Page 32 of 98
Posted: Wed, 20 Jan 2016, 21:23
Northland wrote: righto. That's open circuit. So why not just close the circuit? A resistor or capacitor to draw the voltage down, then drop it out of the circuit once the SCC kicks in? Now where can we find a big fat resistor rated for 120v and thousands of watts? 2 birds - one stone eh?
I don't think you could get that to work practically. There is almost no power in the panels in the early morning, in fact, just enough to blow up your SCC, in accordance with Murphy's law. So you'd need something else to actually heat your water. I don't think you want a relay switching either AC or panels to the element; I can't think how you'd make that vaguely safe.
Besides, if the early morning panels put out over 145 V, even a 24 ohm (2400 W) element will attempt to draw over 6 amps (around 900 W), which the panels won't be able to supply at that time of day, so the voltage will collapse. It may well collapse too low for the PIP to use at all. You also have the problem of chasing your own tail; with the load on the voltage is too low, with the load off the voltage is too high, so the relay chatters until its contacts burn up. Plus, it would have to be a DC rated relay or contactor, which aren't very cheap.
Posted: Wed, 20 Jan 2016, 22:28
Yes probably creates more problems than it solves. OK well what about a zener with reverse bias?
Posted: Wed, 20 Jan 2016, 22:47
Northland wrote: Yes probably creates more problems than it solves. OK well what about a zener with reverse bias?
I really don't think you have to worry, since they are 60 cell PV modules. But if you want to tell me their make and model number I'll do the calculation for you.
Posted: Thu, 21 Jan 2016, 00:25
wow! Too easy....and cheap. Thanks. I might use 2, one as "normal" and another set higher for "booster". Also found same thing but in a case http://www.ebay.com/itm/Digital-Voltage ... SwgkRVUE1p
Posted: Thu, 21 Jan 2016, 03:01
I have my PIP-4048s (2 sets of them 200Km apart) hooked up to TCPIP it can be wireless or cabled
Requires no modifications to the unit
I do use the power from the PIP on the RJ45 to run the wireless unit with a small switch mode regulator module this is the same power that runs the remote.
Watchpower just sees them as serial ports as would any other software on a PC.
I am working with the AVR to communicate with the PIP I am using an Arduino 2560 board but not the Arduino system I write the code in Bascom which is much more stable and a lot easier to use and read.
The AVRs are ideal for this as they can interface to the real world with their inbuilt drivers and ADCs.
The wiznet chips give them the TCPIP
I already have a home automation web server running so I only have to get communication working with the PIPs as Bascom already has Modbus built in as well as a number of CRC functions I should be able to get it working.
My ultimate aim is to add a remote display that uses LCD display that shows operation much like watchpower but is variable.
Note when I get the server running with it you could just use your smart-Phone or any device that has a web browser.
Now just to get enough time to do this.
Posted: Thu, 21 Jan 2016, 03:26
Now as far as the number of panels in series.
On one system
I started with two in series x 6 sets 12 panels ran that for 12 months.
It worked ok and I was able to get the maximum power from the panels.
I have now changed it to three in series x 4 sets still the same 12 panels and it works giving a little more power about 6amps more.
On one day I actually had 47watts more than the power the 12 x 240watt panels should output, 2927watts and they were hot as well!
The maximum voltage the panels output is 105v with three open circuit.
The larger system with sets of three (one PIP at present) 27 x 240watt panels regularly puts a steady 62amps into the batteries.
Note with strings of two the charger heat sink only got hot when the ambient temp was over 30c but with threes it get fairly hot at around 30amps and very hot a 60amps
I have a self regulating fan sitting on the top large systems heat sink.
Posted: Thu, 21 Jan 2016, 03:33
Someone said in nz you get more output than rated. Today my 3kw was showing 3.24kw at one point. So how much can the PIP really do? anyone tested this?
Posted: Thu, 21 Jan 2016, 10:26
OK so I was unable to find a phase control module on ebay.
nevermind, I shall make one.
