PIP-4048MS and PIP-5048MS inverters

[Monkeytom]

Pulled my two PIP4048HS (PWM SCC) apart yesterday and turned the fans over and installed the Parallel boards. Here is a pic of the inside with the PWM SCC, Mounted them and connected them up to the LifePO4 15S2P I need some more battery straps to go 3P, all good and very impressed, the slave takes about 8 seconds to come on line after being switched on, So I am going to use my digital current switch to sense load and switch on the second inverter when the load gets to about 2kw and save some idle current.
Don't be concerned with the cell voltage, I had an external battery charger on to test the BMS and Cell balance and it all Works Great, cells were within .001 Volts of each other at the top after a bottom balance. I have the charging set to 55.5 and float at 53.5 so the cells are at 3.5 Volts. Well did the smoke test today. With base load 400w + washing machine + dryer + toaster + electric Jug + oven and two thumbs up 7Kw on my wireless energy meter and all good.

[Monkeytom]

Hi All,

I have searched - but could not find any particular reference to this.

Has anyone ever run the PIP-4048 with Li-Ion cells (Tesla style) - in something such as a 14S configuration?.
Any thoughts?

Edmund

Check this out make your own

https://www.youtube.com/watch?v=JQa5gn-7D74

[Monkeytom]

Just a Quick update with my Parallel PIPs. All good
After the inverters are wired and set to Parallel mode, which ever inverter is turned on first becomes the Master, once the master is active and the Slave is turned on it takes about 8 seconds to come on line and share the load.
If the Slave is turned off the AC output stops straight away and the inverter takes about 15 seconds to shut down.
Once both inverters are switched on they stay on and they draw about 1/2 Amp each on no load or 25 watts each.
What I have done is change the jumpers on the comms board so I lose the Gen start feature on the Slave inverter and it becomes a remote on/off switch. I wired to the N.C./N.O. terminals. Note there is battery voltage 48v on the on/off switch and remote terminals.
I have mounted a digital current switch (a current sensing relay with an adjustable set point)and run the inverter output active wires thru its toroid pickup. I have the setpoint at about 2300 watts and the slave inverter then switches on. This can be over ridden by the normal on switch but it would need an extra switch in the current switch line to over ride the off,(I will do that). I am happy to report it works nicely.
I was thinking in a year or so I will reverse the Master/slave and share the Hours around
One hiccup I had,
I had each inverter AC output wired to a separate breaker.
on testing I had the Master inverter breaker closed and the Slave breaker open and both fired up OK but when I switched on a load (the Dryer) both inverters went into fault.
Both inverter AC outputs need to be wired together thru a single larger Breaker.

[weber]

Well done, Monkeytom. What voltage did you bottom-balance at, and how did you do it? And what voltage were they at when they were within 0.001 V at the top?

I understand that with the EV-Power CMUs you need to take them up to 3.7 V pc to guarantee top balancing, hence the absorb voltage of 55.5 V, but I don't see any need for your float voltage to be so high. It can be as low as 3.36 V per cell for LiFePO4, i.e. 50.4 V for 15 cells. This will avoid the premature aging caused by holding them at a higher voltage.

Are you using Coulomb's patched firmware? If not, have you ever seen them actually go up to the absorb voltage, using PV input alone?

[Edit: I just realised you have the 4048HS, not the 4048MS, so Coulomb's patched firmware would not work, and it probably doesn't have that no-absorb bug in the first place.]

[Northland]

For those considering the coulometer from a few pages back, I have one. Its been monitoring my lifepo4 for around 7 months now and appears to be very accurate. I asked for a 150a shunt and they were very obliging.

Just received mine. Seems to work well. Except....I can't seem to access the menu. Instructions say to hold all 3 buttons for a few seconds after power up. Well I've tried everything and can't get that menu. So for now it's stuck on 100ah. Any ideas?

