Page 43 of 102
Posted: Thu, 02 Jun 2016, 02:52
Its not a finished version just a beginning using parts of the code I have already written.
Its too much work to make a USB master so get some photos of the interface board in that inverter I think there will be rs232 on it to connect to.
Now I need to know what LCD you have, photo would be nice or link to its connections I think you said it was 4 x 20 but is it a shield or bare LCD
There are not enough pins on a nano for a graphic LCD it needs a full 8 bit I looked at that.
I already have the info from the inverter displayed on LCD
Test file for Xloader blinks led on nano board
Posted: Thu, 02 Jun 2016, 03:53
I haven't looked at the 4048, but on this one there is a mini USB connection at the back of the bigger board the cable connects to. And it has 5v. Might be useful.
My lcd is a 20x4 i2c (SCA, SCL pins) but I was going to add 18B20 reading (pin D2) and current through a hall sensor (for hot water, not pinned yet), and add soc calc to the display.
So my characters would be like this :
I can give you my soc code, I made one for lead acid and another for lifepo4. But adding all this stuff is probably asking too much, which is why I thought it would be better as a module. I (or someone else) wouldn't want to ask you to recode it every time a mod is required (unless you would enjoy it). I could have my sensors and lcd on a second board, and recode myself. Also I've ordered 2 of the multiplexer boards for the batteries so that will use another nano. And they all are going to need to talk to the main (mega) board which has the ethernet shield
Posted: Thu, 02 Jun 2016, 04:25
Dear Northland, I'm not trying to sell you anything. I understood you were planning to design your own BMS from scratch. You specifically explained why you did not
want to buy one off the shelf here
. So I offered you an existing well-tested design for free. If you can't see the value in that, it's no skin off my nose. I don't understand the hostility.
Northland wrote:I bet you could turn an arduino into a BMS in 5 minutes flat. And for about $5. You make it sound like an ordeal. As I said above it's a nano, a relay and a resistor. It's probably 10 lines of code.
No, I could not. No one could. It's never that simple.
Posted: Thu, 02 Jun 2016, 12:54
Sorry, I don't recall being hostile. You asked "why not use this?" and I explained why. Because it's only a design, not a tangible product I can use. You techy guys often assume everyone has your level of skill and your level of knowledge. But that's not reality.
I even offered to help you get it adopted as a standard fitting. I suggested a kickstarter campaign (this was extremely successful for electrodacus SBMS) and to show your design to a manufacturer. And somehow you take that as hostility?
I doubt you guys went to all the trouble of designing it for only a few installs? You clearly want it to be adopted widely or you wouldn't have made it open source, nor would you have said "here, use this". I was merely pointing out that I was utterly confused about the concept, economic viability, time schedule, procedure, state of completion etc. If you were to add this information to the thread it may help it become more widely adopted. The fact that we are discussing it should give you some kind of satisfaction. Would you prefer we ignored it completely? It's not like I said "what a crappy product" or "what idiot thought this up"... that would be hostility. I'm not as smart as you. I don't get it.
I probably don't communicate as well as others and sometimes my comments are interpreted differently than intended.
"no one could"
"it's never that simple"
Well you are the guru. Care to be more specific? Or is flat discouragement your final response?
Here's what I have learned. Lifepo4 have a flat discharge curve. So trying to determine state of charge during 90% of the discharge will be pointless because voltage is fairly constant. So you either have to balance at the top or the bottom. Each method has their supporters who swear it's best. I assume (I don't recall you stating this) your design is top balance because you seem to have resistors to discharge. And I can't see how you could boost charge when only connected to a single cell. So when voltage gets higher than other cells it bypasses some current to prevent overcharging. Seems pretty simple to me. Why is that beyond the processing ability of an arduino? Why is that beyond the skill of a programmer?
If volts > 3.5 and if volts x-amount higher than other cells do this : make pin z go high which engages a relay parallel to the cell with a resistor in series.
If volts x-amount higher than other cells make the pin go low again.
If cell > 3.6 tell the charger to stop. Until the difference is x-amount.
Record cell voltage and send the data here:
If volts < 2.6 sound an alarm
If volts < 2.5 make the batteries disconnect
Now the figures might need slight tweaking but it's far from impossible or complicated. Or am I missing something?
Posted: Thu, 02 Jun 2016, 14:48
That board appears to be basically same as the one in the PIPs there are two opto couplers to isolate the high voltages from getting to a person via the USB chip 13361
The end of these near that IC I think will be at 5V the other side may be much higher so be careful if you take measurements when it is on.
