PIP-4048MS inverter

Post by **coulomb** » Sat, 10 Feb 2018, 06:56

lion6912 wrote: ↑
Thu, 08 Feb 2018, 19:53
I would live to know the change of scc firmware ? I have the 1.24

Files for updating the SCC firmware are in this post; instructions are in the "SCC Updating" section at the end. Note that if you use patched firmware version 72.70c or later, there is no advantage to updating the SCC firmware. Weber and I added a lot of patches to work around issues related to the SCC firmware version. Updating the SCC firmware is also a little more tricky than updating the main inverter's firmware.

Post by **coulomb** » Sat, 10 Feb 2018, 09:11

There has been some discussion recently, here and in the South African Power Forum, about which versions of firmware are able to co-operate with which other versions, when paralleling more than one inverter. The same issues would affect three-phase systems. I assume that all three-phase systems would have sets of three machines bought at the same time, but for very large systems, another set of three might be added later.

When all the machines run the exact same firmware version (e.g. 72.70c with 72.70c, not 72.70b or 72.70), then there is no problem with paralleling. However, this post shows how some different firmware versions can be paralleled successfully.

This post refers exclusively to 48 V machines rated at 5 kVA inverter output (both 4 kW and 5 kW models).

There have been several major version numbers (the number before the dot) in official firmware versions:

52.XX: these came with older machines, whose SCCs were limited to 60 A. The maximum AC charging has always been 60 A, so the total charge current (setting 2) was limited to 120 A. Updating to newer firmware on such machines gives you the option of increasing the maximum battery current to 140 A, but it will never achieve this. So such updating is harmless. These older machines were all 4 kW/5 kVA.
72.XX: this series replaced 52.XX. 72.40 is still limited to 120 A maximum total charge current, hence 60 A for the SCC. 72.6X and 72.70 have the maximum charge current at 140 A, and allow SCC charging to 80 A if the SCC firmware allows it.
73.XX: This appears to be an update to 72.70, which controls a new relay that connects neutral to earth when the inverter is in battery mode, and leaves them unconnected otherwise. Running 73.XX on machines that came with 72.XX and even 52.XX should be okay, since it presumably will only attempt to turn on a relay that isn't there. It looks like 73.XX still operates the "dry contact" relay if setting 38 (Neutral to Earth Connection) is "EnA".
74.XX: This comes with the 5 kW models (PF1.0).
75.XX: This comes with the dual/triple SCC models. It has a different protocol for the SCC commands, and a different LCD segment mapping, so it's not compatible with models with single SCCs.

For what can parallel with what, the firmware uses a constant that it expects the other machines to send in a special CAN packet. These are the constants for the firmwares I have examined, which therefore separate the firmware versions into "parallel compatibility groups". There are additional factors related to our patches that further subdivide the groups. Only firmwares on the same line are "parallel compatible".

Constant      Firmware version(s)
112           52.02
116           52.30, 72.40, Pb_Cv.PCH, 72.60, Pb1_72.60a, Pb1_72.60b, 72.61, 73.00
116f          Li_FE.PO4, Li1_72.60a, Li1_72.60b
116fz         LF1_73.00a
116z          LC1_73.00a
118           72.70, LC1_72.70a, LC1_72.70b, 74.10 (presumed)
118f          LF1_72.70a, LF1_72.70b
118fz         LF1_72.70c
118z          LC1_72.70c
130           75.31, LC1_75.31a

Note how 72.70 is in a different parallel compatibility group to the other 72.XX versions.

So it seems that it should be possible to parallel 5 kW (PF1.0) machines with 4 kW machines (PF0.8), as long as the 4 kW machines are running 72.70.

It seems that dual/triple SCC models can't be paralleled with anything else, at least at this stage.

Post by **coulomb** » Sat, 10 Feb 2018, 11:27

As far as I know, the output voltage setting procedure is not documented at all. My initial research indicated that you have to send it a "V220" command to change the output voltage to 220 V. However, this turns out to be dead code in the firmware; it has no effect other than very slightly wearing out the EEPROM.

Subsequent research reveals that you need a POPV2200 command to change the output voltage to 220.0 V. The other two possibilities are POPV2300 and POPV2400 for 230 V and 240 V respectively.

