PIP-4048MS and PIP-5048MS inverters

[weber]

Is it correct, that I can't disable KettleKomp™?

That is correct. But you can reduce its effect to a minimum, by setting the low cutoff voltage [29] to a value that ends in .0 or .5, and setting the max charge current [2] on all inverters to the maximum safe value.

Are there other differences between 72.20e and 72.20c besides KettleKomp™? If not I simply have to reflash to version 72.20c to disable it?

The only difference between 72.20c and 72.20e is KettleKomp™, which includes the improved SoC reading from the inverter.

I'm unsure if I like KettleKomp™ (they way I understood it) or not. I'll have an eye on the voltages told by the inverter and these told by the batteries.

KettleKomp™ has no effect on the ordinary voltage reading from the inverter. It merely provides a compensated voltage reading in addition to the ordinary voltage reading. The compensated reading is only used for 3 purposes: comparison with the low cutoff voltage and associated warning voltages, comparison with the back-to-utility voltage, and calculation of the crude state-of-charge reading. This compensated voltage reading is never returned by a QPIGS or QPGS command, and you know if you're looking at it on the LCD because of the blinking "k".

For my installation with 5 (yet) Pylontech US2000b I use the voltage settings suggested here.
What would be the suggested cut-off voltage for 15s LFP batteries (without KettleKomp™)?
It was 47.5V in the original firmware. Is this a good value?

47.5 V is a good cut-off voltage for 15s LFP cells. Without KettleKomp™ I'd expect there to be nuisance cutoffs due to short-term heavy loads with that setting. If so, you could set it 1 or 2 volts lower.

[teba]

This compensated voltage reading is never returned by a QPIGS or QPGS command, and you know if you're looking at it on the LCD because of the blinking "k".

Thanks for this hint. I'll enhance my loggin to also log QBV.

47.5 V is a good cut-off voltage for 15s LFP cells. Without KettleKomp™ I'd expect there to be nuisance cutoffs due to short-term heavy loads with that setting. If so, you could set it 1 or 2 volts lower.

Thanks. But what about the statement that there has to be a difference of at least 1,5V to setting 12 in this topic? Is this only correct for the type of inverter discussed there? Or is it because of the missing KettleKomp™?
I've put 48V to setting 12.

@coulomb: I see a smaller difference between apparent and real power at higher loads. Maybe this has also to do with it?

Thanks again! You both are great!

Greetings
Lars

[coulomb]

Thanks. But what about the statement that there has to be a difference of at least 1,5V to setting 12 in this topic?

Ah. That's only for non-LFP flavoured patched firmware, or non-patched firmware. You're running (I assume) LFP flavoured patched firmware, so that doesn't apply. It's a bit of a long story as to why 1.5 V might sometimes be useful, as opposed to 2.0 V, and I try to keep that table as simple as possible (even so, it still looks very cluttered to me).

[Kurnol]

Evening all,

I have spent the last 2 days trying to find an answer to a question online as well as going through this whole thread, I’m confident one of you guys would be able to help me out.

My set up, two 4048’s in parallel, 6kw of panels, two banks 8 x 6v 200AH AGMs.

I have to replace the batteries after 2 years (the old adage a little bit of knowledge) I had the charging set to 56.4v bulk and 54v float. I realize now I had set them to a back up type system rather than an off grid cycling system.

The set up is on our weekend bush place, so the system was cycling maybe 3 weekends out of 4 plus holidays and we also used to leave a fridge on.

If we hadn’t left the fridge on would the system be classed as a cycling one or back up.

The cycling voltage recommends a max of 58.4v so I’m assuming when I replace the batteries I set the bulk to 58.4v (assuming battery spec the same), would the float be set to 54v or higher?

Cheers

[solamahn]

I use 58.4 for bulk and 56 for float in cold weather. 55 in warm weather and 54 when away. You should have your 2 batteries in parallel and then series 2P8S. Also check for battery cell voltage balance at full every now and then. Also set LVD at 48 and back to grid at 49. Back to battery 53

[solamahn]

Sorry Kurnol. I just re read your post. You have no grid backup. Set LVD to 46 if using factory firmware.

[Kurnol]

I use 58.4 for bulk and 56 for float in cold weather. 55 in warm weather and 54 when away. You should have your 2 batteries in parallel and then series 2P8S. Also check for battery cell voltage balance at full every now and then. Also set LVD at 48 and back to grid at 49. Back to battery 53

Hi Solamahn,

System is 2P8S, thanks for your settings, I have been dropping the float when we leave, I think I’ll turn the fridge off also to reduce the amounts of cycles.

