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.
Post Reply
rinaldoparaipan
Noobie
Posts: 25
Joined: Mon, 26 Oct 2015, 03:26
Real Name: Dan Paraipan
Location: Romania
Contact:

Re: PIP inverter repairs and hardware modifications

Post by rinaldoparaipan »

And Part2
Part2.pdf
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 »

rinaldoparaipan wrote: I think is useful.

Definitely! Thanks, Rinaldo.
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
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Re: PIP inverter repairs and hardware modifications

Post by weber »

Last Wednesday I admitted defeat and made the trip out to the bush to remove the ultra-quiet Noctua fans from the PIP in Black Monolith #1 and replace them with the original fans from the manufacturer (but blowing upward).

The installation of the Noctua fans is described starting here:
http://forums.aeva.asn.au/viewtopic.php?p=54806&t=4332#p54806

The reason for restoring the original fans is that the Noctuas would occasionally trigger a "Locked fan" warning (code 01) even though they were working perfectly. Apparently this only became a problem for the system owner after we upgraded the PIP firmware some months back (to use Coulomb's LiFePO4 patch). After that, the PIP would beep loudly whenever this warning occurred and would continue doing so until the system was shut down and restarted. But a few days later it would happen again.

Beeps were supposed to be turned off by our BMS master sending commands to the PIP on system startup. It would wait 10 seconds for the PIP to be ready to receive commands, and the "beeps off" command happened to be the first one it would send. But after the firmware upgrade the PIP would take longer than 10 seconds to be ready and the "beeps off" command would be missed (nothing to do with Coulomb's patch). We now wait 17 seconds before sending commands to the PIP.

But even with the beeps off, the customer worried that other warnings might be masked by the ever-present false-alarm from the fans. I thought that surely the PIP would rotate through multiple warning codes, but Coulomb read the firmware and found that a fan-locked warning will indeed mask everything else except for an overload warning.

We hypothesised that when the PIP requested the lowest speed from the Noctuas via their control input, they might be running marginally too slowly to satisfy the PIP, via their tacho output, that they were working.

So Coulomb also did some heroic circuit tracing and analysis of the part of the firmware that detected the supposed "fan-locked" condition, but it wasn't obvious how we could make use of this information, and circuit board modifications onsite were impractical. Hence the decision to restore the original fans. Since then there has been no recurrence of the fan-locked warning.

The original reason for replacing the fans has been mitigated by the firmware upgrade. The fans no longer roar to jet engine ferocity at the slightest hint of a load.

[ Edited Coulomb: Changed link to https, corrected spelling error ]
Last edited by weber on Tue, 28 Jun 2016, 09:35, edited 1 time in total.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
Adverse Effects
Senior Member
Posts: 1172
Joined: Sat, 01 Jan 2011, 03:30
Real Name: Adverse Effects
Location: Brisbane

Re: PIP inverter repairs and hardware modifications

Post by Adverse Effects »

could you build a small unit to sit between the fan and the control board that would report a fan speed 1.5 x RPM that way you would not have that problem
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 »

weber wrote: We hypothesised that when the PIP requested the lowest speed from the Noctuas via their control input, they might be running marginally too slowly to satisfy the PIP, via their tacho output, that they were working.
Well, actually, moderate fan speeds. When the requested fan speed is less than 30% of full speed (i.e. with < 30% PWM), then the fan lock detect warning is cleared.
So Coulomb also did some heroic circuit tracing ...
Well, I don't know that the tracing was all that heroic:
Image

There must be something else that grounds pin 1 of the fans to turn them on. There are pairs of tall transistors nearby, but they didn't seem to connect, though I didn't try very hard.