The on/off switch at the bottom of the inverter needs to be in the off position for this command to succeed, otherwise the inverter will NAK the command and ignore it. I believe that with the switch in the off position, the inverter will shut down in about a minute or so, so don't switch off till you have everything ready.

You will need a computer with a USB port and a USB to RS232 adapter, and a communications program capable of sending 8-bit characters. The 8-bit characters are required for the CRC characters that are needed at the end of each command.

[ Edit : the USB to RS232 adapter needs to connect to the inverter-charger via the provided D9 to RJ45 cable. ]

Choose one of these methods (or use some other communications program that is capable of sending 8-bit characters):

See the end of this post for setting up Tera Term. Note: to enable 8-bit character transmission, you need to find and edit the initialisation file, as per the instructions in the linked post.

Having set up Tera Term, enter the following for 220 V output:
shift+P shift+O shift+P shift+V 2 2 0 0 m right alt+shift+@ enter

For 230 V:
shift+P shift+O shift+P shift+V 2 3 0 0 shift+Z right alt+p enter (CRC corrected)

For 240 V:
shift+P shift+O shift+P shift+V 2 4 0 0 right alt+shift+_ (underscore, usually shift "-") ` (back quote, usually to the left of the "1" key) enter
Alternatively, this post has the CrcGen app for calculating the CRCs and pasting the command with CRCs as hexadecimal into the clipboard. Use with Realterm. NOTE: I could not figure out how to turn on Realterm's hex mode, so if using Realterm, instead of using Scott's CrcGen app, you could paste one of the following strings into the appropriate drop-down box on Realterm's Send tab:
0x50 0x4F 0x50 0x56 0x32 0x32 0x30 0x30 0x6D 0xC0 (for 220 V; don't copy these comments in brackets)
0x50 0x4F 0x50 0x56 0x32 0x32 0x30 0x30 0x5A 0xF0 (for 230 V; CRC corrected)
0x50 0x4F 0x50 0x56 0x32 0x32 0x30 0x30 0xDF 0x60 (for 240 V)

Either way, you should see an "(ACK9 " response (no quotes), and not a "(NAKss".

Unfortunately, I am not in a position to test the above, so I may have slipped up. The commands themselves have been tested, running at 240 V overnight, and at 220 V for several days (thanks, Weber!), but we have a different way of sending the commands, with CRCs generated automatically by our BMS master.

When paralleling machines, they all have to be set to the same voltage. If they are not, an undocumented Error 73 results. The obvious solution is to manually change the voltage settings as required.

If you were running WatchPower, you'll have to exit it (not just close the window). If unsure, see the how to update firmware post for exactly how.

[ Edit: @ is usually shifted. ]
[ Edit: corrected ACK and NAK responses. ]
[ Edit: CRCs calculated in BMS master, not Node-red. ]
[ Edit: arrgh! CRC was wrong for the POPV2300 command. Thanks, Weber! ]
[ Edit: added note re Error 73. ]

Post by **lion6912** » Sat, 10 Feb 2018, 15:50

Thanks coulonb. Very very good.
I can try to change 230V to 220V, but I dont know Tera Term, CrcGen, realterm.
It is possible to read EEPROM ?
I have launch Tera term, I select COM1, and I have a dark screen.
What do you do now ?

Post by **lion6912** » Sat, 10 Feb 2018, 16:27

I try realterm, but Where can put " 0x50 0x4F 0x50 0x56 0x32 0x32 0x30 0x30 0x6D 0xC0 "

Post by **coulomb** » Sat, 10 Feb 2018, 16:57

lion6912 wrote: ↑
Sat, 10 Feb 2018, 16:27
I try realterm, but Where can put " 0x50 ... "

First, ensure that you have the correct bit rate and port selected, using the Port tab (brown):

: Realterm set 220Vb.png (15.02 KiB) Viewed 2319 times

Note that 2400 is right down the bottom of the "Baud" drop-down box. Select the appropriate port; it varies from machine to machine, and also depends on which USB port you plug your USB to RS232 adapter. (I forgot to mention that you need to connect to the inverter via the supplied D9 to RJ45 cable; I've now edited the original post to include this step.) I usually use Tera Term to tell me which port my USB to serial adapter is on; for me it's port 10 but it could be anything else for you. Sometimes Realterm includes a description of the device connected to the port, sometimes it doesn't. You may have to press Open or Change to actually connect; you'll know when it does connect by the message in the bottom right corner of the Realterm window:

: Realterm set 220Vc.png (7.29 KiB) Viewed 2319 times

If it says "Port: closed" then you haven't opened it successfully yet. The "10" could be any number, of course, depending on your computer.