I bought 4 of these SCR dimmers:
http://pages.ebay.com/link/?nav=item.vi ... 29&alt=web
And 5 of the voltage sensors in cases as above.
Each voltage sensor relay will energize it's own "dimmer", which will each have different output settings. The 5th voltage sensor will output full 2.4kw (I will use 1 element).
Rather than sensing battery voltage I'm thinking a shunt off the battery measuring input current. There will be issues sensing battery voltage depending on what mode the charger is in.
I will want 1v range rather than the normal 500mv range as 0.1v increments will be too fine, so will find the right shunt. The relay timers will be used, maybe a 5 sec delay. Otherwise the extra load will reduce the charge current turning off the relay. Also hoping 5 sec is enough for the SCC to allow more power in if it's been keeping it out because it wasn't required.
So here's the voltage triggers, charging currents and the corresponding dimmer output:
1. 0.2v 5A 200w
2. 0.4v 10A 400w
3. 0.6v 15A 800w
4. 0.8v 20A 1600w
5. 1v 25A 2400w (no dimmer required)
If this doesn't work I can always go back to the battery voltage method
Posted: Thu, 21 Jan 2016, 12:48
I used to use strings of 2 x36v 280w panels. Now I have changed to 3 in series. The heatsinks run hotter but I get more power out. You would think that the lose due to heat would mean less power out. A HS model would work better with 2 panels in series. 15 panels 3 x 5 is the most I ever connect. I always check the tightness of all the DC cable lug screws before installation. Sometimes these screws are not tight enough especially on earlier models. I complained to MPP and now I notice that these screws are tight on later models. Some of the lugs are a bit undersize. The 2 connecting from the scc to the main pcb. Using 12 280w panels would be safer. I have ordered some MSD models but there is a delay in delivery due to FW problem. 24 x 280w on a MSD would be really good.
Posted: Thu, 21 Jan 2016, 13:57
Northland wrote: Someone said in nz you get more output than rated.
From your panels, probably. Colder with good sun in summer. Colder panels produce more power. One of the best places for solar panels is in Nepal; almost equatorial yet cold due to the elevation.
Today my 3kw was showing 3.24kw at one point. So how much can the PIP really do? anyone tested this?
I note that the oft-quoted power limit of 3000 W is really a 60 battery amp limit, using a round 50.0 V as a round voltage when charging. Towards the end of the bulk stage, the battery might actually be at 54.0 V. Then 54.0 x 60 = 3240 W = 3.24 kW.
I'm a little surprised at 62 A charging, as reported by Paulvk. Perhaps that's measuring with an external shunt. I believe that the PIP will aim for 60.0 A, from its own reading. So I would expect the PIP to report around 60 A maximum, even if it's more like 62 A as read externally.
Posted: Thu, 21 Jan 2016, 14:13
Northland your question about cycle life of AGM
The life of your batteries will depend on the depth of discharge if you only use 20% of the available storage you can get 10 to 20 years depending on the type of battery.
Now you have 400ah so using only 80ah worth of power (about 4Kwh) is a good target for long life.
You can not use the figures on watchpower for capacity as a lead acid battery must be at rest for a number of hours before measuring the voltage to get state of charge.
I myself compared the watchpower state of charge to specific gravity readings on my flooded battery and they proved to be meaningless when I measured a state of charge of 80% using specific gravity watchpower showed less than 50%.
Posted: Thu, 21 Jan 2016, 15:23
20 years ...over 70000 cycles at 20% DOD from a AGM battery I would like to see that!
I would think 10 years would be a struggle.
Typically I see graphs topping out at 5000 cycles at 10% DOD dropping to around 2500 cycles at 20% DOD. Also this is assuming the cells are kept at 25C all there life (not often achieved in Australia)
7- 8 years at 20% DOD perhaps less if they are not in a temp controlled environment.
