Page 41 of 111
Posted: Mon, 09 May 2016, 19:55
So it is a M2560 there were a couple of diffrent mega Arduinos on sale.
Would be best to get the W5100 Ethernet Shield , very cheap and this will enable you to use any device with a web browser to connect to the Arduino
The Mega 2560 has 4 serial ports I set one up as a command line interface which enables you to plug a serial port cable into it then use a computer to set up it IP address and also command it to do various things which are also available on the web page interfaces
Loading the software is simple I send you the binary file, you download the free version of Bascom and use it to put the binary file into the Arduino via the usb cable.
You will then have a working Arduino web server that can send commands to the PIP at a preset interval just like the supplied software on the CD
I have not yet written the commands to change settings but it is simple to do as the software I have written already takes care of all the communication CRC etc.
If you also get a LCD display like this https://www.rockby.com.au/catresultsSql ... k_no=37420
you will have all the voltages , currents , power output etc displayed on it no need to press the up/down buttons to see what is going on.
Note Bascom is basic type of language much like visual basic with libraries written in assembly that are automatically called up when a command is used this results in powerful compact compiled code for the AVR
Might find some work for the uno's as well.
Posted: Mon, 09 May 2016, 20:42
I have a 16x2 display
already but just ordered a 20x4 display with adapter.
I don't have a desktop computer so no serial port. Can I solder a male network plug on to the comms pins or is there some reason that won't work?
Correction - the other arduino boards are nano (5pc), not uno
Posted: Tue, 10 May 2016, 01:14
Do you have a computer?
A usb to serial adapter with 5v interface will work fine. like this
http://www.ebay.com.au/itm/New-6Pin-USB ... SwymxVJOX8
You will need a rs232 interface for the PIP as it puts 15v on the serial lines this would do http://www.ebay.com.au/itm/MAX3232-RS23 ... Swu1VW8SVq
You then get a RJ45 LAN cable cut one end off put a male DB9 on the end
The PIP has a power output that can run the Arduino so a voltage regulator like this will give you 5v to run it, best not to use the one on the Arduino as its very inefficient http://www.ebay.com.au/itm/LM2596-DC-Bu ... xycmBS6ewl
Yes I said the uno would be useful for that reason and the one wire devices are finicky so best to have a dedicated micro talking to them if you put the rs232 interfaces on them you can talk between the arduino's over hundreds of meters on cat6 cable
I am going to have a dedicated AVR talking to up to 20 temp sensors on my batteries sending data to the main server.
Since you have Arduino boards and a 16x2 display why not download Bascom free version and have a play it will even run on windows XP
I have code for this display http://www.ebay.com/itm/3208-Matrix-Clo ... Swu4BV0zBy
So it could be used to display info and with an old infrared remote control you could select what to display Note I have a few of these I wrote the code for a working clock with remote control and temperature display
Posted: Tue, 10 May 2016, 02:44
paulvk wrote: if you put the rs232 interfaces on them you can talk between the arduino's over hundreds of meters on cat6 cable
Does that mean I need more than 1?
I have the MPP remote, so rj45 cable already run.
I've ordered those extra bits. Cheers
Posted: Tue, 10 May 2016, 04:02
Yes at least 3, one for the PIP and pairs for Arduino to arduino comms you will use one of the 4 serial ports for the PIP on the mega another is for the command interface that leaves two for communication to other devices.
The rs232 to TTL boards convert the signals to 15volts and have protection against voltage spikes this allows for communication over much longer distances.
The good thing about the Arduino is the parts are cheap and there are lots of them.
If you want to have a look at what I have been doing with the AVRs/Arduino have a look through the bascom forum http://www.mcselec.com/index2.php?optio ... &Itemid=59
Note on page 40 of this discussion I posted a screen shot of a simple web page I made to test the Arduino mega talking to the PIP the Arduino actually returns the page from its SD card filling in the results from the PIP.
Posted: Tue, 10 May 2016, 13:17
Came across http://arduino.stackexchange.com/questi ... o-arduinos
and since my distance is about 15m between arduinos I think it will be fine.
Posted: Tue, 10 May 2016, 15:27
Posted: Fri, 13 May 2016, 01:07
I spent many hours looking into how I could control the SCR with arduino. Complex circuits, resistors & relays, opto-isolators....ugh.
Eventually I stumbled across this arduino controlled triac
and at 4A per channel I can control a 16A element much more simply
Posted: Sat, 21 May 2016, 22:59
I just learned that there are some reputable 60 cell polycrystalline solar panels where you can string 3 of them in series and still stay under 120 V open circuit, provided your coldest expected dawn temperature is no lower than 4 °C. They should be fine near the Australian coast at latitudes north of about 32° S. But if you ever get frosts, forget it.
They are the REC PE (Peak Energy) series.
http://www.gold-coast-solar-power-solut ... _sheet.pdf
This is useful, as it is illegal (since some time last year) to connect an array whose open circuit voltage can be expected to exceed 120 V, to a PIP-4048 in Australia or New Zealand, because the PIP lacks insulation monitoring.