Now go to the Send tab (red highlight):

: Realterm set 220Va.png (22.03 KiB) Viewed 2319 times

Tick the two boxes indicated in green. Paste into the dropdown box where the blue is, using the standard Windows key combination of ctrl + v. Then press the Send Numbers button (orange). You should see "(ACK9" as a response. If you see "(NAKss", make sure that the inverter's on/off switch is off. If you see no response, check the bit rate and port settings, the USB to RS232 adapter, and cable to the inverter.

[ Edit: added third screenshot. ]

Post by **lion6912** » Sat, 10 Feb 2018, 17:26

It is GREAT.
Very thanks Mister coulomb.
I am at 220V.
But if I can manualy 220V, I could put over than 58.4V for battery ?

Post by **coulomb** » Sat, 10 Feb 2018, 18:37

lion6912 wrote: ↑
Sat, 10 Feb 2018, 17:26
I am at 220V.

Well done!

But if I can manualy 220V, I could put over than 58.4V for battery ?

There is a special-order model of the PIP-4048 that allows up to 64 V of charging, if you think you really need to go beyond 58.4 V. I don't know anything else about it, not even a name for it. I assume it has different MOSFETs and higher voltage capacitors, and different firmware.

The attachment is from http://www.voltronicpower.com/oCart2/in ... uct_id=132 .

Post by **lion6912** » Sat, 10 Feb 2018, 21:17

I dont want reach 64V but only near 59.4V . But I dont want crah mosfet or capacitors.
Thanks

Post by **coulomb** » Mon, 12 Feb 2018, 14:56

Release Version of Patched Firmware 73.00b

[Edit: This post was originally for the beta version of 73.00a, but after being in use for 6 weeks, and being downloaded 38 times for LFP and 55 times for other chemistries, we updated the zip files to "release" status. Then a few days later we learned about a problem where 73.00 (both original and patched) would stop working after 60 days (with fault code 90) when run on older hardware. 73.00b fixes that problem. It was a one-word change, so we didn't bother with a beta.]

Weber and I have finally finished testing our patched firmware based on factory firmware version 73.00. Hardware that comes with factory firmware 73.00 has a neutral to earth relay, which is active only when the inverter is in battery mode. 73.00 also modifies the behaviour of the fans, spinning the fans even at no load, and more vigorously during heavy solar charging. There may be other improvements as well. 73.00 is parallel compatible with some older firmwares: 52.30, 72.40, and 72.6X, but not 72.70.

This patched firmware has all the same patches as 72.70c and earlier patched firmwares, including Dynamic Current Control and AussieView™. Naturally, this includes fixing the infamous premature float bugs.

As before, there is an LFP (Lithium Ferrous Phosphate) version, and the "all other chemistries" version (LC for Lead aCid, Lithium Cobalt blends, and lithium titanate).

Note: This firmware is only suitable for the 48 V models with a single low-voltage MPPT and a power factor of 0.8 (4 kW / 5 kVA), not those with dual or triple MPPTs or a power factor of 1.0 (5 kW / 5 kVA), and not those with a maximum PV array open circuit voltage greater than 145 V DC.

For lithium ferrous phosphate (LFP) (16S or 15S)

dsp_LF1_73.00b.zip: (1.53 MiB) Downloaded 43 times

For lead acid (24S), lithium cobalt-blends (LCO, NMC, NCM, NCA) (14S), and lithium titanate (LTO) (21S)

dsp_LC1_73.00b.zip: (1.53 MiB) Downloaded 87 times

Note: Lithium ferrous phosphate (LFP) is the same as lithium iron phosphate or LiFePO₄. The name "lithium ferrous phosphate" is preferred, as it avoids the confusion between "iron" and "ion".

For details on reflashing, and what hardware this firmware will and won't run on, see the 72.70b reflashing instructions, but ignore the files there.

[ Edit: Links to text files for DCC and AussieView™ ]

Post by **coulomb** » Mon, 12 Feb 2018, 18:01

I've updated the list of parallel compatibility groups.