Here's link to 100A version with better shunt
DC80V/100A Battery Indicator Battery Voltage/Current/Power/Capacity/Charging Time Tester/Monitor Battery Coulometer
http://s.aliexpress.com/Uvq6Bfaq

Edit: disconnected from 48v, connected to 12v battery. Got the menu up after many attempts. Changed to 400ah. Can also change shunt resistance. Unable to replicate the process. Seems like having a charger or load connected inhibits the ability to access the menu.

Since the switch cable is 6 core I might use the other 4 wires for this meter (48v + 2 shunt). Will make a panel to house the remote and coulumb meter. Also thinking of stealing the display from the pip so will end up with 3 displays in the panel. Really want to see pv watts and load watts without clicking buttons. Could be tricky.

[Monkeytom]

Thanks Weber your detective work helped a lot with the remote switch.
I bought 43 used 90Ah sky cells they are 6 years old but I was told they had only been cycled about 20 times they seem good so far with one a little swollen and not the same as the others.
I bottom Balanced at 2.80 V I brought the pack down on the inverter and then found it very easy and quick to move the voltage of each cell around with an Ebay purchase a battery load tester it has a coat hanger type element and draws 50Amps after burning out the switch on the first few uses, I replaced it with a made in USA switch from my parts junk box and it now works great. I have been using a 5 Amp LiPo charger to bring them up to about 2.8 then I put the whole pack in parallel for a day and left it sit at about 2.8v 1S30P, then I assembled the pack 15S2P and mounted a BMS module on every second cell
I was running the inverter on my LA Pack in the mean time.
I bought a used set of 45 CM90 EV Power BMS modules and found one dead one the others tested ok. I am Planning on using two sets of BMS one set on all the odd cells 1,3,5,7,9. etc and a second string on all the even cells and will have built in redundancy I believe the BMS is very important.
My idea using a SSR to disconnect the solar was a fizzer, and I burnt my finger. The SSR was mounted on a heatsink and I had 20 Amps from the solar panels to the PIP thru the SSR, The BMS Module switches balancing on and a red LED at 3,7 as you mention and the green LED goes off at 4.0 and the N.C. daisy chain goes open which was wired to the control of the SSR it turned off the red LED on the SSR good but the SSR did not switch the solar off and got very hot very quick as I found when I touched the heat sink (bacon smell and taste)so I need a contactor or some made in USA SSR's or something else....
The cells were within .001 of each other the next morning after sitting the night after the BMS balance at the Top.
Yes these are the HS version inverters with the Pulse Width Modulation SCC,there was two reasons I bought them, first was they were cheaper and well under the GST free $1000 limit where the MPPT were well over it and would attract another 10%. Reason two was after reading as much as I could on the PIPs. I thought there may have been a problem with cloud edge effect on the stand alone 60Amp MPPT SCC and thought the PIP MS used the same SCC. And with enough panels, I should be charged by mid morning most sunny days anyway, so i spent the saved money on another 1.7Kw of Thin film for $319.
No I am not using Coulomb's patched software, and Thanks I will drop my float down to 50.4 as you mention as the higher voltage is long term trouble, No I haven't seen them go to absorb voltage on PV yet, hoping tomorrow as I have been fiddling with stuff today and had the solar switched off most of the morning.
All in all I am Super happy with this setup. I have done it on the cheap it now owes me $6050.00 all up, so an 8kw inverter system with 6Kw solar and usable 7Kwh LiFePO4 Battery with another 15 cells needing battery straps (probably have to replace one cell). and with my solar FIT it should be paid for in three years? fingers crossed. And to top it off I sold my Lead Acid pack on the weekend.