There is another copy of this circuit (except for C20) that connects to FAN2. So when either fan provides pulses of a certain duration, they are integrated by this circuit. But unless C26 is a lot larger than I measure, the RC time constant is very short (microseconds). The software doesn't seem to integrate the signal, it just has to see a one at that port (GPIO13) once in a while to reset a counter. But if the counter gets to 250 (i.e. no fan signal is detected for 250 consecutive readings), then a locked fan warning is generated. (If the PIP is in a certain state, this warning can become an error). If this code is called 30 times per second (as most timing code seems to), this represents about 8.3 seconds.

My guess is that the original Adda fans generate a "movement detected" signal, but most fans including the Noctuas generate a signal that is on twice per revolution. So the reason that it takes about a day for the warning to come up is twofold: firstly, the load has to be substantial enough to make the fans come on at >= 30% speed, and secondly, you have to be unlucky to see 250 zeroes in a row when the fans are outputting a one twice a revolution. I believe that the width of the pulse is not mandated; just that there have to be two pulses per revolution of the fan blades. Perhaps the Noctuas generate a very narrow pulse, such that there is a 1% chance of seeing a one when polling the output regularly. So you are 99% likely to read a zero. So there is a (0.99)^250 = 0.08 = 8% chance of seeing 250 zeroes in a row, or one in 12.3 or so. So it might take on average about 8.3 times 12.3 ~= 103 seconds of moderate fan speed. Maybe the pulse is wider than 1%, so it takes longer to miss out on 250 pulses in a row. My apologies if my simplistic maths is wrong here; I think I'm assuming that the counter is reset to zero at the end of every 250 samples, but it isn't. But you get the idea.

One fix would be to replace C26 with a much larger value. It's a M1608 (imperial 0603) part, so it may be necessary to solder in a leaded electrolytic capacitor to get enough capacitance to guarantee seeing pulses whenever they are there. If the capacitor is too small, it would just be annoying, as it might just increase the average length of time till you fluke 250 zeroes in a row to say one week. So then you have an intermittent fault that takes a long time to confirm.

The relevant components (under the main PIP PCB) are these:

Image

Note that C26 (lower middle) is missing. I attempted to remove it from a spare PIP board that I have, but unfortunately it snapped off its end cap in the process. So I can't measure its value; the 1-3 nF indicated are some of the readings I got with a multimeter lead pushed against the end with the missing end cap. [ Edit: our best guess at present is 4.7 nF. ]
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 »

Here is a photo of the PIP main board near the fan outputs:

Image

It's not obvious to me whether U9 is involved or not. [ Edit: it is not; it's for the DC-DC converters. ] It's an SG3525AN, a pulse width modulator control chip, so it seems likely. Though the DSP microcontroller chip does PWM, so I don't see why it's needed. Q15/Q16/Q41/Q43 are the tall medium power type, which may well drive the fans via pin 1.

[ Edit: so that suggests that the PIP is producing a 0-12 V signal from the PWM signal, to send to the fans. Maybe the fans aren't driven with the PWM input at all. I can't find much information on the original ADDA fans. ]

My apologies for the extraneous artifact, no thicker than a dog hair :-)
Last edited by coulomb on Tue, 28 Jun 2016, 11:31, 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
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 »

Adverse Effects wrote: could you build a small unit to sit between the fan and the control board that would report a fan speed 1.5 x RPM that way you would not have that problem

I doubt that 1.5x would be enough, and it would be simpler to simply stretch the pulses with a 555 or 556 timer chip. or just increase the value of C26.

If the original fans prove to be too noisy after all, we may need to resort to something like that.
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.
paulvk
Senior Member
Posts: 457
Joined: Fri, 23 Oct 2015, 23:45
Real Name: Paul
Location: Sydney

Re: PIP inverter repairs and hardware modifications

Post by paulvk »

With the new firmware the fans are not noisy infact I think they run a bit too slow as the heatsink gets to 50c and they are still not running flat out.
I get a few extraneous alarms on watch power , output short circuited, fan locked and a battery fault but this maybe watch power it may not check the crc for valid data Note this is on different inverters which all appear to be working fine.
I have not had time to finish building my AVR web server which will log the data to SD card and checks the crc for valid data.
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 »

For those interested in the serial saga of the serial (RS232 communications) boards, I've added another photo, courtesty of RiaaanH:

Scott's comparison of 4 (now 6) RS232 boards.