Post by **al76** » Wed, 14 Feb 2018, 10:25

weber wrote: ↑
Wed, 31 Jan 2018, 15:19

al76 wrote: ↑
Wed, 31 Jan 2018, 13:57
The OC voltage reported by the Pip is about 130v Once mppt kicks in with charging it drops to about 110v
From this graph you can see that Volts sits about 130v until about 10am the mppt/charging kicked in.
...
I turned off the solar about 4pm ( on graph you can see Volts drop to zero) to try and change the string but I had a connector issue then it it started raining
Hi Alan, If your connector issue was difficulty in getting MC4 connectors apart, I note that a 16 mm (or 5/8") open-end spanner works well to compress the barbs and allow them to come apart. Much better than fingers (too fat to fit in the notches), and much better than pointy nose pliers (jaws not parallel). But even with the spanner compressing the barbs, it can take a lot of pulling and wiggling to get them apart.
I turned the solr back on and as you can see the V went back to about 125v and the relay just clicks away again.
Given the OC volts is well below max of 145v is it possible there is some issue with the pip?
I'll be adding some more panels soon so I may have to wait to do the string mod test.
It does seem that your PIP is more sensitive to high voltage than is claimed in the manual, but for all I know, they may all be like that. What are your U1 and U2 firmware version numbers?

But no matter whether the PIP is out of spec or not, there is no doubt that the optimum number of 72-cell panels in series, for a PIP-4048MS, is 2. Please just do the experiment.

I finally changed the strings this morning to 2 strings of 2 panels was 2 strings of 3. The problem persisted with the relay clicking away. After approx an hr the charging started.

Post by **paulvk** » Wed, 14 Feb 2018, 12:13

Have a problem with 72.70 the Max. Charging Current setting has become a max current setting for solar charger so if you set it at 30amps that is all the solar charger will put out instead of being max charge to battery.
Only found this out because I loaded them up to 4500watts and was getting a battery discharge then saw that only 30amps of total PV was there.
Two pips in parra with 225ah of LA flooded.

Post by **weber** » Wed, 14 Feb 2018, 13:47

al76 wrote: ↑
Wed, 14 Feb 2018, 10:25
I finally changed the strings this morning to 2 strings of 2 panels was 2 strings of 3. The problem persisted with the relay clicking away. After approx an hr the charging started.

OK. So now we know there is definitely something wrong with your machine. This is definitely not expected behaviour. My next suspect is the relays in the SCC. Maybe their contacts are badly burned and making intermittent contact?

Post by **weber** » Wed, 14 Feb 2018, 13:56

paulvk wrote: ↑
Wed, 14 Feb 2018, 12:13
Have a problem with 72.70 the Max. Charging Current setting has become a max current setting for solar charger so if you set it at 30amps that is all the solar charger will put out instead of being max charge to battery.
Only found this out because I loaded them up to 4500watts and was getting a battery discharge then saw that only 30amps of total PV was there.
Two pips in parra with 225ah of LA flooded.

It sounds like you have the Solar Balance parameter [31] set to Sbd. If so, you need to set it to SbE (the default).
See the bAL [31] section of the AussieView manual or page 23 of the manufacturer's manual.

Post by **al76** » Wed, 14 Feb 2018, 19:17

weber wrote: ↑
Wed, 14 Feb 2018, 13:47

al76 wrote: ↑
Wed, 14 Feb 2018, 10:25
I finally changed the strings this morning to 2 strings of 2 panels was 2 strings of 3. The problem persisted with the relay clicking away. After approx an hr the charging started.
OK. So now we know there is definitely something wrong with your machine. This is definitely not expected behaviour. My next suspect is the relays in the SCC. Maybe their contacts are badly burned and making intermittent contact?

Is it worth following this up with mpp solar or move fwd with buying another unit and have this unit repaired in Aus? I was planning on a parallel system plus I will be needing a second independent system soon anyway.

Post by **weber** » Wed, 14 Feb 2018, 19:39

al76 wrote: ↑
Wed, 14 Feb 2018, 19:17
Is it worth following this up with mpp solar or move fwd with buying another unit and have this unit repaired in Aus? I was planning on a parallel system plus I will be needing a second independent system soon anyway.