[weber]

Sounds like you need one of these
http://www.evworks.com.au/gigavac-minitactor-50amp

[Northland]

So a while back I asked how I could have a bypass switch to quickly go to utility and back. I have a remote unit and just took it apart to see what the switch at the bottom is connected to



seems the switch is normally open with 2 wires connected. It goes via 2 tracks to the "switch in" connector where only 2 of the 6 contacts are used. Question is, can I open the pip,locate the phase contactor of the 230v input, break the circuit track (12v) and connect that to the remote switch? Anyone see a problem? It would require "utility" mode instead of "sbu"

[coulomb]

Question is, can I open the pip,locate the phase contactor of the 230v input, break the circuit track (12v) and connect that to the remote switch? Anyone see a problem?

I'm quite uneasy about this idea, though I can't pinpoint a concrete reason just now.

The firmware may throw an error, but if so, I guess you haven't lost much other than some effort to get to the 230 V relay(s).

[Monkeytom]

Thanks Weber I will get a couple of these maybe three and do three separate strings 15S1P two on Anderson plugs.{so I can use 30 in the car}
I followed the PIP charging today all good to Absorb at 55.5V although I will raise it to 55.8 as I measured the voltage at the battery terminals and it was 0.35 V lower than shown on the PIP, voltage drop on the cable I guess. Most of the BMS were in balance mode. The PIP then dropped back to float now at 50.5v All Good.
I have relative short cables they are 2AWG 35mm2 1.2 meters long x 2 they should be big enough.

[weber]

I was actually suggesting the Minitactors for disconnecting PV array, since they are only good for 50 amps and have tiny terminals. I'd usually recommend one of the larger Gigavacs, GV200 or GX14, for batteries.

Yes, your battery cables are big enough, and short. I'll bet most of that voltage drop is across that crazy arrangement that is required by the EV-Power BMS installation manual, because the EV-Power CMUs have no reverse voltage protection, and people can't be trusted to disconnect all loads while installing or removing CMUs or battery cables.

The crazy arrangement I'm talking about is the one where you have to run the entire battery current through an M8 stainless-steel stud at each end of the battery. Stainless-steel has only 1/40th of the conductivity of copper. So when you consider the thread depth, that's like 1 mm dia copper! You could check with a multimeter, how much voltage is dropped across those.

[offgridQLD]

I noticed that in the instructions to Weber. Though I didn't use studs on my bank for that reason. I want my lugs clamped fast to the battery terminal face not conducting through the threads and stud.

Its the golden rule when welding never try and earth through a threaded component or a component that isn't a solid part of what your welding.

Kurt

[Monkeytom]

Thanks Weber I was thinking two for the solar one on each PIP, and one big one for battery disconnect. I was thinking of Hard wiring one pack and the other two on Anderson connectors the big ones 135A I saw some where.
I will Give Tim at EV Works a call next week he has been very helpful.
Thanks for the tip on the stainless conductivity or lack there of,
These cells came with studs with plated copper straps flat, spring and Half nuts. I have the BMS on top of the half nuts with another Half nut on top of them.
I was thinking of strapping the cells together in groups of 5 cells using a steel banding packing system [old school] I can get some 12mm stainless strap I saw on ebay, 5cells would not be too heavy to shift if needed.
I was doing some numbers on the power saving not having Both PIPs running all the time, I have estimated that the second is on for less than an hour a day, say 24 watts in idle saved for 23 hours a day
works out at over 200Kwh a year saving not bad. shows how the small loads over a long period sure add up.

[paulvk]

I have anderson connectors on my inverters 50amp on the panels and 175 amp on the battery I also have plugs and sockets on the mains so that in 10 min I can replace the units with a spare. I got the anderson connectors from here http://www.larpro.com.au/

I now have the inverters running in parallel with 72.40 software so now I can write the instructions into the AVR web server for parallel units.