I put the photo in the earlier post, so that there are now 6 versions (from 6 years, amazing), all in one place, accessible from the index.

It's a bit strange that it has no transformer or 2-pin connector for power (marked HFPW), so now I'm wondering if it's different to the 4 kW / 5 kVA models. Can anyone confirm?
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
Tejota
Noobie
Posts: 24
Joined: Tue, 29 Nov 2016, 12:07
Real Name: MDKTejota
Location: Spain
Contact:

Re: PIP inverter repairs and hardware modifications

Post by Tejota »

Scott wrote:
coulomb wrote: after looking at your photos again


Thanks for the photos, they're very helpful.

That black box is a relay (see here) and it looks to me like they're routing it out to the RJ45 on pins 3,4,5 although it's difficult to see clearly with all the silkscreen.

Image

I haven't measured but I'm pretty sure the HFPW pins are 'high frequency power' which drive the primary side of the isolation transformer. The secondary side would the be center tapped and the two zener diodes clamp it to plus and minus 12VDC for the RS232 voltages. The two caps would then smooth the two supplies.

I managed to find a destructive test report that had photos of the '2012' board. It doesn't use the separate 2 way HFPW connector but still uses HFPW+ on the other 6 way connector. I assume that the + line wrt ground will generate enough power to drive the RS232 line, but not enough to drive an external display. It doesn't look like there's any power line on the 2012 board's RJ45 at all.

Here's the photos I have of all the boards, the dates on the transformers match up.
Image

EDIT: Added photos of 2015 board

[ Edit Coulomb: added 2016 board below. It's from a 3 kVA machine, in case it matters. Thanks to RiaanH for the images. ]

Image


Could I use 2015 comm board on 2013 PIP instead 2013 comm board? I mean, I need USB port on 2013 PIP (mine is manufactured on March 2014 but It has got 2013 comm board picture). I must do hole to adapt usb connector but no problem. I think switch power conn is only to manage dry connect relay so I wouldnt need this conn. Rest of internal conn are the same as 2015, 2013 comm boards.
Any advise?
paulvk
Senior Member
Posts: 457
Joined: Fri, 23 Oct 2015, 23:45
Real Name: Paul
Location: Sydney

Re: PIP inverter repairs and hardware modifications

Post by paulvk »

I have been looking at the solar charger:
Two dual TO220 diodes V30200C 15Amp x 2 at 200v thus a limit of 60 amps (another part running at full rating)
Current sense is 6 x R002 surface mount (.002 ohms) in two sets of 3

There are two chargers on the board running in parallel.
So by adding another R002 on each set we should come close to getting 80 amps when it thinks its 60 amps.
But we need to change the two dual diodes so would need to go to dual 25amp
Looking I can only find dual 40amp 170V in TO-3PW package but the mosfets IRFB4321 are only 150V 85amps so they should be ok at 170v.

Might do one and see how it goes.
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Re: PIP inverter repairs and hardware modifications

Post by weber »

Hi paulvk. That's interesting. I suspect you might have to upgrade the MOSFETs to a higher current rating as well, despite their nominal rating.

On the subject of circuit modifications: I recently became curious as to where that 50 watts is going when the PIP is in battery mode with no load. It has to appear as heat somewhere. So I left the inverter doing nothing for a few hours, then I took the cover off and waved an IR thermometer around. The hottest thing by far was the large toroidal inductor in the upper left of the main PCB. It was running at over 100°C.