I can't really advise you on such decisions. Except to say that I don't see that you have anything to lose by raising the problem with MPP Solar.

How long has the PIP-4048MS been in use? Could there be a fault in the isolation device between your PV array and the PIP?

Post by **simat** » Thu, 15 Feb 2018, 22:19

Hi,
I am really impressed by the amount of excellent information in this thread and especially at all the work done by Coulomb and Weber. This is forums at their best! Thank you!

I have been asked by a friend of a friend who I designed and installed an LFP based off grid system for in 2013 if I would help him get an off grid system up and running.

This thread has made me think that it might be worth considering using the PIP4048MS or maybe the 5048 rather than a separate charge controller and inverter from more mainstream manufacturers. I am considering the Victron Phoenix inverters and EPSolar charge controllers.

The price of the PIP4048MS makes it worthwhile buying a spare and all the information in this thread will make it easier to interface my BMS to the SCC and inverter in the PIP and fix it if there are any problems.

My plan would be buy two units, run one for 3-5 years or until it blows up (hopefully not!) and then swap it with the spare. I would replace the main caps as detailed in this thread in the unit that I had swapped out.

I was thinking of configuring the LFP battery as 15s rather than the normal 16s to drop the voltage a bit. Only issue I can see with this is that it will increase the current. Which is preferable, decreasing the voltage stress or increasing the ripple currents and component temperatures due to the increased current?

I wanted to shut down the SCC via a signal from my BMS. Am I right in thinking that I can do this by sending a command to the PIP to raise the bulk and float voltages, or is there another way to do this?

I also wanted to be able to shut down the inverter. After reading the thread a few times! I am not sure if this can be done by sending a command but is possible by simulating the On/Off switch on the PIP being turned off, am I right?

Simon

Post by **T1 Terry** » Fri, 16 Feb 2018, 12:33

Hi Simon, I've had success with the 15 cell arrangement charging via a PWM controller using nom. 12v panels (18v Vmp) in series. The 24v units come in 5000va (4kW) models and they don't seem to have current rated issues so I can't see the 48v model having issues. The days of needing high voltage battery packs to reduce the current requirements really date back to the lead acid battery days, using LYP cells where high current draw is not an issue the days of requiring the use of a 48v battery pack and charging components might really be a carry over rather than a necessity.
I recently installed a 32 x 200Ah cell battery (4p8s 24v) with a 4kW PiP inverter and 5.6kW of solar and the system is out performing any and all expectations with none of the herding feral cats using Dingoes type problems the 48v systems seem to enjoy.

T1 Terry

Post by **lopezjm2001** » Thu, 01 Mar 2018, 13:24

coulomb wrote: ↑
Mon, 12 Feb 2018, 14:56
Beta Version of Patched Firmware 73.00a

Thanks for the great work you guys do, Coulomb and Weber. I updated my PIP4048MS to your latest beta firmware a couple of days ago. Not sure whether I needed it or not as I was running the latter version.
You can see a graph of your beta 73.00a firmware at work at https://drive.google.com/open?id=1U7GgT ... pzOlgMBowp, it works well. Can also view my PIP4048MS at https://emoncms.org/dashboard/view?id=33822 as is setup for public viewing. You can see how it goes into float mode at the right time. Does bulk charge followed by float mode even on cloudy days. No problem.
I like how my PIP4048MS LCD screen now shows the temperature of the unit... nice.

You guys should try to get crowd funding for this..

Post by **weber** » Sun, 04 Mar 2018, 09:46

Thanks for that, lopezjm2001. As we don't ask for money, it is kind words like yours that keep us going. I'm glad you like the "AussieView" LCD improvements. I'm surprised at how little feedback we've had about those. But we try to assume: "No news is good news".

Post by **paulvk** » Mon, 05 Mar 2018, 20:12

I have 72.70 running on the systems and the battery performance appears to be much better but note I also have the 9 panels with the e-smart3 as well.
The battery is staying at a higher voltage for much longer while under discharge at night, the balance of the current from/to the banks is much better.
As a companion to the PIP the E-smart3 is a good match giving the extra 40amps to take solar to 100amps which gives the system enough to run the loads and charge the battery properly.
As far as the now usable LCD goes I have not had enough time to do a lot with it but it is nice to be seeing batt volts , solar current and batt current on two PIPs
Note I have VRLA

I am going to try and find time to finish the remote display I started on which uses a dotmatrix LCD to display most of the parameters all at once from the QPIGS command.