I do have a manual back up generator which I have been having to run more and more lately. What voltage would be 60% dod?

Cheers

[martin_f]

Hello all,

this is my first post and I want to begin with appreciating the great work of coulomb and weber allowing me to create a cheap LiFePo4 isle system which is just working fine.

I started experimenting 2 years ago with a PIP-5048MS, 15x300W modules and 4x12V lead acid camping batteries and spent 1 year of upgrading the system to what I have now...

Here is a short overview of my system (running fine for a year now):

I have 4 PIP-5048MS (patched firmware 72.20e) working in parallel mode charging 8 LiFePo4 15S Pylontech US2000b+ batterie modules (17.6 KWh, constant voltage 52.7V, float voltage 51.7V) (no dynamic charge control yet).
The inverters are connected to 4 PV-arrays of sizes 6.3, 5.4, 3.6 and 3.6 KWp.
Furthermore 2 of 4 inverters are connected to grid-power to be able to bypass the isle system when the batterie's SOC gets below 20%.
All 4 inverters power the house, a Tesla Model 3 and crypto miners creating heat to warm up the basement , if there is too much energy provided by the PV modules.
Several relais controlled by a raspberrypi help to cut off the PV-arrays and disable/enable the "active" switch of the inverters to shut them down or bring them up.
A state model programmed in python decides how much inverters are active depending on the load or when to shut down inverter 3 and 4 to make inverter 1 and 2 switch to grid when batterie SOC gets below 20%.
For this "controlling" the batteries and inverters are read out every 5 seconds by rs232 and rs485.

I am very happy with what I have create so far. Sometimes the fuses between the inverters and batterie might trigger, when the load increases dramatically and the bringup of more inverters takes too long. The same sometimes happens when the PV-arrays produce low power or for any strange reason I do not understand the SCC suddenly reconnects to the PV-array shutting down PV power completley for a few seconds (no settings have been changed). I think I can fix the fuse issue by utilizing more expensive ones. The fuses that are currently applied should trigger at 120A (25mm² cables) but they already trigger at 70A - 80A because of thermal heat generated at the contactors of the fuses...

However the reason for my post should address the RF noise generated by the PIP-5048 inverters.
I read this post and I can confirm that I have much noise in the range from 6Mhz to 18MHz and higher disturbing my DSL signal.

Well I have had the two PIP 5048MS inverters installed a couple of weeks now, and have noticed two interesting things. First the mppt solar charge controllers do not input nearly as much power to the battery as the old PIP 4048HS units. My solar panels are running about 70 volt dc in. And second to that and more importantly to me the PIP 5048MS are taking out my ham radio, they are very dirty RF wise. And generate lotsa QRM. So I will go back to the PIP4048HS charge controllers and see if what I think is happening is? Anybody else experience this? The noise is particularly bad on 6 meters my favorite band. Tom

See my picture to get an idea of how the DSL signal is disturbed by the inverters.
The upper diagram shows the noise margin of the signal at different frequencies. The higher the yellow and purple margin the better.
The orange line shows the maximum and minimum margin over a whole sunny day. At night the orange noise-margin is at maximum and the spectrum displays smooth blocks.
But at day the blocks transform to a very curly shape starting dramatically from 6MHz up to the end of the diagram.
The minimum orange line is reached when the PV-arrays produce most power. At this point, my DSL signal only achieves about 70 MBit/s from about 120 MBit/s maximum.

dsl_spectrum.PNG (158.19 KiB) Viewed 1726 times

I noticed this last week, when I moved the location of my internet modem/router direclty to the telephone wallbox of the house to get the maximum broadband speed.
The router is about 5 meters away from the inverters. Cutting the PV-arrays off completely eliminates the noise and the DSL signal is perfect.
Most of the noise is coupled by an ethernet cable that runs parallel with the PV strings for a few meters. But there is still some radiation noise the affects the router when the ethernet cable is disconnected.

Obviously the switching of high currents within the inverters leads to RF noise that is escaping the inverters using the PV strings as antennas.

So I am trying to reach out to you if anyone else is fighting with RF problems and might have some hints to compensate this problem.

Next week I will try to put ferrites around the string cables.

Regards,
Martin

[solamahn]

Hi Solamahn,

System is 2P8S, thanks for your settings, I have been dropping the float when we leave, I think I’ll turn the fridge off also to reduce the amounts of cycles.

