Axpert KS PV sensing

[Guest]

Hi to all, I am new to the forum and have been reading allot on previous topics. I would also like to contribute to the forum and share some of my own experiences and knowledge of these Chinese INVERTER/CHARGE controllers.

Installed hardware:

PV system: 6 x 255Wp in parallel.

Battery system: ~3.24kWh capacity (54 x 5Ah arranged for 24V)

Inverter setup:

MAIN CPU (U1): 15 68

SECONDARY CPU (U2): 03 23

PROGRAM	SELECTION	COMMENTS
00	SOL
02	30A
03	UPS
04	DISABLED
05	USE
06	DISABLED
07	DISABLED
09	50Hz
11	30A
12	24V	TO MAINTAIN 50% SOC? (voltage method)?
13	FULL	DIFFERENCE BETWEEN FULL & 29V?
16	SOLAR & ULITITY
18	ALARM ON
19	STAY LATEST SCREEN
20	BACKLIGHT OFF
22	ON
23	ENABLED
25	ENABLED
26	28.2V
27	27V
29	22V

The issue i do have is with the firmware programmed to switch the inverter mode from LINE to BATT mode when it detects the voltage of the PV to be around 24Vd.c.

At 24Vd.c there is no current produced by the panels and thus draining my BATT bank unnecessary.

Is there a work -around to this?

I've seen some patch firmware for the 4/5 kVA units, is there something for the 3kVA out?

 

PS: I've managed to source the actual service manuals for all 1-5kVA Axpert models, these includes repair procedures and detailed schematic diagrams with component data.

[Guest]

Welcome vandyh.

I am sure the master will be around momentarily.

So, do I gather you can repair these units if you wanted to?

[Chris Hobson]

Hi Vandyd

Welcome.

Something concerns me about your battery bank. Are they standard 2V lead acid batteries? If so you are you are short a couple of batteries. If they are some other chemistry you need to look carefully at your bulking and float settings.If the are Lead Acid then your charging rate is too high. How the hell do you wire up something like that since I am sure the terminals are not designed for the thickness of cable you will end up needing.

[___]

9 hours ago, Chris Hobson said:

Something concerns me about your battery bank.

He said 54 x 5Ah. I hope that's a mistake, that it is not what I think it is... 27 strings of 2 x 12V batteries, each rated at 5Ah. Or in other words, a massive amount of strings made out of alarm batteries...

[Guest]

I do have 54 x (12V 5Ah) batteries (they came out of a broken UPS), 2 batteries wired in series to have 24V, then 27 sets wired in parallel.

This is not ideal, but they cost me nothing, so for now it must do.

I will post all photos tomorrow 

[Guest]

I would not do that if I was you, why, for I tried that already. 

The max charge current for those little batteries are very low. Midpoint voltages are going to be a bastard to sort. 

But, you have them, you want to try ... have fun. 

[Guest]

20 hours ago, The Terrible Triplett said:

I would not do that if I was you, why, for I tried that already. 

The max charge current for those little batteries are very low. Midpoint voltages are going to be a bastard to sort. 

But, you have them, you want to try ... have fun. 

The max current per batt is 1.5A. Thus the bank can be charged at 40A. Thats why setting 11 is set to 30A and not the max 50A.

BP5-12.pdf

[Guest]

2 hours ago, vandyh said:

Thus the bank can be charged at 40A. Thats why setting 11 is set to 30A and not the max 50A.

I am going out on a limb here.

Once had a 3000VA online UPS with internal and external battery pack. It had a 96v bank consisting of 32 x 9ah batteries, AND it could take another 16 external batteries. So a LOT of batteries as you have.

So I got a brainwave, which turned out to be  a brainfart: Why all these small batts!? Why not connect one time 8 x 105ah batteries?

Got hold of the manufacturer asking them what is the charge current on the UPS, for the manuals where useless.

Now IF I recall correctly, it was awhile ago, it was something like 2amps or thereabout for ALL the batteries connected to the machine.

2 hours ago, vandyh said:

The max current per batt is 1.5A.

So I want to go out on a limb again and say, as you quoted, 1.5amps per batt is your charge limit?

[Guest]

On 08/05/2016 at 10:13 AM, The Terrible Triplett said:

I am going out on a limb here.

Once had a 3000VA online UPS with internal and external battery pack. It had a 96v bank consisting of 32 x 9ah batteries, AND it could take another 16 external batteries. So a LOT of batteries as you have.

So I got a brainwave, which turned out to be  a brainfart: Why all these small batts!? Why not connect one time 8 x 105ah batteries?

Got hold of the manufacturer asking them what is the charge current on the UPS, for the manuals where useless.

Now IF I recall correctly, it was awhile ago, it was something like 2amps or thereabout for ALL the batteries connected to the machine.

So I want to go out on a limb again and say, as you quoted, 1.5amps per batt is your charge limit?

The batt bank i've made up works perfectly and it cost me nothing more than a few lugs and scrap cable.

Yes I would like to go for the 105Ah option or even the LiFePO4 (with a new charge controller), but for now it works fine.

[___]

Very neatly done. One small issue, the batteries on the eight will work a lot harder than the ones on the left. Longer path through the left side, and electricity takes the shorter path. An easy way to resolve that is to connect positive and negative on opposite sides of the bank so the differences in length cancel each other. Even then, the outside batteries will work harder than the inside ones.

The biggest problem with this, I think, is if one of the batteries fail. Kind of reminds me of discussions about why RAID5* is no longer a good idea but I digress. Basically, assuming a uniform chance of internal failure/short, the more strings you have, the higher the overall probability. When any particular cell fails, it will draw down all the others. Anyway, you know it's suboptimal, I'm not going to harp on that any more :-)

Side discussion on RAID5. So RAID5 is n+1, you have n data disks and one parity disk (parity is usually spread over all the disks, to increase write speed). You can lose any one disk and construct the data out of the remaining n disks, by using the parity. However: The probability of failure has stayed pretty constant over the years, but the disk size has increased a lot. As a result, the probability of having two overlapping failures are now in the double-digits. Use of RAID5 is therefore discouraged.

[Guest]

5 hours ago, plonkster said:

Very neatly done. One small issue, the batteries on the eight will work a lot harder than the ones on the left. Longer path through the left side, and electricity takes the shorter path. An easy way to resolve that is to connect positive and negative on opposite sides of the bank so the differences in length cancel each other. Even then, the outside batteries will work harder than the inside ones.

The biggest problem with this, I think, is if one of the batteries fail. Kind of reminds me of discussions about why RAID5* is no longer a good idea but I digress. Basically, assuming a uniform chance of internal failure/short, the more strings you have, the higher the overall probability. When any particular cell fails, it will draw down all the others. Anyway, you know it's suboptimal, I'm not going to harp on that any more :-)

Side discussion on RAID5. So RAID5 is n+1, you have n data disks and one parity disk (parity is usually spread over all the disks, to increase write speed). You can lose any one disk and construct the data out of the remaining n disks, by using the parity. However: The probability of failure has stayed pretty constant over the years, but the disk size has increased a lot. As a result, the probability of having two overlapping failures are now in the double-digits. Use of RAID5 is therefore discouraged.

Thank you for sharing. I will definitely rearrange the bank to compesate.