Posted: Thu, 21 Jan 2016, 17:05
offgridQLD wrote: 20 years ...over 70000 cycles at 20% DOD from a AGM battery I would like to see that!
As I said depending on the type of battery large (1000Ah) 2volt flooded cells with pure lead lead plates can last 20years or even more.
I got 13years out of my last lot of VRLA batteries but note those 100Ah 12v batteries cost over $1000 each but due to physical damage to the cases I got them free.
Posted: Thu, 21 Jan 2016, 17:22
I had (in the past) 1300ah 2 volt cells with huge thick tubular plates about as good as lead gets from A big leading brand Name. But they only spec them for 3000 cycle life at 20% DOD (again if they are kept under 25c). That's 8.2 years.
At 10% DOD they spec 5000 cycles 13 years.
Using just 10% DOD gets a bit silly so typically people see 8 - 10 years max at 20% DOD In AU.
Just don't want people getting the wrong Idea thinking they are going to get 20 years of useful service life from lead acid cells.
Count on 10 years if your lucky.
Posted: Thu, 21 Jan 2016, 20:09
coulomb wrote:I note that the oft-quoted power limit of 3000 W is really a 60 battery amp limit, using a round 50.0 V as a round voltage when charging. Towards the end of the bulk stage, the battery might actually be at 54.0 V. Then 54.0 x 60 = 3240 W = 3.24 kW.
So, if there's 3kw load and only 2kw charging will it take 5kw panels? What's the actual mppt limit?
Posted: Thu, 21 Jan 2016, 20:39
coulomb wrote:I note that the oft-quoted power limit of 3000 W is really a 60 battery amp limit...
So, if there's 3kw load and only 2kw charging will it take 5kw panels? What's the actual mppt limit?
Sorry, perhaps I wasn't clear. The SCC measures its power on the battery side, where the current is higher than on the PV side. Example: 3000 W @ 100 V on the PV side is 30 A plus losses, say 32 A; on the battery side, if the battery is 50 V, it will be 60 A. But if the battery is say 55 V, you'll get 10% extra power from the 10% extra voltage, so 3300 W.
So the real limit is 60 A of SCC charge current on the battery side; that's all the MOSFETs can handle.
So any load on the battery is irrelevant to this limit, as is the total current into or out of the battery. In your example of 3 kW load and 2 kW charging, there will be 1 kW net out of the battery, but you still have the ability to increase that charge current from 2 kW to whatever 60 A times the battery voltage comes to.
The nominal power of the PV panels also doesn't matter. If you have 5 kW of panels, then the SCC won't "work harder" than it takes to make 60 A flow into the battery from the panels (via the "DC transformer" of the buck converter inside the SCC). Likewise, if you have 2 kW of panels in ideal conditions, or 5 kW of panels with shade such that there is only 2 kW available, the SCC will hunt around till it finds the most power it can get, which hopefully will be 2 kW. It will essentially work harder and harder until it finds that the power actually starts dropping, and will wind back to where the maximum power point was. As the sun and shade moves, the panels heat up, and so on, the maximum power point will move about, but the maximum power point tracking software keeps working the panels close to the maximum power point all the time.
The advantage of more than 3 kW of panels is that instead of having 3 kW only for half an hour on ideal days, you'll get 3 kW from say 9am to 3 pm (ignoring daylight saving), and only start falling off either side of those hours. You'll never get more than the 3 kW (or whatever 60 A times the battery comes to), but you'll get more "value" from your panels in non-ideal conditions.
Posted: Fri, 22 Jan 2016, 17:05
"(quiz question) what should you do if you turn such a breaker off and hear it continuously arcing or smoke is coming out?"
I would probably (with a wooden stick) have a go at switching the breaker back on again. The arc/smoke and all that fun stuff is from trying to break the contacts (unsuccessfully)restore the contact and the arc should go away.
You will most likely never be able switch it off again but I'm sure you are beyond wanting to do that again until you replace it with a DC rated breaker.
That our Isolate the PV load as this will eliminate any current flow.
