP inverter repairs and hardware modifications

Post by Vissie »

So it looks like I have to use the main board only with no processor board or solar charger board connected and then connect an external current limited supply of at least 31V to the battery contacts of the main board. I will also remove all IGBTs on the dc-dc converter and the one on buck converter. Then I short pins 3 and 4 of opto U19 and scope the gates of the IGBTs
User avatar
Vissie
Groupie
Posts: 50
Joined: Sun, 02 Sep 2018, 07:24
Location: South Africa

Re: PIP inverter repairs and hardware modifications

Post by Vissie »

Morning
I received another inverter for repairs
This one had ghekko's inside that shorted the bus. The battery side fets and the 4 igbts of the inverter full bridge and the buck igbt were blown. It also burned a hole in the main board. Also a few drive resistors were gone.
Well I managed to scope the signal like Coulomb described and the signal looks fine. I get the gate signal on the bat side fets and on the dc dc igbt's
I do not get a pulse for the igbts on the inverter full bridge Igbts but it seems like its driven from a small ic or maybe the control board
So I am bit scared to switch it on without knowing if they get signal. I think after building all the faulty parts back I should only connect the current limited dc supply on the battery terminals and see what happens
Please can some one tell me how to post photos. I got it right the first time but not again
User avatar
coulomb
Site Admin
Posts: 6357
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: PIP inverter repairs and hardware modifications

Post by coulomb »

Vissie wrote: Tue, 11 Sep 2018, 19:27 This one had ghekko's inside that shorted the bus.
Wow. I bet the critter involved had a bad day too :o
The battery side fets and the 4 igbts of the inverter full bridge and the buck igbt were blown. It also burned a hole in the main board. Also a few drive resistors were gone.
Well I managed to scope the signal like Coulomb described and the signal looks fine. I get the gate signal on the bat side fets and on the dc dc igbt's
Ok, that's good.
I do not get a pulse for the igbts on the inverter full bridge Igbts but it seems like its driven from a small ic or maybe the control board
Yes, the main inverter IGBTs are driven by ePWM1 and ePWM2 (both A and B outputs, so 4 signals total) of the microcontroller. So that's from the control (daughter) board, probably buffered on either the control board or the main board, or possibly even both.
So I am bit scared to switch it on without knowing if they get signal. I think after building all the faulty parts back I should only connect the current limited dc supply on the battery terminals and see what happens
That's wise. Unfortunately, I don't think we've figured out how to do that as yet.

The answer would seem to be a sort of dummy daughter board, which might be little more than a piece of Veroboard with appropriate pins (maybe less than a dozen); you'd want to make sure that no gate driver inputs are left floating. Also ensure that the bus soft start circuit is disabled. It would be nice if it's difficult to plug in such that the pins are one hole away from where they should be.

The board might have a pair of 555 ICs, and maybe a switch, so you can pulse the upper IGBTs and lower IGBTs separately or together. Of course, have generous dead time built into the circuit, so the upper and lower transistors of a pair never turn on together. This would all be run with about 12 VDC current limited to 0.5A or so on the DC bus.

If you work it out, please publish the schematic here.
Please can some one tell me how to post photos. I got it right the first time but not again
Just drag an image into the edit pane. Or, since you've done it before, the main trick to remember is that the needed attachment tab is often off the page, and you need to use scroll bars to get to it.

However, we have had some users claim that sometimes the attachment tab simply refuses to appear. It's never happened to me, but I haven't been posting many images or other attachments on this forum lately. Our web master, Rob, is aware of this issue, but last I know (a week or so ago) he was at a conference and had little spare time for web administration. It's also rather baffling. It's possible that you can drop an image into the edit pane even if the attachments tab isn't visible.

There is also my New Forum Tips post. Scroll down to the "How to post an image" section.
MG ZS EV 2021 April 2021. Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
Patching PIP-4048/5048 inverter-chargers.
If you appreciate my work, you can buy me a coffee.
User avatar
Vissie
Groupie
Posts: 50
Joined: Sun, 02 Sep 2018, 07:24
Location: South Africa

Re: PIP inverter repairs and hardware modifications

Post by Vissie »

Good day
I have a question. Can I test the main board by just plugging the control board in and connect the 48V batteries and shorting the switch pins?
Or must it be screwed onto the casing for the 4 earth screws to make connections?
I repaired 2 boards and was trying to test it like this after I was sure that the signals was going to the gates like in previous posts
The lcd comes on but it only shows the battery on screen and will not switch on the inverter
The one board had a 10k resistor and one O/P diode on the switched power supply blown. The other one blew one of the small transistors that drives the small signal transformers that supply the square wave to the fet gates. All the fets and igbts were ok
I need to know as it is a pain to screw everything back every time to test.