I previously thought the only option there was 2 x 72-cell panels, or 3 x 54-cell panels (which are somewhat rare).
Posted: Sat, 21 May 2016, 23:40
Well this was interesting
So for the past week I've spent quite a few hours relearning arduino coding and building a few sketches.
-Built soc calculator which takes into account things like puekurts law and the efficiency of lead acid.
-phase control of a 16A load dependant on battery charge current (8 stage @ 2A)
Those probably would have been an hours work for someone else. I managed to find a bug in arduino that wasted about 6 hrs.
And yes Paul, I did go to the bascom forum but since nowhere (on that forum, Google or YouTube) is there any kind of simple explanation of what bascom is and what it does I read some posts then my brain hurt.
Can someone please provide me with simple code to send qpigs command from arduino to the pip and read the response?
Posted: Mon, 23 May 2016, 02:09
I have written it in Basic (bascom)
Bascom is Basic syntax very much like Visual basic
To print to the RS232 port 1 QPIGS
To print to a LCD
Lcd "Hello World"
I find "C" very obscure I can see how a maths person would like it though.
Being a powerful compiler there have been some very complex things written with bascom so just reading the forum would make things hard.
Did you register on the forum
Did you download the demo the files contain many examples that are meant to help in learning how to use it.
If you go to the bascom forum and make a post in the various section and mention the PIP I can help you by putting the code I have already written as a binary file you can just load into the board no reason you should have to reinvent the wheel
Note if you search the bascom forum for MPP Solar there is a program I posted that uses 2 ports on an arduino that lets you use a terminal program on a PC to send the commands to the PIP it does the CRC calculation
It would be too much off subject to continue here.
Posted: Tue, 24 May 2016, 04:38
I'm asking this question on behalf of another person who is running a PIP4048ms. I have quoted in below some of there experience with there consent.
"Today had the right conditions to upset the inverter/charger, bright sunny morning followed by fast moving dark clouds; came home to find the unit had not restarted. I checked the screen and an error 8 was displayed (BUS over voltage) It's like the mppt charger can't adapt to fast changing PV voltages which is why I was considering changing to a standalone charging unit."
"I've sat in front of the display during cloudy days hoping to catch it but it just won't do it when I'm there or the clouds aren't moving fast enough"
"OK I'm having a clouded sunny day again and my PV voltage went from 71v to an absolute maximum on a dark cloud edge of voltage of 111v lighter clouds were up to 105v before settling down to 98-100v in full sunlight; No shutdowns so far.
The only difference today is the clouds are moving slower, when the shut downs occur the clouds are moving faster which is why I thought the unit was not coping with a sudden change in PV voltage?
My panels are Solarborn SL250TU-30P. Max power current Imp 8.06. Max power volts Vmp 31.02. Short circuit Isc 8.62 and VOC 36.99"
He is running the panels as three in series. I think the voltage looks ok to me in spec.
Has anyone had this issue with the pip shutting down under similar conditions? Like he mentioned its not often but often enough for it to be a issue. No one wants to come back to a offgrid system and find the inverter has shut its self down.
Posted: Tue, 24 May 2016, 05:12
I haven't had that problem. How many panels? What is the battery -- chemistry, amp-hours? Does it only happen when the battery is full? Absorb voltage? I'm thinking it may be a combination of the solar charge controller not backing off quickly enough, and a high impedance battery.
Posted: Tue, 24 May 2016, 11:52
This was a problem with earlier 4048's. What year of manufacture and firmware revisions.
Posted: Tue, 24 May 2016, 14:22
He has a total of 3000w of PV. The above mentioned 250w panels set up as 3s four string of them. The same issue was accruing when he only had one string and 750w of PV conected.
His batterys are 220amp/hrs @48v of AGM Batteries. He intends to get more batterys when the system is switched over to full time use.
The age of the unit is 9 - 2014. Similar to the one I have.
He mentioned that some times it recovers after the hiccup and restarts by it's self but some times it's enough that it just shuts down.
He is reasonably sure the unit was still in absorb.
Posted: Tue, 24 May 2016, 14:37
Definitely not enough batteries but that is not what is causing the problem because it used to happen with less panels. The scc is trying to limit battery charge current and voltage. That's why the PV voltage is high 111v. This could also indicate that the batteries are full. The problem that I experienced with a 2013 model 4048 was when the batteries were full and a load was disconnected like a fridge cycling off. The battery voltage would climb to 60v and the inverter would shut down. I changed the CPU from an R1 to R4 and the problem disappeared.
Posted: Tue, 24 May 2016, 14:52
The bus over voltage may not be referring to the solar side
The bus voltage reading I get from my PIPs with the QPIGS is the high voltage DC bus that is used to generate the 230V AC
I had nearly 6KW of panels on one PIP for months with no problems it was charging at 62amps for hours at a time with the two PIPs in parallel now I had 96amps from the same panels
My smaller system has 12 x 240 watt panels in sets of 3 and 225Ah trojans it has not missed a beat in 18months
Posted: Tue, 24 May 2016, 15:02
I guess the question then is why is it coinciding with strong cloud edge effect for him. It seems to be more than just by chance.