There should also be a transformer on that board to isolate the power supplied to the board but is no seen in the photo.
So there should be 5V to run a nano and RS232 at 5V to connect to it.
Note also those like myself who have PIPs with serial ports and would like USB the nano could do this I will have a look.
Posted: Thu, 02 Jun 2016, 18:41
I had a look at coms board discussion
but my board is different. Definitely no transformer on it. I wonder if that's because it's a 12v inverter and the voltage doesn't need to step down
Posted: Thu, 02 Jun 2016, 19:51
Yes you could be right being 12v they did not need to reduce the voltage with a transformer.
But still it is best to treat it as though there may be high voltage there they have put the opto couplers in and slotted the PCB
Note the 18B20s are a bit touchy with timing on the 1 wire buss this is why I am going to use a dedicated nano to talk to them then send the data back
Did you try the blink with Xloader
Posted: Thu, 02 Jun 2016, 20:40
yes, 7th blink is a double
on that 6 pin cable, pins 1,2,3 (top) seem to have 0-3v across them (bouncing wildly) with watchpower running. If I disconnect the cable, the meter says 103v. Is it worth a try to connect those 3 pins as Grd,TX,RX in each combination or risky? I have 5 nanos so not a big deal if one pops.
The optocouplers seem to have 100v on both sides, but if I go between them I can get 2-3v
Posted: Thu, 02 Jun 2016, 21:32
Your 103v maybe leakage with no earth on the neutral output of inverter
The PIP-4048 has a similar problem if you put an earth on its neutral output and intended to supply its 240v input from a system with an earthed neutral you end up with two earths on the neutral in two different places so a relay is need to remove the earth on the output when in bypass mode.
We may be over thinking this because you could just use one of those RS232 to 5V units to give you a RS232 port on it
Posted: Thu, 02 Jun 2016, 22:50
The new Monolith cabinets finally arrived from the fabricators. Hoorah! It's been 3.5 months since I sent them the first drawing! But they did a good job at reasonable cost. Thanks to Damon Rahmate for recommending the fabricator and powder-coater, and to him and Andrew Davey for transport.
These cabinets are designed to take two PIP-4048s and switchboards in the upper section, and 18 kWh of LiFePO4 in the lower section, and to be installed outdoors on a concrete pad against a wall, with 30 mm standoffs for ventilation via the perforations you can see in the back panel.
Posted: Thu, 02 Jun 2016, 23:19
nice clean layout
Posted: Fri, 03 Jun 2016, 00:47
Weber make sure ants can not get in they love the warmth for the babies!
Posted: Fri, 03 Jun 2016, 00:53
paulvk wrote: Weber make sure ants can not get in they love the warmth for the babies!
That's a good thought, paulvk. But how do I do that while still having ventilation? If I have mesh so fine that small black ants can't get through it, it will clog with dust in no time.
Posted: Fri, 03 Jun 2016, 01:37
Thats a nice simple layout. How is the end user accessing the PIP screen and buttons along with the switchboard? As it looks like the door is bolted on sheet of steel.
Perhaps if it sits on adjustable legs you could place each leg on something like a water filled plate..something better than that but just starting of the idea.
The only issue with that is you have cables that need to enter the cabinet that becomes a bridge for the ants.
I have 1000's of them just near my power room but I havn't had then in the power room perhaps to cold now it's refrigerated or just good luck.
Posted: Fri, 03 Jun 2016, 02:55
offgridQLD wrote: Thats a nice simple layout. How is the end user accessing the PIP screen and buttons along with the switchboard? As it looks like the door is bolted on sheet of steel.
The whole thing is 2 mm aluminium. The plan is for the end user _not_ to access the PIP screen and buttons. It should all be done via WiFi or ethernet. A Beagle-Bone Black inside serving up a web page and talking to the BMS master which talks to the PIP.
As in the off-grid Monolith, the switchboards will have breakers and fuses for overcurrent protection and short-circuit protection, but isolation will be by contactors, so they can be operated by a big red button on the wall beside the monolith.
Posted: Fri, 03 Jun 2016, 04:16
I am using a mini-PC
(stuck to the wall with velcro) to monitor and control my PIP4048. Also logging using the AICC Solar Monitor Free Software
which logs all data including BMV702 which automatically uploads logging every five minutes to PVoutput.org. All done in the kitchen.
