PIP-4048MS and PIP-5048MS inverters

[Enviroweb]

Yes it will be free for personal use,
I will just have to tidy up the code and put together install instructions

[reecho]

Very nice..!!

[paulvk]

"
auto switches to battery \ Utility mode for peak/offpeak times"

Have to think about the durability of the memory in the inverter for where it stores its settings

[Enviroweb]

The software script manages the switching by sending commands to the inv.

cheers

[coulomb]

The software script manages the switching by sending commands to the inv.

Yes, but it's not clear to me whether sending such commands causes extra writes to the EEPROM chip that saves parameters. Certainly commands such as POP (setting Output source Priority) and PCP (setting Charger Priority) override settings read from the EEPROM. There is some evidence that settings are updated continuously, but it seems likely that if so the EEPROM device checks for whether changes have been made before actual flash writing.

Most EEPROM or flash devices have a write limit; writing more than a certain number of times may wear the device out and eventually cause errors.

[Enviroweb]

Im finding out from technical support if it writes to EEPROM,
worst case 35 writes per day I'm hoping to expect 7~8yrs before issues start fingers crossed it lasts that long

[offgridQLD]

I have a potential project that could require 3 phase power offgrid.

I could end up with a industrial Panasonic 6 acces Robotic welding machine later this week robot welder





I already do a lot of manual welding up to 250A from single phase offgrid though this is stepping it up a little. Actually I use One PIP to power my tig welder for some time now and it's is ok up to 175A for reasonably long periods ( as in I get worn out befor the pip has issues) 200A has it maxed out pulling 4000w. Over all hit handles the messy loads of pulse welding in AC well.

The welding unit on the Panasonic is 350A max but I'm not fabricating bridge structures or earth moving equipment. so all my work will be sub 200A. predominantly around the 100A range on thin wall tubing. ( always under 4000w load)

Funny how things get out of hand so fast. How can I keep a serious face when I tell me daughter to switch her 12w led bedroom light off when she isn't in there as she is wasting battery power when dad is running a production line in the garage

I have 3 phase power available to the shed from my diesel Kubota generator (that I never use as it's not very green, cost defective or pleasant to listen to) though it will be ok for initial setup and trials it's the last thing I want long term.

So I have one PIP4048 now and would need two more to give me 12kw and 3phase for the shed.

Anyone working with PIPS in a 3 phase application?

I will report back on the thread as the story unfolds.

[antiscab]

does that robotic welder require 3-phase internally?
or does it just rectify it out to a 600V dc bus?

in the later, it may be more cost effective to power it from a 600v battery

[solamahn]

[offgridQLD]

I can't say for sure. Though I can't see how a 600v battery of reasonable capacity would be less expensive than two extra PIP4048's at about $1000 each to add to the PIP I already to create three phase AC power.

+ the three phase power is handy for other things.

[offgridQLD]

Nice one Solarmahn. What 3 phase loads are the 3 pips in that photo responsible for. Refrigeration?

[solamahn]

It's actually running in single phase now. Had some problems balancing the loads with the 3 phase plus they are using a single phase generator for backup. It runs an office block and a workshop

[Adverse Effects]

why not get a SINGLE PHASE TO 3 THREE PHASE CONVERTER that way you can just run 1 phase and use this to run anything that needs 3 phase
[tube]https://www.youtube.com/watch?time_cont ... LqIAPP3G6A[/tube]

[offgridQLD]

That's is a interesting way around things.

Just in my head the machien's max rating is 350A at 36v from memory 12,600w. Though like my other welders you don't have to run them at full capacity. 150A is 5400w, 100A - 3600w.

To answer the question about 600v dc bus.
The servo motors on the robot arm are AC.

Considering I all ready have one PIP4048 to play with that is just dedicated to powering the now. I think purchasing 2 additional PIPs at $2000 would be the most cost effective way to get 3 phase and have a respectable 12kw output 24kw surge rating. + the extra 9kw of pv charging potential they offer as a set of three is always a nice to have down the track.

Though the issues that solamahn had with load balancing sounds concerning.

Kurt

[lopezjm2001]

Does someone know under exactly what conditions the PIP4048MS will change from CV(constant voltage) to float charge?
Thanks.

[coulomb]

Does someone know under exactly what conditions the PIP4048MS will change from CV(constant voltage) to float charge?

I suspect I know more about that question than most. This is a place where Weber and I believe there to be a bug in the official firmware; certainly for us, when charging a LiFePO4 battery, changing this part of the algorithm seems to work much better. See more on that below.

There is a global flag, g_uwChgFloatFlg, which controls the change to float (internally, wInvChgStatus is set to 13; when this variable is 12 it's still in bulk/absorb mode). When this flag is set and it's in bulk/absorb mode, counter wFloatChgCnt$1 is incremented; when it reaches 1500 (in 50 seconds; this code seems to get called 30 times per second), that counter is cleared and wInvChgStatus is set to 13 (i.e. the charger goes to float mode). If at any time during that 30 seconds g_uwChgFloatFlg is cleared, the counter is reset to 30. So g_uwChgFloatFlg has to be set for 30 consecutive seconds to go to float mode.