Ok. I build it in and there is no difference. One of the control boards gives a fault code ( HS) I cannot find it in the service manual. Does anybody know what it is?
User avatar
coulomb
Site Admin
Posts: 6357
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: PIP inverter repairs and hardware modifications

Post by coulomb »

Vissie wrote: Wed, 26 Sep 2018, 20:43 Can I test the main board by just plugging the control board in and connect the 48V batteries and shorting the switch pins?
Or must it be screwed onto the casing for the 4 earth screws to make connections?
Yes, you can test on the bench; the earth connections are not required for operation. You can even run it without the control board for certain tests. Obviously you need to be very careful, and I would not run a large load while on the bench.
I repaired 2 boards and was trying to test it like this after I was sure that the signals was going to the gates like in previous posts
Well done finding those faults, even if you haven't seen them working yet.
The lcd comes on but it only shows the battery on screen and will not switch on the inverter
All I can think of is you might not have the right settings to cause the inverter to come on. Or it might have judged the battery to be "weak". Run through the settings for anything that might be stopping the inverter from coming on.
The one board had a 10k resistor and one O/P diode on the switched power supply blown. The other one blew one of the small transistors that drives the small signal transformers that supply the square wave to the fet gates. All the fets and igbts were ok
Interesting. I've seen more of the larger components (IGBTs and MOSFETs) blowing up, taking out the smaller components.
One of the control boards gives a fault code ( HS) I cannot find it in the service manual.
That's actually not a fault code; it's only a fault code if the fault LED (red) is lit, and the "ERROR" indicator is showing. The "HS" means it's a MaSter; setting 28 must be set to PAR. It's just so hard to get something that looks like an "M" into seven segments. Our patched firmware has an "M" in "MS" that doesn't look like a "H", which is hopefully less confusing. If this machine isn't supposed to be paralleled (does it have the paralleling board installed?) then setting 28 is wrong, and there may be others incorrectly set that is preventing the inverter from turning on.

Good luck; it's good to see people attempting to fix these machines instead of just throwing them away.
MG ZS EV 2021 April 2021. Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
Patching PIP-4048/5048 inverter-chargers.
If you appreciate my work, you can buy me a coffee.
User avatar
Vissie
Groupie
Posts: 50
Joined: Sun, 02 Sep 2018, 07:24
Location: South Africa

Re: PIP inverter repairs and hardware modifications

Post by Vissie »

Coulomb. Thank you for your answers
Is there a way to reset this control boards to the factory setting?
User avatar
coulomb
Site Admin
Posts: 6357
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: PIP inverter repairs and hardware modifications

Post by coulomb »

Vissie wrote: Thu, 27 Sep 2018, 15:24 Is there a way to reset this control boards to the factory setting?
Yes, there is the PF command (checksum 0x0439). The easiest way to reset to defaults is to download the free WatchPower program on a connected computer, and use one of the menus or buttons to reset all settings to defaults. I don't know the details since I don't use the program myself.
MG ZS EV 2021 April 2021. Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
Patching PIP-4048/5048 inverter-chargers.
If you appreciate my work, you can buy me a coffee.
User avatar
Vissie
Groupie
Posts: 50
Joined: Sun, 02 Sep 2018, 07:24
Location: South Africa

Re: PIP inverter repairs and hardware modifications

Post by Vissie »

Thank you. I will try it
Savage1ZA
Noobie
Posts: 3
Joined: Sun, 07 Oct 2018, 01:31
Real Name: Chris knipe

Re: PIP inverter repairs and hardware modifications

Post by Savage1ZA »

Hi Guys,

Is there any new / additional information available about the error 51? My 5KW Axpert also now out of the blue, started with this very intermittently. What's also very strange to me, is that it only occurs during VERY hot and sunny conditions (30C+ ambient). Irrespective of load, and never during the night. In fact, it's always between round about 11AM and 3PM. Just as quick as the Inverter will start giving this fault condition (on and off, intermittently), just as quick it will just magically stop and return to normal pro-longed operation.