I use Microsoft remote desktop program to access the mini-PC from another PC or my tablet via Wifi. I can also setup the mini-PC to be accessed remotely via internet using a VPN. Work in progress.
You can view my logging on PVoutput.org which is available to public view at My PVoutput Site
You can also follow me on PVoutput.org
I just thought to post my setup to maybe benefit others.
Posted: Fri, 03 Jun 2016, 05:50
The ants are a real danger I caught them trying to get warm so I sprayed the wall & cables around the PIPs with a surface spray.
On the VPN side I have had a VPN network for many years I have two networks 200km apart linked via VPN into one.
I can also dial in from any computer and become part of the network.
The best way is to use a router with inbuilt VPN servers at one end I now have NBN fiber so had to replace my 10year old ADSL VPN router, I got a TP-Link TL-R600VPN router it cost less than $100 and has dial in and point to point VPN servers with encryption.
When I finish writing the software for the AVR server I will be able to dial in and it will serve the data to me, it will also log to an SD card and will send that data if requested it also has automation so I can turn things on or off as well, I already have one working I turn my xmas lights off/on when I am not home and can shut off my watering system.
Posted: Fri, 03 Jun 2016, 07:01
weber wrote: installed outdoors on a concrete pad against a wall, with 30 mm standoffs for ventilation via the perforations you can see in the back panel.
dose that include the monolith will be 30mm off the slab as well?
think of the way light poles are installed in concrete with bolts
and have some sort of tray or something around the bolts that you could now and then add ant sand or the like in to
Posted: Fri, 03 Jun 2016, 16:39
Thanks for the anti-ant ideas paulvk, offgridQLD and Adverse. Unfortunately, any kind of tray underneath will be defeated by the bolts and standoffs tying it to the wall so it can't fall over, as well as the aforementioned cable conduits. So the best option sounds like surface spray on the vents and joints and drain holes, on the inside of the cabinet. And regular inspection by the owner, looking for ant-trails leading into the cabinet. And a cover-off inspection and respray at least once a year.
Because of the sheer weight of batteries, and to maintain the illusion that it just grows out of the slab, we are planning to use six of these slide glides
under it. They are only about 10 mm high. They will be packed up with some durable kind of shim, as needed to compensate for any unevenness of the slab.
Damon Rahmate made an excellent suggestion for improving weather resistance without impacting the free-standing-monolith-like appearance, in the case where it is not already underneath eaves. Namely a canopy
attached to the wall above it.
Posted: Fri, 03 Jun 2016, 19:54
The other thing to keep in mind with such an enclosure outside is condensation and humidity inside the cabinet if you get a day with dew in the air it will be drawn into the cabinet.
With such a cabinet I would have mounted the inverters on a swing out panel to make them easy work on and had forced air ventilation incorporated into it.
Also what I have done with my units is to make them not hard wired by fitting anderson power connectors on the DC and 25 amp connectors on mains in and out.
It takes me 10 minutes to change an inverter
When I get home I will find the part numbers for the mains connectors.
Posted: Sun, 05 Jun 2016, 01:54
Received an email from MPP Solar regarding their 2016 inverter range. Higher input voltage, higher charging rates, I wonder how much of the firmware has been changed or if the cabinet size has changed, each new model so far appears to be in a different size casing
Posted: Sun, 05 Jun 2016, 23:37
Terry, I told them (mpp on eBay) I had 3kw model and wanted to upgrade the scc, they said it was not compatible and would not fit anyway because the new case is bigger. They offered me the new model for 999aud including shipping. Might be useful for someone looking to buy
Posted: Mon, 06 Jun 2016, 07:32
First of all, I would like to say this thread is very informative. I want to ask if anyone here has been using two inverters in parallel and getting an error 80?
I have two 5kva inverters in parallel mode, which run fine but randomly give me error 80 (CAN data loss). I have checked my parallel kit cables & current sharing cables. All cables seem to be fine but I can't figure out why I get this error. Both the inverters are on same firmware 52.30 and same scc firmware as well.
If anyone has knowledge on this, please guide me.
Posted: Fri, 10 Jun 2016, 00:47
Nasha, everyone on this list is always very helpful and forthcoming. So I think you can take it from the silence, that none of us have a clue. Sorry.
Posted: Fri, 10 Jun 2016, 01:55
I have installed more than 10 parallel systems, 4048 and 3248 without any problems. Only suggestion I could make is use both parallel inverter connecting cables and join ac out cables together