2 Quick Questions

Ako

Taken from the M* ts-60 manual  " DAY AND NIGHT CONDITIONS ARE DETECTED BY THE TRISTAR AND NO BLOCKING DIODS ARE USED IN THE POWER PATH " , i thought they were necessary to stop the batteries draining ,at night , i know chargers i have had previously mentioned the fact and reading specks on others they mostly mention they are incorporated to stop this . Should i assume that the M* ts60 relies on the user having the diods in each pannel or somewhere in the line between them and the CC .

The other question is how can my batteries charge when the input voltage from the CC is lower than the Batteries as shown in the attached screenshot , this is the situation every day when the sun starts to go down or even days when its overcast .

littleharbor2

Looks to me like you have 60 cell panels trying to charge a 24 volt bank. Once they warm up the voltage drops too low to maintain a 24 volt battery bank.  What panels are you using?

Estragon

I assume you had loads running during the log period shown? Looks like the pv is putting out ~9a at high enough voltage to power loads, but not high enough to do much charging.

Agree with lh2 -panel specs would help.

Ako

Another question but i think like me it will be an a logical assumption , the log shown high and low voltage , would they be the battery voltages taken from the CC terminals or the sense wires . Logically it would be the sense wires as thats where the control voltages are taken from but the log does say Vb Min/Max so for someone with a large voltage drop between CC and batteries then Battery voltage would be totally misleading especially at the most important time which for me would be first thing in the morning before charging begins and there has been nothing in or out for several hours with the exception of the inverter which states " Consumption ( in detection )
< 70 mA " , im assuming this means self consumption because ts switched on . I apologise for coming back with so many questions that might be obvious to many but im trying to get a fuller understanding of everything . Will look at the panels and post the info for the previous replies from Littleharbour and Estragon

Ako

Littleharbour , " Looks to me like you have 60 cell panels trying to charge a 24 volt bank. Once they warm up the voltage drops too low to maintain a 24 volt battery bank.  What panels are you using? "

There are 2 identical strings consisting of 2 x 140w x 24v ( 72 cell )  , 4 x 75w x 12v wired in pairs for 24 v (36 cell ) ,  2 x 120w x 12v wired in pairs for 24 v ,(36 cell )  all  poly same manufacturer and age  , both strings run in parallel from the Array then 3 meters before the CC then they go into one larger 6AWG cable . The difference between the arrays is no more then half a volt open circuit approx39 -  40v .

Estragon , I assume you had loads running during the log period shown? Looks like the pv is putting out ~9a at high enough voltage to power loads, but not high enough to do much charging

The previous screenshot is when the panels are cold ,  late afternoon/early evening , load varies then from 50 to 250 watts ., Battery bank size is 600 amp/hr c20 is 468 amp/hr .
There is no load other then the inverter early morning prior to charging ,( screenshot attached ) but the same condition exists .

mcgivor

Looking at the screen shots, nothing makes sense, the battery voltage is sometimes higher than the array but still putting out current, other times it looks correct, the array voltage has to be higher than the battery in order to charge, it would be one thing if it were consistent, but it's not. Without relying on MS View, have you tried two digital volt meters, one on the battery and one on the solar imput, my guess is that you will find the array voltage is always slightly higher, when charging. Personally I've never been a fan of using a computer to get accurate measurements, the RM2  meter though,  is very accurate in a real time comparison, do you have one? 
Think in  the end you'll find the controller is doing exactly what it's supposed to do and the strange reading are  computer software related issues.

mcgivor

The other question re diodes, they are not required.

Ako

No , i dont have a meter . The voltage being lower than the Array is what concerns me also but looking at the Amp/Hrs they still show an increase during those times for some reason . The situation is always the same , array voltage lower than battery bank early morning and early evening which i would have thought was natural , has to start and end with nothing coming in but that wouldent explain why the Amps increase , not a lot but they do increase and its not the battery settling or anything like that as early morning its had hours to settle and evening there is a load but both times the scenario is identical , i did wonder if it only showed array and battery voltages but did nothing with them until the array ones became greater then the Battery bank . Tonight i will take voltages with a meter again at the battery bank terminals and on the battery and array terminals inside the CC but im sure they will all concur except for being around 0.2v lower then the msview voltages show them to be .

Why are diods not required , i was always lead to understand that without them the batteries would discharge backwards when the voltage coming in was lower then there voltages .

littleharbor2

+1, what Mcgivor said.

I am a little confused at how you get 2 "identical" strings. It looks like you have 3 not identical strings, unless each string has all the panels listed x 2. Either way you are costing yourself potential power. The lower amperage, (looks like the 140 watt), panel will drag down the output of the 150 watt pairs and 240 watt pairs to whatever the 140 watt is putting out.

 What, exactly is your method for connecting these panels into strings?

mcgivor

It seems that when the sun is low the readings are strange, but when producing good current they look good. Perhaps the swithing of the CC, PWM after all is just a switch, is sending pulses and the readings captured are the peaks and valleys therefore sometimes correct, sometimes not, just my theroy.