No change to any settings, no change to any loads, no change to batteries. As randomly as it starts, just as randomly it stops. My PV is also very close to the max acceptable parameters (as someone else posted) - I too, will be disconnecting my PV tomorrow to see whether the fault / issue persists without the PV connected. Someone else said they stopped experiencing the problem after disconnecting the PV.
User avatar
coulomb
Site Admin
Posts: 6357
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: PIP inverter repairs and hardware modifications

Post by coulomb »

Savage1ZA wrote: Sun, 07 Oct 2018, 01:37 Hi Guys,
Welcome to the forum.
Is there any new / additional information available about the error 51?
You may be providing it here; see below. Error 51, for other readers, is documented in the user manuals as "Over current or surge". It seems to relate specifically to inverter current. There are separate limits for battery and line (bypass) modes; I assume that your problem is only when in battery mode.
My 5KW Axpert also now out of the blue, started with this very intermittently. What's also very strange to me, is that it only occurs during VERY hot and sunny conditions (30C+ ambient).
This suggests something is drifting with temperature, and the temperature may be elevating because the solar charge controller is working hard.
Irrespective of load, and never during the night.

If there really is no correlation with actual load, then that strongly suggests that the problem is a measurement problem, not an operating condition problem.
My PV is also very close to the max acceptable parameters (as someone else posted) - I too, will be disconnecting my PV tomorrow to see whether the fault / issue persists without the PV connected. Someone else said they stopped experiencing the problem after disconnecting the PV.
The correlation with PV has been mentioned before. I can't see how it makes any difference to the main firmware. So it may well be that the SCC (Solar Charge Controller), which is now packed in rather tightly with the main board components, is causing heating of some measurement components, e.g. the hall effect current sensor for inverter current.

What is your main firmware (U1) version number?

If it's earlier than 73.00, the it may not be working the fans hard enough. In 73.00, they increased the fan speed for a given PV power. Initially, I thought that this was a way of making the inverter cool harder during the day, and be more quiet at night when people appreciate the quiet more. But now it looks like the SCC might need more cooling than it used to, certainly more than back when the SCC heatsink used to be on top of the inverter. So updating to factory or patched firmware based on 73.00 might be a simple solution. It might also be worth noting the inverter temperature when the overload happens.

Instead of a sensor drift error, it might be that the inverter automatically derates its maximum inverter current limit as the temperature increases. I've not seen this in the firmware so far, but I haven't been looking for it, and might just not have come across it as yet.

The other possibility is that some EEPROM settings might be corrupted somehow, causing up to about 25% scale errors, and possibly offset errors as well. There is a command to reset these to factory defaults, but I would not recommend using it except as a last resort. You would lose any factory calibrations by using that command. There are commands to adjust the scale and offset eeprom calibration values; this would be better if you notice that power readings are consistently in error, and suspect that these errors are pushing your inverter to think that it's over the limit. But you say that there is no correlation with load, so this seems unlikely in your case.

[ Edit: Added sentence about losing factory calibrations if using the force all EEPROM to defaults command. ]
MG ZS EV 2021 April 2021. Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
Patching PIP-4048/5048 inverter-chargers.
If you appreciate my work, you can buy me a coffee.
User avatar
coulomb
Site Admin
Posts: 6357
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: PIP inverter repairs and hardware modifications

Post by coulomb »

Savage1ZA wrote: Sun, 07 Oct 2018, 01:37 Is there any new / additional information available about the error 51?
We have continued this discussion on the Power Forum. To summarise,
  • Error 51 is not restartable, whereas "ordinary overload" (error 7) is restartable.
  • So this is a "never exceed" situation, protecting the circuitry from catastrophic failure.
  • This error seems more common in South Africa, where mains surges seem more prevalent than elsewhere.
  • My latest suspect is the MOV (Metal Oxide Varistor), associated with the output of the inverter.
From http://www.resistorguide.com/varistor/ :
MOVs degrade when exposed to repeated surges. After each surge the MOVs clamping voltage moves a little lower, how much depends on the joule rating of the MOV in relation to the pulse. As the clamping voltage falls lower and lower, a possible failure mode is a partial or complete short circuit, when the clamping voltage falls below the protected line voltage.