The fridge would not normally be a problem with that size system. Your battery bank is small at 400Ah so you would have to be modest with your energy use at night. The 56.4 bulk and 54 float might not have caused your batteries to fail prematurely. It might have been because the batteries have been getting over discharged. What LVD setting did you use. It is a bit hard to work out 60% DOD using voltage as an indicator on a working system. Coulomb and Weber’s excellent kettlecomp patched firmware would help. You need to make sure your panels are providing enough charge to get the batteries full during the day. What ocv are you using for the panels? You could try saving the batteries by charging each 6v pair individually to 8v but watching them so you can disconnect the charger if they vent. Best setup would be to add 600w of panels to each inverter so total of 7,200w, 16 x 400Ah Lifepo4 and patched firmware.

[ Edited Coulomb: Added quote for some context. ]

[paulvk]

Voltages of 56.4 bulk and 54 float are close to what I have had my flooded set to and they are 6 years old now
they are showing their age and I only have 8 x T105 225Ah so half of your banks with two 4048s
The specs for flooded trojans T105 is for equalization at 62v periodically the pips can not do that
and this is why the newer inverters have 64v after I sent MPP Solar the specs.
Sealed batteries can not go that high so the way for them is pulse charging to 62v so gases have time to recombine.
You could also charge higher by disconnecting one bank and charging it I do this in my big system with 11 banks of 100Ah VRLA
pulse charging each bank they are now 9 years old (3 years when I got them) .
But the big problem with the sealed ones is low current availability for example your limited to 40 amps so if you have been
going beyond this then they will fail early by drying out.

[Kurnol]

Hi Solamahn,

System is 2P8S, thanks for your settings, I have been dropping the float when we leave, I think I’ll turn the fridge off also to reduce the amounts of cycles.

The fridge would not normally be a problem with that size system. Your battery bank is small at 400Ah so you would have to be modest with your energy use at night. The 56.4 bulk and 54 float might not have caused your batteries to fail prematurely. It might have been because the batteries have been getting over discharged. What LVD setting did you use. It is a bit hard to work out 60% DOD using voltage as an indicator on a working system. Coulomb and Weber’s excellent kettlecomp patched firmware would help. You need to make sure your panels are providing enough charge to get the batteries full during the day. What ocv are you using for the panels? You could try saving the batteries by charging each 6v pair individually to 8v but watching them so you can disconnect the charger if they vent. Best setup would be to add 600w of panels to each inverter so total of 7,200w, 16 x 400Ah Lifepo4 and patched firmware.

[ Edited Coulomb: Added quote for some context. ]

LVD was 46v, the OCV of the panels was either 38v - 39v. 250w panels in banks of 3. The battery bank was just over 19kwh I figured 6 or 7 kWh should be enough for overnight. The batteries always seemed to be sitting on float (54v) by the end of the day with very little current going to them so I assumed they were charged. They did always drop to 51v as soon as the sun went down.
I was guessing I ruined the batteries by under charging and then over discharging. All the panels face north with no shading.

[Kurnol]

Voltages of 56.4 bulk and 54 float are close to what I have had my flooded set to and they are 6 years old now
they are showing their age and I only have 8 x T105 225Ah so half of your banks with two 4048s
The specs for flooded trojans T105 is for equalization at 62v periodically the pips can not do that
and this is why the newer inverters have 64v after I sent MPP Solar the specs.
Sealed batteries can not go that high so the way for them is pulse charging to 62v so gases have time to recombine.
You could also charge higher by disconnecting one bank and charging it I do this in my big system with 11 banks of 100Ah VRLA
pulse charging each bank they are now 9 years old (3 years when I got them) .
But the big problem with the sealed ones is low current availability for example your limited to 40 amps so if you have been
going beyond this then they will fail early by drying out.

Gday Paul are your batteries cycling every night or are they a back up system. What do you mean by current availability?

[paulvk]

are your batteries cycling every night or are they a back up system. What do you mean by current availability?