I'm guessing it is part of an LC filter on the inverter output, intended to block the PWM frequency, leaving clean 50 Hz. I assume the PWM frequency is more than 20 kHz. But because the inductor is wound with solid 1.3 mm diameter wire [Edit: Turned out to be trifilar 1.45 mm], not litz wire like the other inductors, it is behaving more like a resistor than an inductor, due to skin effect at the PWM frequency.

I calculated that litz wire made of 35 strands of 0.2 mm dia should be suitable.
http://www.ebay.com.au/itm/Litz-Wire-35-32-Enameled-copper-wire-twisted-pair-AWG32-X-35-Strands-A37J-LW-/172081474460?var=&hash=item6da56366ac

But I've balked at disassembling a PIP and desoldering and unwinding the inductor to (a) count the turns, and (b) measure the wire length.

If someone else was to do this. Then take the risk to order the litz wire and rewind the inductor, and measure the new standby battery consumption, they might become a PIP/Axpert legend. Image
Last edited by weber on Fri, 20 Jan 2017, 07:36, edited 1 time in total.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
offgridQLD
Senior Member
Posts: 1827
Joined: Tue, 23 Jul 2013, 16:05
Real Name: Kurt
Location: Fleurieu Peninsula, SA

Re: PIP inverter repairs and hardware modifications

Post by offgridQLD »

Weber, I have a spare pip that hasn't been installed yet. I am happy to try it on. Thanks for doing all the ground work. I will order the litz wire and update when it arrives.

Thanks

Kurt.
Last edited by offgridQLD on Fri, 13 Jan 2017, 18:23, edited 1 time in total.
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Re: PIP inverter repairs and hardware modifications

Post by weber »

offgridQLD wrote: Weber, I have a spare pip that hasn't been installed yet. I am happy to try it on. Thanks for doing all the ground work. I will order the litz wire and update when it arrives.

That's awesome Kurt!

I tried to estimate the length of wire needed, from measurements of the wound toroid, but it's completely covered by multiple layers, so it's impossible to tell how much is core and how much is wire. So the only way to be sure is to unwind it and measure it. At least the re-winding should be much easier than the un-winding (and the original winding) as the litz wire will be so much more flexible.

Of course the whole 50 watts won't be dissipated in that inductor. But I wouldn't be surprised to learn that half of it is.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
paulvk
Senior Member
Posts: 457
Joined: Fri, 23 Oct 2015, 23:45
Real Name: Paul
Location: Sydney

Re: PIP inverter repairs and hardware modifications

Post by paulvk »

As I have a PIP apart I can see that the inductor is potted at the bottom so not going to be easy to take apart.

Looking at the SCC board it has the two connections to the inverter marked as "To Axpert" so we know who came first.

On the subject of mosfets it would be nice to know what ones and how many in the 80amp versions.
Last edited by paulvk on Sat, 14 Jan 2017, 03:30, edited 1 time in total.
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 »

offgridQLD wrote: Weber, I have a spare pip that hasn't been installed yet. I am happy to try it on. Thanks for doing all the ground work. I will order the litz wire and update when it arrives.

Yes, thanks, Kurt! If it can be done with low enough pain, it could become another automatic modification, like a lot of us invert the fan direction.

I have an older model main board here for parts, so in the unlikely event that your efforts completely fail, the inductor from this unit should work. This main board has no processor daughter board or case, so it's difficult to test with.
Last edited by coulomb on Sat, 14 Jan 2017, 03:36, 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
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 »

They have made some attempt at reducing the skin effect; the inductor in question is wound "three in hand" (tri-filar, three wires wound together and in parallel):

Image     Image

It's a little difficult to see from the first photo whether it's two or three in hand, but the second photo from underneath the board makes it very clear.