So it's all about when g_uwChgFloatFlg is set. The only place seems to be when the total charge current is less than the following: the maximum of either 5 A per paralleled unit, and (g_uwSysMaxChgCurrSum / 5.0). I don't recall exactly, but I think that g_uwSysMaxChgCurrSum must be the sum of the maximum charge current settings for all paralleled units. Whenever I've been watching the value of these variables and charging from solar, it seemed to be comparing the charge current with one fifth of the maximum charge current setting, or 5 amps, whichever is the higher.

The idea seems to be that when you've been in absorb long enough, then the charge current will fall to either a fifth of the maximum, or to 5 amps (10 amps for two paralleled units, and so on). That's all well and good, but at the start of the day, having discharged the battery overnight, the battery may well need a charge, but soon after dawn the solar charge current could be quite low. It could easily be below these thresholds for 30 seconds. So part of the LiFePO4 firmware patch is to add an additional criterion: the battery voltage has to be no more than half a volt lower than the CV voltage setting [require batt_volts >= (CV_setting - 0.5)]. So in the early morning scenario, this is not met, so the charger stays charging until both the charge current is low, and the battery voltage is at or near the CV target voltage.

[ Edit2: Even if charging from the mains, which is not subject to a lack of sun in the morning, the charging current could still be ramping up to its ultimate value when the 30 seconds are up. So often even the mains would not charge the battery for more than briefly. ]

As part of the LiFePO4 patch, we also set the current limit lower: to the maximum of one amp and the maximum charge current setting divided by 30. So that makes it harder to exit bulk/absorb and enter float.

[ Edit 1: the LiFePO4 patch also changes the 30 second "qualifying period" from 30 seconds to two minutes. ] [ Edit 1a: I forgot that while I intended to change this value, it somehow got dropped, so the qualifying time remains, for now, at 50 seconds. ]

Without the LiFePO4 patch, we found that the PIP would exit to float mode way too soon, and the battery would never get properly charged from solar [ edit: or even from mains].

[lopezjm2001]

As part of the LiFePO4 patch, we also set the current limit lower: to the maximum of one amp and the maximum charge current setting divided by 30. So that makes it harder to exit bulk/float and enter float

Thanks for your detailed answer.

I am using the Lifepo4 patch and I assume from your quote above that the PIP4048MS charger will not go into float mode until the charger current lowers to 1 amp.

I have just started using the charger part of my PIP4048MS today, it seemed to go on for hours in CV mode with a charge current of about 4 amps and it never went into float mode. Then sundown came.

I have three 80ah 48V lifepo4 (using pouch cells) battery packs connected in parallel. Each battery pack has a centralised mini-BMS using top balancing of 0.75 amps maximum, so 2.25 amps total maximum. Also current is drawn by two 48V DC converters that supply 12Vdc to each BMS and two DC contactor coils.

I suspect that all this current draw that is seen by the charger as charge current is preventing the PIP408MS charger from going into float mode.

Is there a "get out of gaol card" that will allow charger to enter float mode, no matter what?

[offgridQLD]

The way I see it is there has to be a overriding time value for absorb. For example 2 hours. Typically you have a minimum amp charge rate value that needs to be met and maintained for a short countdown time (say 30 seconds) .

The 2 hr timer avoids the situation of live loads being above the min charge amps required to trigger float. This method is still crude though as often live loads (or base load) need to be guessed and added to the min float trigger amp setting .

The Best solution that I am using in my system is a adjustable minimum float trigger charge amps. Though the charge amps that the charge controller reads are the true net charge amps to the battery . As it is monitored through a shunt by the controller. This way all live loads on the battery are factored in when float trigger min amps are triggered.

I still have a max time setting ( though it is just as a back up and in practice is never used)

With the PIP your going to have to have vertualy no live loads on the battery at absorb stage to trigger float via the min charge current setting (1amp 50watts) so it's always going to by triggered by the count down timer.

Kurt

[coulomb]

I am using the Lifepo4 patch and I assume from your quote above that the PIP4048MS charger will not go into float mode until the charger current lowers to 1 amp.

It depends on your maximum charge current setting, parameter 02. If that's 120, for example, then the limit will be 4 A, since max(4, 1) = 4.

I have just started using the charger part of my PIP4048MS today, it seemed to go on for hours in CV mode with a charge current of about 4 amps and it never went into float mode. Then sundown came.

It sounds like the charge limit constants (1 and 30) may need some tweaking. We were hoping to be using our PIPs daily by now, even long ago, but many things have gotten in the way.

I suspect that all this current draw that is seen by the charger as charge current is preventing the PIP408MS charger from going into float mode.