Yes its an off grid system that uses the grid as backup so while there is enough solar and battery no grid power is used.
It runs the house fridge , freezer , computer , tv , lights , ACs (2 or 3 split 700 watts each) , kettle , toaster and so on.
So when running and clouds pass over a large current of 60 amps could be drawn from the battery
to do this with sealed batteries I would need 600 Ah or more.
The batteries were not purchased for this task so are a bit too small in Ah for 8000 watts of inverters
two sets would have been best 450Ah so 6 years of life I think is OK.
Now current availability is how much amps you can draw from the battery without damaging it
As stated sealed lead acid is very limited as you can not have gassing occur which creates
a loss of water drying out the battery as you can not add water this limits the current
to 1/10 of the Ah capacity so you have 2 x 200Ah that is 2 x 20 amps so total of 40 amps
that you can draw if you draw more they will gas and dry out.
If you tap your batteries on the sides and they sound empty then they are drying out

What you need to do now is to get a battery load tester and test each one to see how it performs
you can then get all the best ones and make a bank from them.
I did this recently with my banks (11 x 100A 44 12v batteries) and have 3 banks that perform very well
even though they are all the same age 9 years old.

[solamahn]

AGM batteries do not gas during discharge and can be charged at up to .3C rate. Charge rate will taper after meeting bulk charge voltage setting.

[paulvk]

AGM batteries do not gas during discharge and can be charged at up to .3C rate. Charge rate will taper after meeting bulk charge voltage setting.

They most defiantly will if you try to draw a lot of current from them the gas can only recombine at a certain rate same goes for charging.
AGM are VRLA they can supply a lot of current for a very short time but not continuously or repetitively.

[MarkusRE]

Hi all,

let me introduce first.
I‘m Markus from Germany and hopefully soon owner of a pip5048.
I buyed one from wattstunde, which seems to be a generic Voltronic. I had an hour phone call with one of their technicans who was very afraid of the wrong Statements on their website and also their shop site. Complete chaotic. So I hope to get a good one.

Steca and all of the rest are not possible to deliver before end of june, thats why i decided to go with this one.
https://www.wattstunde.de/produkte/elektronik.html

As a second one i want to wait until EASun can deliver again. As it is mentioned a few pages before, the EASun is also a generic one, but 300€ cheaper.

https://www.ebay.de/itm/5000w-PV-4000w- ... 2f306e2064

What do you think about this?
You guys are really great and I had a lot of things to read through this forum.
Btw: The Steca is a 66V Model, also mentioned shortly by a german colleague.

And last but not least, what do you think about this parallel kit?

https://www.ebay.de/itm/202680485800

In my opinion it should fit to all generic models.

Could you confirm this, because its very cheap.

Thanks in advance

Markus
P.S.: Sorry for my bad english

[coulomb]

Hi all,

Welcome to the forum! (finally; I know you had a lot of trouble registering, part of which was my fault).

I buyed one from wattstunde, which seems to be a generic Voltronic. I had an hour phone call with one of their technicans who was very afraid of the wrong Statements on their website and also their shop site. Complete chaotic. So I hope to get a good one.
Steca and all of the rest are not possible to deliver before end of june, thats why i decided to go with this one.
https://www.wattstunde.de/produkte/elektronik.html

Hmmm. The photo shows a model with the heatsink on top. If' it's a real Voltronic, that means it was manufactured about August 2015 or earlier. Also, the web page doesn't mention a PIP-5048MS-like model; the highest power model seems to be a PIP-4048MS or clone thereof. It will be interesting to see the sticker when it arrives; it should have the manufacturing year and month encoded in the serial number. It will also be interesting to see what version of main firmware it comes with.

As a second one i want to wait until EASun can deliver again.
https://www.ebay.de/itm/5000w-PV-4000w- ... 2f306e2064
What do you think about this?

Those seem to be PIP-5048 models, i.e. maximum apparent and real power are 5 kVA and 5 kW respectively. The Wattsunde seems like it will be a 5 kVA but 4 kW model. Those likely can't be paralleled, unless you take the risky step of upgrading the 4 kW model to 5 kW firmware.

And last but not least, what do you think about this parallel kit?
https://www.ebay.de/itm/202680485800
In my opinion it should fit to all generic models.

There isn't all that much to the parallel boards; I'd say it's a safe bet that they will work just fine in any suitable genuine Voltronic Power inverter-charger, and probably most of the clones and work-alikes. But the proof will be in the actual using, perhaps after using for several months.

[weber]

The only thing I have to add to Coulomb's excellent post, is that you can't parallel a 64/66 V model with a 58.4/60 V model. The 58.4 V model is preferred, because it can maintain a proper sine wave down to lower battery voltages.

[paulvk]

The only thing I have to add to Coulomb's excellent post, is that you can't parallel a 64/66 V model with a 58.4/60 V model. The 58.4 V model is preferred, because it can maintain a proper sine wave down to lower battery voltages.