Edit: you can also see some of the black potting material Paulvk mentioned at the bottom of the first photo. Unfortunately, it's the rock hard type, not the stiff rubbery kind. But Kurt is a resourceful fellow, so I'm sure he'll figure out a way to remove the potting material without mangling the wire too much. A heat gun might be the first line of attack.
Last edited by coulomb on Sat, 14 Jan 2017, 05:37, 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
offgridQLD
Senior Member
Posts: 1827
Joined: Tue, 23 Jul 2013, 16:05
Real Name: Kurt
Location: Fleurieu Peninsula, SA

Re: PIP inverter repairs and hardware modifications

Post by offgridQLD »

Thanks for the pics coulomb.

I assume the potting material was used just to mechanically fix it to the board and limit the inductor from resonating to much.

I guess I will need to obtain some kind of potting material to achieve the same once it's rewound.

I notice the original solid wire has a resin coating. Is the litz wire in the ebay listing raw. Edit: I see now it's listed as enamel coated.

Kurt.
Last edited by offgridQLD on Sat, 14 Jan 2017, 06:18, edited 1 time in total.
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Re: PIP inverter repairs and hardware modifications

Post by weber »

3-in-hand makes a lot more sense. Thanks Coulomb. I was wondering how they could get away with putting 20 amps through 1.3 mm² in such a packed winding. But in fact it's 5 mm². They may not have done it with skin effect in mind. They would need to do that just so it's flexible enough to physically do the winding.

So Kurt, that means you should use 100 strand litz wire, like this:
http://www.ebay.com/itm/Litz-Wire-100-32-Enameled-copper-wire-twisted-pair-AWG-32-X-100-Strands-A37P-LW-/182007478284?var=&hash=item2a607db80c:m:mr5eLYLsHiJuBzOqCTn1BhA

That's only 4 mm² in total copper cross section [Edit: actually it's only 3.1 mm², as I point out in a later post], but if you used 120 strands (to get nearer 5 mm²) I don't think you would fit the same number of turns as the original, due to every strand being individually enameled. As it is, you will need to wind it firmly and evenly. But if we succeed in significantly reducing the heating due to PWM then we'll have more heat dissipation capacity remaining to cope with the maximum 50 Hz load.
offgridQLD wrote:I assume the potting material was used just to mechanically fix it to the board and limit the inductor from resonating to much.

I guess I will need to obtain some kind of potting material to achieve the same once it's rewound.
Yes. It's purely mechanical fixing. If it was needed to stop the inductor from emitting sound, it would fill the whole winding. Yes, you can worry about that if and when it works. But any two-part epoxy such as Araldite will do.
I notice the original solid wire has a resin coating. Is the litz wire in the ebay listing raw. Edit: I see now it's listed as enamel coated.

Yes. It wouldn't be litz wire if every strand wasn't individually insulated from the others. It must also be arranged so that no strand spends more time near the centre of the bundle. So they can't just twist the lot. They have to twist them in groups of 5 or less first, then twist 5 or less of those bundles together and so on.
Last edited by weber on Sun, 15 Jan 2017, 11:42, edited 1 time in total.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
offgridQLD
Senior Member
Posts: 1827
Joined: Tue, 23 Jul 2013, 16:05
Real Name: Kurt
Location: Fleurieu Peninsula, SA

Re: PIP inverter repairs and hardware modifications

Post by offgridQLD »

Dam it I ordered the original ebay listing.

Edit: ok I see it's from the same seller see if they can emend that.

Kurt
Last edited by offgridQLD on Sat, 14 Jan 2017, 08:54, edited 1 time in total.
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 »

I thought of a way of estimating the length of wire needed for the PIP's large inductor, before anyone pulls one apart. Presuming Weber's claim of 1.3 mm² wire (about 1.3 mm diameter as it happens), and using Wikipedia's tables for 16 AWG wire (13.17 mΩ/m each wire, or 13.17/3 = 4.39 mΩ/m for three in hand), so each metre of trifilar wound wire would have 4.39 mΩ of resistance. I guessed the toroid's cross section at 30 x 15 mm, for a perimeter of 90 mm. Call it 100 mm with thick wire bending radii. So ten turns to the metre, or 0.439 mΩ per turn.