It goes on to say that this situation could lead to a fire hazard. Fortunately, the PIP/Axpert firmware is monitoring the peak current, and when it exceeds a threshold (it seems that the threshold starts at 40 A instantaneous, and increases to 80 A instantaneous if all goes well), fault code 51 (error 51) is issued, saving the electronics from catastrophic failure, possibly including fire.

MOVs have a marked negative temperature coefficient. As temperature increases, their threshold voltage decreases, and their leakage current increases. Both these effects would lead to the error 51 coming up more frequently when the inverter is hot, e.g. due to high solar charge current during the hotter part of the day.

Edit: Note please also @JvdSpoel and @ejb, since my guess has changed.
MG ZS EV 2021 April 2021. Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
Patching PIP-4048/5048 inverter-chargers.
If you appreciate my work, you can buy me a coffee.
Savage1ZA
Noobie
Posts: 3
Joined: Sun, 07 Oct 2018, 01:31
Real Name: Chris knipe

Re: PIP inverter repairs and hardware modifications

Post by Savage1ZA »

@coulomb started to open up and test various components inside the inverter today as per the service manual. The entire battery side bus of the main board, is effectively toast and blown :shock: Right up to (and including the) PWM CNTL.

There is absolutely NO visible damage / burns anywhere, but testing the components resulted in the one failure after the next... Right on Axpert for the Error 51, it is indeed an indication of something rather seriously wrong with the Inverter.

Needless to say, I am in the market for a new main board, and failing that, a new Inverter.
User avatar
coulomb
Site Admin
Posts: 6357
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: PIP inverter repairs and hardware modifications

Post by coulomb »

Savage1ZA wrote: Wed, 10 Oct 2018, 18:28 coulomb, [ I ] started to open up and test various components inside the inverter today as per the service manual. The entire battery side bus of the main board, is effectively toast and blown :shock: Right up to (and including the) PWM CNTL.
? But didn't you say that it recovered and never showed the problem at night? It sounds very suspicious that every component you test is faulty, and yet there is no visible sign of damage.

It sounds like you have been misinterpreting the service manual checks. Many of those are "resistance" checks, which are only valid with one model of multimeter. I use those ones more as a guide than as gospel. The diode measurements are far more useful.

Also, I would not expect the battery side components to be faulty, though it is possible that they failed as well as the inverter components, buck transistors, and battery-side to high-side full bridge.

I still think it's worth replacing the MOVs. I'd replace them all, since I suspect they all will have seen much the same overload events. It may well be that all your semiconductors are fine, saved by the Firmware calling fault code 51.
MG ZS EV 2021 April 2021. Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
Patching PIP-4048/5048 inverter-chargers.
If you appreciate my work, you can buy me a coffee.
User avatar
coulomb
Site Admin
Posts: 6357
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: PIP inverter repairs and hardware modifications

Post by coulomb »

The MOVs seem to be Brightking brand, part number 561KN20. Specifications here; basic specs:
350 VAC, 504-616 V for 1mA, clamp at 920 V and 100 A, 280 J at 2 ms. 20 mm diameter, thermal fuse protected.

The nearest I can find in a quick search is this part from Mouser (comes up for me in AU$; use the Mouser part number 576-TMOV20RP385E in your local Mouser web page). Basic specs:
385 VAC, 558-682 V @ 1mA, clamp at 1025 V and 100 A, 300 J at 2 ms. 20 mm diameter, thermal fuse protected.

[ Edit: this part (Mouser part number 576-TMOV20RP320E) has lower voltages, and may be more suitable:
320 VAC, 459-561 V for 1mA, clamp at 840 V and 100 A, 270 J at 2 ms, 20 mm diameter, thermal fuse protected. ]

Image
MG ZS EV 2021 April 2021. Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
Patching PIP-4048/5048 inverter-chargers.
If you appreciate my work, you can buy me a coffee.
Savage1ZA
Noobie
Posts: 3
Joined: Sun, 07 Oct 2018, 01:31
Real Name: Chris knipe

Re: PIP inverter repairs and hardware modifications

Post by Savage1ZA »

coulomb wrote: Wed, 10 Oct 2018, 18:58
Savage1ZA wrote: Wed, 10 Oct 2018, 18:28 coulomb, [ I ] started to open up and test various components inside the inverter today as per the service manual. The entire battery side bus of the main board, is effectively toast and blown :shock: Right up to (and including the) PWM CNTL.
? But didn't you say that it recovered and never showed the problem at night? It sounds very suspicious that every component you test is faulty, and yet there is no visible sign of damage.
That's correct yes. It only showed the error during the day, and on warm days. It's not every single component, but it is a fair amount of them... A summary (posted extensively about this on the other forum)...