Have you measured the lowest it goes before distortion ?

[MarkusRE]

Hi all,

Welcome to the forum! (finally; I know you had a lot of trouble registering, part of which was my fault).

I buyed one from wattstunde, which seems to be a generic Voltronic. I had an hour phone call with one of their technicans who was very afraid of the wrong Statements on their website and also their shop site. Complete chaotic. So I hope to get a good one.
Steca and all of the rest are not possible to deliver before end of june, thats why i decided to go with this one.
https://www.wattstunde.de/produkte/elektronik.html

Hmmm. The photo shows a model with the heatsink on top. If' it's a real Voltronic, that means it was manufactured about August 2015 or earlier. Also, the web page doesn't mention a PIP-5048MS-like model; the highest power model seems to be a PIP-4048MS or clone thereof. It will be interesting to see the sticker when it arrives; it should have the manufacturing year and month encoded in the serial number. It will also be interesting to see what version of main firmware it comes with.

I hope a good one, because on their shop site they show a different picture. Also they speak from 1500Wp over 2500Wp to 3000Wp on the same page On phone they confirmed the data from a Voltronic one.

https://solarkontor.de/Kombi-Wechselric ... A-48V-MPPT
Here the sity of their shop, there is no heatsink ontop.

As a second one i want to wait until EASun can deliver again.
https://www.ebay.de/itm/5000w-PV-4000w- ... 2f306e2064
What do you think about this?

Those seem to be PIP-5048 models, i.e. maximum apparent and real power are 5 kVA and 5 kW respectively. The Wattsunde seems like it will be a 5 kVA but 4 kW model. Those likely can't be paralleled, unless you take the risky step of upgrading the 4 kW model to 5 kW firmware.

And last but not least, what do you think about this parallel kit?
https://www.ebay.de/itm/202680485800
In my opinion it should fit to all generic models.

There isn't all that much to the parallel boards; I'd say it's a safe bet that they will work just fine in any suitable genuine Voltronic Power inverter-charger, and probably most of the clones and work-alikes. But the proof will be in the actual using, perhaps after using for several months.

Sorry, it seems there went something wrong with quote. [ Edited Coulomb: hopefully fixed the quoting. ]

Regards
Markus

[coulomb]

on their shop site they show a different picture. Also they speak from 1500Wp over 2500Wp to 3000Wp on the same page

Ok, that looks more hopeful. Looks like I saw an old photo.

Even so, whether this will parallel with an EASun that is more likely a genuine Voltronic Power... who knows. My guess is that it will (if the firmwares are sufficiently similar, I can make a better guess when you report the main firmware for the first unit), but they also might both blow up. Blowing up seems unlikely, but it's a non-zero risk. I would stick to the same brand to parallel them. [ Edit: unless the label on the first one looks genuine, in which case they should parallel no problem. So please post a photo of the label as well as the firmware version. ]

[weber]

Have you measured the lowest it goes before distortion ?

Yes. Coulomb and I posted about that back in November. 48.1 V for the 64 V model. 42.1 V for the 58.4 V model.

[paulvk]

Yes. Coulomb and I posted about that back in November. 48.1 V for the 64 V model. 42.1 V for the 58.4 V model.

OK that means I can experiment with 36 NiFe cells and set the back to grid at 44v and bulk at 58v
just waiting on the new caps to arrive to replace the 63v ones.
Also going to replace the two 63v ones in the Esmart-3

[Martino.Q]

........... And yes, it WORKED !!!! Now I've sniffed all communication so I can try to write a better flasher software.
I attach some image to see Aduino connected to inverter +++++++

I substitute the RS232 communication board blow whit a a magnetic isolator

ADUM1201 Dual Channel Digital Magnetic Isolator

Used for test this https://community.blynk.cc/t/problem-wi ... dded/31864
and connected arduino to inverter.
I tested is OK!!!!


Thank you!!

[MarkusRE]

Hello all,

today I received my inverter from Wattstunde and what should I tell you.... one with heatsink on top, what they denied. So they get it back immediatly.

I ordered two EASUN SM5K with powerfactor 1 now.

AF160ED0-8214-438D-8DE1-0070F694429B.png (1.59 MiB) Viewed 1587 times

1A996112-E645-4B72-81E9-7DDA22D4B366.png (1.34 MiB) Viewed 1587 times

Hopefully I‘ll be happy with this, but waiting begins one more time.

Have a nice weekend