[solamahn]

I got some 4048msd and 4048mst

[Northland]

OK so a review of the coulumb meter http://s.aliexpress.com/Uvq6Bfaq

So I had the pip remote next to the new meter. I was wasting a lot of time clicking to the screen I wanted on the pip.
The new meter is more accurate than the pip. It is faster to respond and updates in real time and has better resolution. The pip can be +/- 1A and also rounds to the nearest A. So on 4.55A it might say 6A. The pip seems to have a 5 second lag and only updates every 3 or 4 seconds.
I noticed I was pressing the pip remote buttons less and less, then not at all. Finally, I just turned it off and unplugged it because I stopped caring about the figures it showed. I may sell it.
soc in % terms and ah terms seems to work perfectly and is far easier to digest.
I now spend 2 or 3 seconds monitoring the system, where I was wasting 30 seconds. Multiply that by 8-12x a day....

[paulvk]

Hello Northland
Yes I think the PIPs displays are very poor I can not understand why they did not use a graphics display module for a mobile phone as these are cheap have touch screens and the display can show anything just by software.
Have not had much time lately but hopefully will get there in the next month or two and have this server working nicely
Note it has also a graphical LCD on which it displays the PIP data including the stuff that the watchpower dose not like heatsink temp and bus voltage.

[solamahn]

Installed a 4048msd with 24 x 280w panels, 12 at 15 degrees due east and 12 at 15 degrees due west with 16 CALB CA400. Set bulk at 54.5, float at 53.5 and lvd at 48. Works really good but I would like the lvd to be about 50v

[coulomb]

You can set the Low Voltage Disconnect to values up to 52.0 V, if you're willing to try my modified firmware on what sounds like brand new hardware.

I meant to contact someone at Voltronic about these issues, so that new machines could come with fixes making them more suitable for lithium ion batteries. But alas I never got around to it.

So it doesn't exit bulk PV charging too soon? If not, I wonder if it's because of your larger 400 Ah battery (we tested on an aged 160 Ah battery).

[coulomb]

Oh, just now I realise that MSD probably means the dual MPPT version. [ Edit: twenty-four panels should have been a small clue ]. I don't think that firmware 72.40 can handle dual MPPT PV chargers. Did you note the firmware version? If it's later than 72.40, then it's not a good idea to use my modified 72.40 firmware. But I think you realised that already, hence the inability to go over 48.0 V LVD.

[solamahn]

First I had it set to 56 and 54 but that seemed too high. With the PV breakers off the resting battery voltage was 53.6. I was aiming for 53.3 so that's why I tried 54.5 and 53.5. I need to make more observations to see how this setting goes. I will also set the generator start to 50 and stop at around 53. Generator charge current is set at 20A so 53 should be ok. Again I need to observe how it goes. I did not note the firmware revision but it's not 72.40

[weber]

Thanks Weber your detective work helped a lot with the remote switch.

I don't think I had anything to do with that. That was down to Coulomb and lopezjm2001.
viewtopic.php?title=pip4048ms-inverter& ... 332#p60733

[Monkeytom]

Thanks Coulomb and Weber and others for all your work and posts on these inverters, It has helped me get mine up and running well,
I have bought some more used batteries last weekend 43 x 100Ah Blue Calb
and will use 30 x 90Ah yellow TS in Parallel with 30 x 100Ah Calb in the Karman Ghia split in two 48v15S Packs to be charged by solar by being hooked up to the house pack which will be the rest 15S of each in Parallel.

[solamahn]

4048msd firmware
U1 75 00
U2 01 01
U3 01 01

[coulomb]

4048msd firmware
U1 75 00

I look forward to reading that firmware one day.

U2 01 01

Assuming that's not a typo for 04 01, that seems to imply that the dual MPPT firmware is brand new, not an evolution of the single MPPT design. (Or just two of the single MPPTs squeezed into one box, as I had guessed.) It will be interesting to see how the new chargers behave with lithium batteries; I would guess much the same. But my guesses have been known to be wrong

Thanks for the info.

Edit: I see the remote display firmware (U3) might be brand new too. Does the system you recorded these version numbers from have a remote display? Does it have the two (and provision for 3) solar panel icons?