Main MOSFETs:
Resistance GS: 11.7K (fine). GD 130k (manual states 250k?), DL O/L Considering I get 130k across ALL the MOSFETs and not one, I am going to assume a reading of 130k instead of 250k is sufficient
Diode: SD 0.43V EXCEPT Q14 where I get a O/L This is definitely a faulty MOSFET.

Drivers:
Q46, Q48, Q41, Q43 Resistance BE 26M (manual states 420k), BC 26M (manual states 420k), CE 30K (correct)
Q46, Q48, Q41, Q43 Diode BE 0.6V (all fine), BC 0.6V (2 fine, 2 dead), CE (all busted)
Q15, Q16, Q41, Q43 Resistance BE 26M (manual states 420k), BC (26M (manual states 420k), CE All over (should be 30K)
Q15, Q16, Q41, Q43 Diode BE 0.6V (all fine), BC 0.6V (2 Fine, 2 busted), CE (all busted)

A few resistors also faulty

Q60/Q61 Mosfets blown

I've pretty much given up measuring at this stage... The part (component) numbers in the service manuals, matches the part (component) numbers in the inverter, so I'm fairly sure I have the correct manuals for the inverter...
Revlac
Groupie
Posts: 87
Joined: Sat, 19 Aug 2017, 19:43
Real Name: Aaron
Location: Scenic Rim

Re: PIP inverter repairs and hardware modifications

Post by Revlac »

Some time ago We had a 5048Hs inverter and a 4048HS running well on the same 300ah battery bank, the 5048 was running the normal house load and the other would run the water pumps and the air conditioner occasionally.
On a day when the 4048 was sitting completely idle and the other was running the house at about 400watts, one of the inverters failed and took out the other one at the same time, so I can confirm this happens.
They were on completely different circuits apart from the battery, I suspect a transient surge on the DC lines.
And mostly the same parts blown on each inverter.

Battery side Power Mosfet's 8 of them busted
Bus Side components IGBT's 2 of them gone
R54 R57 burnt open
D69 open
Roasted the 200amp fuse.
A few other parts and some zenor diodes on the DC/DC IGBT bus side where also open circuit and not in the checklist on the repair manual.

I have repaired the 4048HS and it has been working hard in the shed ever since and very pleased with it, It was not difficult to repair, just time consuming checking all drive components are in working order before any attempt to power up.
So it was well worth the effort to repair it and learn about its working along the way. :D
Will repair the other one later.
Now have a 5048MS unit ready to go in case something happens to the other unit.

Oh I should mention the only other thing that has caused the inverter to stop was A Taurus Hot Air Gun on the low setting, cant remember the error number, but the inverter only disconnects the output relay and within 10 seconds it reconnects and running again.
The little geny didn't like the hot air gun either.

I think the later model Main boards have a bunch of TVS Diodes (under Main board) on the battery side Mosfets?

Cheers
Aaron
4048HS Inverter 2015.
5048HS Inverter with 200AH GFB cells
5048MS Inverter with 200AH GFB cells
Mad Inverter
rezydent
Noobie
Posts: 17
Joined: Tue, 02 Jan 2018, 04:53

Re: PIP inverter repairs and hardware modifications

Post by rezydent »

weber wrote: Wed, 10 Oct 2018, 16:38 Release Version of Patched Firmware 72.20c for PF1 models with the 64V option
...
Note: This firmware is only suitable for the 48 V models with 5kW/5kVA (PF1) and the 64 V maximum battery voltage option. It is not suitable for the older models with a PF0.8 rating (4 kW/5 kVA), nor is it suitable for the newer PF1 models if they don't have the optional 64 V hardware.
I have a question. If, in my PIP4048, I would replace transistors, withstand higher voltage, and load software that would allow up to 66V, everything would work properly? if I understand correctly, I could raise the charging voltage to 66V
Which transistors should I replace?
6240 solar, 2 x PIP5048MG inverter, 20kWh battery
3120 solar, PIP5048MG inverter 6kWh battery
Post Reply