I can't remember now, but I think that the inverter can cancel out current used by the inverter. [ Edit: that is, it can figure out net charge current into the battery, not just using solar or mains charge current. ] But of course, if you have other loads on the battery that the PIP can't see, it can't take those into consideration.

[ Edit: reworded. Sounded like the PIP could magically cancel out pesky current drains, perpetual motion like. ]

[offgridQLD]

" I think that the inverter can cancel out current used by the inverter. "

That's interesting and great if that's so. I will experiment with that on the weekend.

Kurt

[lopezjm2001]

It depends on your maximum charge current setting, parameter 02. If that's 120, for example, then the limit will be 4 A, since max(4, 1) = 4.

Just checked and mine was set to 60 amps. I have now changed it to 120 amps. See how we go later on today. It's a bit cloudy today so it may not happen.

Thanks.

[ Edit: the LiFePO4 patch also changes the 30 second "qualifying period" from 30 seconds to two minutes. ]

Just wanted to know if this 120 second timer is cumulative or does it reset to zero if it stops qualifying at any time during the count. Thanks.

[coulomb]

Just wanted to know if this 120 second timer is cumulative or does it reset to zero if it stops qualifying at any time during the count. Thanks.

It resets the counter to zero if the "go to float" global is reset. So yes, it starts again if the charge current doesn't go low enough, even briefly. So you need a situation where the current falls below the threshold for two consecutive minutes, with no exceptions. So if it's not quite made the criterion, it doesn't count and continues CV charging until it makes the criterion decisively.

[weber]

Coulomb, according to Edit 10 in the LiFePO₄ patch post
viewtopic.php?title=pip4048ms-inverter& ... 332#p59891,
with a single PIP (no parallel PIPs), the PIP has to measure both battery voltage ≥ (Param26 - 0.5) volts and battery current < Max(1, Param2 / 30) amps, continuously for only 50 seconds (not 2 minutes), before it will go to float. And 50 seconds is what it was originally, not 30 seconds.

Of course, current going directly from the solar charge controller subsystem to the inverter subsystem doesn't count for this. That's not battery current. The PIP measures the battery current at its battery terminals as you'd expect. It appears to use a short length of solder-reinforced printed circuit board track as a shunt, to do this, which would explain its poor accuracy and only 1 amp resolution.

So with parameter 2 set to 120 A, the charge current as measured by the PIP must show as 3 amps or less for 50 seconds continuously, to go to float.

We've had a PIP operating with the LiFePO₄ patch for several months now, as the sole source of power for an off-grid household, and the cells are happy and remaining balanced.

The PIP can't cater for all possible non-PIP DC loads that anyone might use. Otherwise, in cases where such loads are minimal, it would end up going to float too soon, and therefore not balancing the cells.

But it does need to allow for BMS bypass current and the power needed for the BMS master and to hold the battery contactors on. That wouldn't normally amount to more than 3 amps (150 watts).

Lopezjm2001, you have an unusually high bypass current of 2.25 A because of your 3 parallel strings. But the BMS master current would usually be less than 0.1 A. So that still leaves about 0.65 amps in the budget for contactors. These typically (EV200 or GX14) pull a large current for a short time when turning on, but use less than 0.2 A each at 12 V to hold. If there were two of them and your DC-DC was only 70% efficient, that would still only be 0.15 A at 50 V.

So it should still come in under 3 A and therefore the existing LiFePO₄ patch should work now that you've bumped parameter 2 up to 120 A.

What kind of contactors do you have?
What voltages do you have for parameters 26 and 27 (absorb and float)?
At what cell voltage does your BMS start to bypass?

[offgridQLD]

"Of course, current going directly from the solar charge controller subsystem to the inverter subsystem doesn't count for this. That's not battery current. The PIP measures the battery current at its battery terminals as you'd expect. It appears to use a short length of solder-reinforced printed circuit board track as a shunt, to do this, which would explain its poor accuracy and only 1 amp resolution."

Thanks for confirming this. That makes things a lot better than I was first thinking and thanks for clearing up the countdown time to float to.

In my system Now that I have my PIP wired in permanently to power my workshop. I don't see why I shouldn't put the PIP's PV charger to some use. I have six spare 2nd hand 200w panels collecting dust 1200w. Though the issue is I have a lot of DC loads the PIP will not be accounting for (The big one another inverter then two charge controllers and a lithium BMS system) That said I don't need to count on it for final absorb or cell balancing as the other two charge controllers can handle that.

So I was thinking perhaps setting the absorb voltage on the PIP lower than the other two controllers. That way it can assist in the bulk stage of the charge in the morning though it will tap out early as the voltage gets just shy of the higher absorb setting on the other two controllers. Then it can assist again later in float covering loads and maintaining float voltage.

Perhaps have the 1200w facing East or NE to let it contribute more in the early stage of the day.

[weber]

Sounds like a good plan, Kurt.