Posted: Tue, 24 May 2016, 16:43
It may be a software problem as previously suggested not able to handle the ups and downs on the DC side so getting to high on the high voltage DC bus
So date of manufacture & software version will be needed.
Posted: Tue, 24 May 2016, 16:46
9 - 2014 is the date of manufacture. I will ask him for what software he is running.
Posted: Tue, 24 May 2016, 17:07
What solamahn and paulvk say, sounds right to me. As paulvk says, the bus voltage it is referring to would be the approximately 400 Vdc internal bus. But my understanding is that that is essentially the battery voltage times the turns ratio of the high-frequency transformer in the bidirectional DC-DC converter stage. So it can be read as saying the battery voltage (not the solar input voltage) has gone too high.
Solamahn, you mention changing the CPU from an R1 to an R4. Does that involve replacing the CPU circuit board that plugs into the main inverter/charger board? Or are you referring to a firmware upgrade?
Assuming that Kurt's friend already has a suitable CPU and firmware, something else he could try would be to lower the absorb voltage. If it is even a little bit too high for the AGMs then towards the end of the absorb stage they may have generated enough gas bubbles on their plates that they have become high resistance and cannot contribute to controlling the bus voltage when a load shuts off, or when a cloud suddenly passes.
If he has parameter 5 set to "AGn" then the absorb voltage is 56.4 V. He could try setting parameter 5 to "USE", then setting parameter 26 down to 55.2 V. Then test it by shutting off some load when the batteries have nearly finished absorb mode (before they go to float).
If the problem definitely goes away with an absorb voltage of 55.2 V, but he thinks that may be too low to keep the AGMs balanced, then he could try splitting the difference and using 55.8 V and seeing if the problem comes back.
[Edit: Corrected grammar, "but thinks" -> "but he thinks"]
Posted: Tue, 24 May 2016, 17:27
Yes. I changed the CPU board that plugs into the main pcb. This problem only occurs with a full battery because that is when the battery does not contribute much to the DC voltage regulation. If the battery was kept below full then it should stop the problem but not good for the batteries. My problem only occurred when there was only one appliance like a fridge being powered. Mine had nothing to do with clouds. It might just be that use is programmed and bulk or float charge voltage settings are too high. I had another site where the batteries were getting tired, bulk was set to 58.4 and the charging would overshoot that value and go to 59 and trip out. Some are 59, some are 60. I think some of the new ones are 61. Changing the bulk to 58 fixed that one.
Posted: Tue, 24 May 2016, 19:42
offgridQLD wrote: (quoting friend) ... "I checked the screen and an error 8 was displayed (BUS over voltage) It's like the mppt charger can't adapt to fast changing PV voltages which is why I was considering changing to a standalone charging unit."
The algorithm handling error 08 changed somewhere between 52.02 and 52.30. In versions 52.30 and later, the criteria are (bus voltage > 500 V for 16.7 seconds) and (rate of reduction in bus voltage > 10.0 V over the last 15 seconds). I suspect that it's the last check that has made the difference with these nuisance fault codes. It seems that under unusual circumstances (e.g. full sun then sudden cloud), the bus voltage can exceed 500 V for a while, as the MPPT or more likely the buck unit may not react fast enough. The electrolytic bus capacitors are rated at 500 V, so they can take an infrequent over-voltage for short periods of time, taking a small hit to life span as long as it doesn't last long. It looks like with the later firmware, they put in an additional check to ensure that the bus voltage is decaying quickly enough to save the capacitors and the silicon.
The problem is that later firmware seems to assume a hardware upgrade, in particular the utility (mains) charging rate from 30 to 60 A, so if you upgrade to 52.30 or 72.40 (no reason not to do the latter, as far as I know), be aware that you should not set parameter 11 to more than 30 A. See Weber's post on this
. [ Edit: oops, I see it's a 2014 model. That means it already has the "high current update" in the hardware, so later firmware should be fine. But I would think that a 2014 model would already have at least 52.30 firmware. ]
Posted: Tue, 24 May 2016, 19:50
Well explained, thanks Coulomb
I think that's the most likely path to head down. I will get confirmation of the software version and if its the older software I will advise updating with the above warning in mind.
I have him reading this thread from cover to cover so he should stumble upon it soon.
Posted: Tue, 24 May 2016, 19:56
I have him reading this thread from cover to cover so he should stumble upon it soon.
Yes, any month now
Please note the edit above. It sounds like he will already have modern firmware. He could possibly update from 52.X (where X is much greater than 2) to 72.40, but that part of the firmware is the same, so it's unlikely to fix the problem. But yes, that's the next step; verify the versions of both firmwares (U1 and U2).