Posted: Fri, 22 Jan 2016, 17:17
If anyone needs DC breakers these are the cheapest http://pages.ebay.com/link/?nav=item.vi ... 20&alt=web
I negotiated and got 10pc various ratings for 30usd incl shipping
Posted: Fri, 22 Jan 2016, 17:33
Good answer, Kurt.
I probably wouldn't bother hunting around for a stick. I'd just flip it back on again immediately. But as we saw in that video Coulomb posted, that may not work. And whether it seems to work or not, we should then, as you say, remove the load from the array by turning off the AC isolator (typically a circuit breaker) in the case of a grid-feed system, or the battery isolator (typically a fuse switch) in the case of a battery system.
The standard shutdown procedure, as posted on a notice nearby, should require turning that AC (or lower voltage DC) isolator off first
, before turning off the PV array isolator. Just in case.
Posted: Fri, 22 Jan 2016, 18:00
Re: http://pages.ebay.com/link/?nav=item.vi ... 20&alt=web
I note that these are polarised DC breakers, and as such are now illegal for use as PV array isolators in AU and NZ. And even when they were legal, you would need a two pole version, unless one side of your array was earthed. But then, with a 2 pole device, you can get a 125 Vdc rating (63 Vdc per pole) from an ordinary non-polarised 250 Vac Hager, Merlin-Gerin or ABB 2-pole breaker.
Edit: For arrays whose open circuit voltage is less than 500 Vdc I have used these Noark brand non-polarised breakers as isolators, with no problems so far.
http://www.offgridbatteries.com.au/noar ... t-breaker/
Posted: Fri, 22 Jan 2016, 18:18
solamahn wrote: I used to use strings of 2 x36v 280w panels. Now I have changed to 3 in series. The heatsinks run hotter but I get more power out.
How many cells are these panels? I assume they are 60 cell, not 72 cell? And I note (for others' benefit) that you're installing them in the tropics.
Posted: Fri, 22 Jan 2016, 19:02
weber wrote: But then, with a 2 pole device, you can get a 125 Vdc rating (63 Vdc per pole) from an ordinary non-polarised 250 Vac Hager, Merlin-Gerin or ABB 2-pole breaker.
The new Hager breakers are no longer DC rated (I have spoken to Hager) even for 48v telecoms where they were used for decades. This gave me problems for my battery & distribution cabinet. Note the old ones were 125v dc rated.
Posted: Fri, 22 Jan 2016, 20:17
Update on AVR (Arduino hardware) to PIP
I have an Arduino M2560 board working as an interface between a PC with terminal and the PIP
Type a command into the terminal press enter and the AVR calculates the CRC adds it to the command and then sends it to the PIP.
The PIP sends its response to the AVR which stores it in a variable then sends it to the PC.
This uses two of the serial ports on the M2560 board the one to the PC is the USB connection to the Arduino the other is port two of the M2560 with a max3232 buffer.
the binary for the M2560 is attached you can use the free version of Bascom to load it into the Arduino M2560
Pc terminal setting 57600 , Synchrone = 0 , Parity = None , Stopbits = 1 , Databits = 8
Note the AVR converts the typed characters to upper case.
Posted: Fri, 22 Jan 2016, 21:48
Posted: Sun, 24 Jan 2016, 02:52
Now have the AVR talking happily to the Inverter
If I get enough time over the next few weeks I will try to get the web server AVR give web page control to the Inverter.
The Q1 command returns this from version 52.30
RETURNED DATA>(02672 00001 00 00 01 000 033 044 048 01 00 000 0000 0000 0000 00.00 13—
And here is the AVR making sense out of the QPIGS when I type info it sends the QPIGS command and then prints this to the terminal
Note the inverter is on standby
AC Output Volts=000.0
AC Output Frequency00.0
Inverter Heat Sink Temp=0048C
PV Current into Battery=0000Amps
Solar Charger Battery Voltage=00.00Volts
Battery Discharge Current=00000Amps