The internal circuitry of the CC  blocks the battery power from feeding back to the panels, adding additional diodes would only create a volt drop during charging, thereby reducing efficiency.

Ako

I hope you can make sense of the screenshot of panel configuration , each string has 8 panels ,  6 are 12 volt so wired in parallel into pairs of 24 volt so each string has 5 positives and 5 negatives connecting . In Bulk i have between 30 and 50 amps coming in depending on the sun and its relationship to the panels ,(  ie: sun comes up around 7.30am and wont shine directly on the panels until midday )  voltages just above the battery voltage in Bulk increasing to around 2 or 3 volts greater than battery when it goes into Adsorption , after a while the amps coming in tapers down .

Always thought diodes were only way to do that and when i read that there were none in the power path it seemed strange for M* not to use them when they use them in the Prostars however you say they are not necessary so thank you that answers my question .

mcgivor

littleharbor2 said:

+1, what Mcgivor said.

I am a little confused at how you get 2 "identical" strings. It looks like you have 3 not identical strings, unless each string has all the panels listed x 2. Either way you are costing yourself potential power. The lower amperage, (looks like the 140 watt), panel will drag down the output of the 150 watt pairs and 240 watt pairs to whatever the 140 watt is putting out.

 What, exactly is your method for connecting these panels into strings?

With PWM, as long as the voltage of all the arrays in parallel are appropriately the same, the current will be the sum total, not like MPPT where there would be losses, think @Ako has it correctly hooked up. 

Ako

Just been up the hill and checked , the 12 volt pairs doubled up are within a volt of the 24 volt panels , Phew !!!!! was a little worried then .

littleharbor2

mcgivor said:

littleharbor2 said:

+1, what Mcgivor said.

I am a little confused at how you get 2 "identical" strings. It looks like you have 3 not identical strings, unless each string has all the panels listed x 2. Either way you are costing yourself potential power. The lower amperage, (looks like the 140 watt), panel will drag down the output of the 150 watt pairs and 240 watt pairs to whatever the 140 watt is putting out.

 What, exactly is your method for connecting these panels into strings?

With PWM, as long as the voltage of all the arrays in parallel are appropriately the same, the current will be the sum total, not like MPPT where there would be losses, think @Ako has it correctly hooked up. 

The terminology may be wrong here and that is what is throwing me.

A STRING is a group of series wired modules. That being said the lowest amperage module will dictate the overall amp output of said string.

mcgivor

littleharbor2 said:

mcgivor said:

littleharbor2 said:

+1, what Mcgivor said.

I am a little confused at how you get 2 "identical" strings. It looks like you have 3 not identical strings, unless each string has all the panels listed x 2. Either way you are costing yourself potential power. The lower amperage, (looks like the 140 watt), panel will drag down the output of the 150 watt pairs and 240 watt pairs to whatever the 140 watt is putting out.

 What, exactly is your method for connecting these panels into strings?

With PWM, as long as the voltage of all the arrays in parallel are appropriately the same, the current will be the sum total, not like MPPT where there would be losses, think @Ako has it correctly hooked up. 

The terminology may be wrong here and that is what is throwing me.

A STRING is a group of series wired modules. That being said the lowest amperage module will dictate the overall amp output of said string.

There are 2×12V panels connected in series, creating 24V, 24V panels in parallel with those, all 24V so the current output will be combined, PWM takes what it can get as long as the voltage is equal, it won't be dragged down by the lowest string /panel, that is the beauty of PWM, it's simple and logical. For example you can use 6W panels with 300W panels, the 6W output will only add to the 300W as long as the voltage is the same, it won't drag the larger panels down, it will only add.  

littleharbor2

SO... the terminology is incorrect .There are not 2 STRINGS. There are 5 combined in parallel in each array. MPPT and PWM have nothing to do with it. MPPT controllers are PWM with MPPT tracking in which the higher voltage/lower amperage input is converted to the optimum voltage for the batteries needs. The PWM charging algorithm is no different.

Estragon

A theory - as the sun comes up, Voc rises to something like panel spec. This triggers CC logic to wake and run through a pulse cycle. The current draw during the pulse cycle pulls Vmp down to just below Vbatt (the controller itself being a load), recording Vpv as part of the pulse cycle. This repeats until there's enough sun hitting the array to sustain high enough Vmp to charge.

If this theory is correct, there is nothing wrong with what you see in logs.

Photowhit

I'm a bit lost, is your charge controller on a hill with the array and how far to your batteries?

I went back and check those panels are correct for a PWM charge controller, but they are also among the lowest VMP (More important number with PWM CC)some under 17 volts.

I see you mention a diode, do you have one in line now? perhaps all combined to make for low voltage, drop across diode, distance from charge controller to batteries, already lower VMP. Add that to the higher voltage your new Rolls batteries like.

In addition 780 watts (2-140 w= 280, 4- 75w= 300, 2-120w= 240) feeding a 468 amp battery bank should likely give you a max of 32 amps on perfect, cold winter days, and more normally 24 amps on normal days about a 5% charge rate (not sure where you are seeing 30-50?).  ...and you say the sun doesn't hit the panels until "midday" are you saying noon?