I put three amps through the coil with my bench power supply, and measured just over 58 mV voltage drop. So that's a resistance of 58/3 = 19.3 mΩ. So that's 19.3 / 0.439 = 44 turns. That seems reasonable. So those 44 turns are going to need some 4.4 metres of wire. There are many approximations in the above, so the result would be +- 25%.

Now to convert to caveman units; I believe that's some 14.4 feet. So Kurt, I hope you ended up with 20' of the 100 strand litz wire, or 50' of the 35 strand wire (which you could wind three in hand).

I didn't see any mention of the actual length of wire you are getting for the stated price in the second Ebay listing.

For the record, my old Digitech multimeter measures the inductance at 0.39 mH (390 μH).

[ Edit: added some context at the start. ]
Last edited by coulomb on Sat, 14 Jan 2017, 11:05, 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.
paulvk
Senior Member
Posts: 457
Joined: Fri, 23 Oct 2015, 23:45
Real Name: Paul
Location: Sydney

Re: PIP inverter repairs and hardware modifications

Post by paulvk »

Before taking the inductor apart I would measure the inductance.
It may also be simpler to just get a core and make a new inductor with so much room even an ferrite "E" core could be used

Also after the inductor there is an AC current measuring sensor then what appears to be a 20uF 550V capacitor after this a common mode choke.
As my 2 year warranty's are now up I will be changing the capacitors I found some 85deg 3300uF 80V, 100V surge, use switching power supplies only 2mm larger in diameter on Element14 when I get back to my large system I will post the part number
Last edited by paulvk on Sat, 14 Jan 2017, 11:31, edited 1 time in total.
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Re: PIP inverter repairs and hardware modifications

Post by weber »

Oops! I messed up. 100 strands of 0.2 dia isn't 4 mm² it's 3.1 mm².

[Edit: I earlier wrote here that 120 strands would be required, but Coulomb convinced me that 3.1 mm² is sufficient and 120 strands won't fit.]

On the eBay listing, there's a "length:" menu just above the price and quantity.

Well done calculating the length Coulomb! So a 30 ft roll will probably do two inductors. [Edit: turned out that's wrong. Keep reading]
Last edited by weber on Fri, 20 Jan 2017, 07:45, edited 1 time in total.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
offgridQLD
Senior Member
Posts: 1827
Joined: Tue, 23 Jul 2013, 16:05
Real Name: Kurt
Location: Fleurieu Peninsula, SA

Re: PIP inverter repairs and hardware modifications

Post by offgridQLD »

At $50 for a rewire and teir 1 PV now avalable at 75 cents a watt it's starting to look less atractive.

Though if your running a small battery or multiple pips then it might be worth a shot.
Last edited by offgridQLD on Sat, 14 Jan 2017, 16:59, edited 1 time in total.
User avatar
offgridQLD
Senior Member
Posts: 1827
Joined: Tue, 23 Jul 2013, 16:05
Real Name: Kurt
Location: Fleurieu Peninsula, SA

Re: PIP inverter repairs and hardware modifications

Post by offgridQLD »

I pulled the front cover off my two pip4048 inverters today to familiarize myself with the inductor in question.

Way larger than I had imagined. I see perhaps one potentual issue when rewiring it with litz wire that will take up more space.

On my 2013 model (with the solar charge controller up top large black heat sink) there is plenty of free space surrounding the inductor.

On on the late 2016 model (solar charge controller down the middle of the unit) the inductor sits very close to the inverters heat sink.

Could this be a issue?

you can see in this pic taken sideways that the silver heat sink is just a few mm away from the inductor wire in its stock form.
Image

Image

Image

Perhaps best to try it on a early model (black heat sink up top) PIP4048 first. Im sure if your keen after you could trim some off the heatsink to create room. Or perhaps lean over the inductor a little? Will all depend on how much it grows when rewound.

Kurt
Post Reply