Post by **Tejota** » Wed, 07 Mar 2018, 05:29

coulomb wrote: ↑
Sat, 10 Feb 2018, 18:37

There is a special-order model of the PIP-4048 that allows up to 64 V of charging, I assume it has different MOSFETs and higher voltage capacitors, and different firmware.

MOSFETS in PF1 5kW up to 64V charging are TK100E08N1

I have seen them through the grid of holes. I dont know if there are 16 or more. I cant watch capacitors.... I would have to open the device invalidating the warranty.

A question about parallelized pips: If mosfets/igbts in one unit of the cluster fail, what happens in the other units of the cluster?

Regards.

Post by **coulomb** » Wed, 07 Mar 2018, 06:23

Tejota wrote: ↑
Wed, 07 Mar 2018, 05:29
MOSFETS in PF1 5kW up to 64V charging are TK100E08N1

Wow! So these are 80V MOSFETs with Vgs=0, but only 60V with Vgs=-20 V. The datasheet doesn't say what voltage is safe with intermediate gate voltages, but suggests (to me) that with any significant negative gate voltage, which I assume is standard, that the limit is 60 V. So there must be a non-standard gate drive for these. Or I'm misunderstanding the datasheet.

I have seen them through the grid of holes.

Well sleuthed! Maybe an optic fibre camera next?

Is there a special firmware revision number? I imagine they would want to distinguish the high voltage versions from others, yet indicate what other features and/or bug fixes it has in common with the nearest non high voltage version.

A question about parallelized pips: If mosfets/igbts in one unit of the cluster fail, what happens in the other units of the cluster?

A great question. Firstly, I'll consider MOSFET failure. I assume that the battery is effectively shorted while the MOSFET leads (the real battery fuse) are melting, which would protect other inverter units in the cluster. But when the last MOSFET lead lets go, the voltage across the battery terminals to the other units would suddenly increase, and I imagine that there would be an inductive overshoot that would severely test the capacitors and MOSFETs in the other units of the cluster. This spike in voltage might or might not be enough to take out some or all of the other units.

Secondly, if the IGBTs fail, then the 400V bus is shorted, which might or might not take out the MOSFETs of the same unit, depending largely on the health of the battery side capacitors (and probably also the health of the bus capacitors as well). This would happen via a surge through the high frequency transformer coupling the bus and battery circuits. This would then propagate or not through to the other units, as above.

Does anyone have first hand experience of this?

Post by **Tejota** » Thu, 08 Mar 2018, 04:07

coulomb wrote: ↑
Wed, 07 Mar 2018, 06:23
Well sleuthed! Maybe an optic fibre camera next?

coulomb wrote: ↑
Wed, 07 Mar 2018, 06:23
Is there a special firmware revision number? I imagine they would want to distinguish the high voltage versions from others, yet indicate what other features and/or bug fixes it has in common with the nearest non high voltage version.

coulomb wrote: ↑
Wed, 07 Mar 2018, 06:23
Does anyone have first hand experience of this?

Yes. I lost 4 (four) pip4048ms on cluster. Full disaster. I dont know what happened....
Working conditions: 1 pip4048 (master) manufactured August 16 (no heatsink on top) + 3 Axpert 5048 (4kW) (slaves) manufactured Mars 2014 (heatsink on top). Firmware 72.70c on all units. Lead Acid Battery 1500 Ah. At midday global solar production about 9 kW 160A (40+40+40+40 more less) to battery and a little inductives load in AC (2 kW). A little cloud effect.... not important. Suddenly ALL DC breakers of slave units OPEN. Only DC breaker of master closed but all units in fault.
I disconnect everything. Two slaves shorted at battery side. One slave starts at fault 09 and shutdown. Master starts at fault 09 and shutdown.
I only have dissambled one shorted slave: 12 mosfets shorted, 2 mosfets fifty fifty, and 2 mosfets correct. Buck IGBT (2) shorted !!!
No visual destructive effects.
Any explanation of this disaster ?? Mosfets and capacitor were original on failed units.

Pd: Yes I bought an axpert 5048 pf1 64V until I repair failed units if I can.

AEVA Forums