Looks like you are setup for chronic undercharging.

mcgivor

To keep it simple forget MPPT, PWM is what is in focus, don't want to get into an argument, there is too much science behind that door. In my early days of experimental findings, I had many panels of various wattage to charge a 12v battery, the more I added only increased the current output, all with a cheap PWM  CC, which I must add, still works to this day, 10 years later, despite donating it to someone who needed basic lighting. Understanding your rationale, there is much to be learned with both technologies, a PWM controller is a simple switch, an MPPT controller is much more sophisticated, similar in some ways, but very different.

Ako

Photowit , the CC is in a building with the batteries etc just below the Array , i dont have any diodes thats why i questioned if i needed to include any . When in absorption i am normally seeing about 34 volts come in from the array . I was unsure why the voltages coming in were lower then battery voltages only at the start and end of the day , during bulk they are normally kept one or two tenths of a volt above battery i think by the CC and allowed to increase to be up to 4 volts higher i think also by the CC as the increase is immediate and at the same time as the conversion from Bulk to Absorption . I am having trouble reconciling 780 watts , my total watts from adding up each individual panel irrespective of voltage is 1640 although at 24 volts i thought i had 1100 watts . I think your right though my input current is lower than ideal averaging 35 amps but by midday 12 ish im always into Absorption for the rest of the day , admit i dont often see Float but SG readings after the voltage in drops to less than the battery bank are 1.28 - 1.285 so i have taken this to mean that i have replaced the 2500-3000  watts drawn each 24 hours and my voltage shown before charging begins in the mornings is always 24.2 - 24.6 .with SG readings around 1.22 . I have considered buying more panels but i dont see any advantage other then to help out on the rare days we have poor weather . Possibly i have bought batteries to large for my consumption however my last set lasted 10 years so i worked on the basis that i should stick with what worked before . I do have a generator with a 20 amp charger and on the odd occasion when i have had very little coming in for a day i have used it and i find 2 or 3 hours early evening and the batteries will give me the same readings in the morning . When i say the sun doesn't hit my panels until midday ( noon 1200 hrs ) i meant that the sun comes up in the west and the panels face south so its midday before the sun is facing them directly . I start getting voltage and current from about 7.30am reaching around 20 amps after an hour and 30 - 35 amps 30 minutes later increasing only until i go into Absorption before midday , i only see 40 or more amps if part of the morning has been a bit dull and Bulk is delayed until after midday such as today has been , those times the sun is giving me more and the CC is still in Bulk to accept ,  the screenshot will show what i mean better . Today is a perfect example , by midday today i had only 60 amps come in so still in Bulk when normally i would have expected to have around 100 amps total in and gone into Absorption so amps would have been tapering off . Today i a still in Bulk when the sun is in the best position so im benefiting from the extra amps produced which i wouldent normally see .

Little harbour , your right , i said strings because i used a string calculator on one set of panels and it gave the result as 5 strings but as you point out its actually different sets of panels running in paralel , the only strings are the 12 volt ones wired into pairs .I think i confused things with saying 2 strings when actually i think there are 10 in total all running parallel , 4 single panels and 6  in pairs . Until a few weeks ago i had 2 identical sets of panels each with 8 panels running 5 positives and 5 negatives each set controlled by its own 30 amp controllers in parallel and into a bus bar . I replaced the two controllers with a single M* ts60 PWM so just before the ts60 i joined the 2 cable runs from the array to make for a single connection into the M* ts60 .

Photowhit

Ako said:

I am having trouble reconciling 780 watts , my total watts from adding up each individual panel irrespective of voltage is 1640 although at 24 volts i thought i had 1100 watts .

Does your statement of your array "... 2 x 140w x 24v ( 72 cell )  , 4 x 75w x 12v wired in pairs for 24 v (36 cell ) ,  2 x 120w x 12v" correct?

If so you have; ...820 watts (2-140w= 280, 4- 75w= 300, 2-120w= 240) 280+300+240=820 watts sorry, I ran the numbers in my head real quick, but I did "show my work" ...lol.

Perhaps we there is something we missed.

I find this interesting "...admit i dont often see Float but SG readings after the voltage in drops to less than the battery bank are 1.28 - 1.285 " Something doesn't sound right.

This definitely sounds wrong...lol "...that the sun comes up in the west and.."

I do understand, just having a little fun, I normally see close to max charging a couple hours either side of "solar noon".

Photowhit

FWIW - I understood and am comfortable with you using "5 strings", Like batteries, solar panels are made up by 'stringing sets of cells together. So 2 - 36 cell panels in a string to make a 72 cell string is pretty much the same as a string of cells in a 72 cell panel.

Ako

Does your statement of your array "... 2 x 140w x 24v ( 72 cell )  , 4 x 75w x 12v wired in pairs for 24 v (36 cell ) ,  2 x 120w x 12v" correct? Yup , its correct 

If so you have; ...820 watts (2-140w= 280, 4- 75w= 300, 2-120w= 240) 280+300+240=820 watts sorry, I ran the numbers in my head real quick, but I did "show my work" ...lol.  You and me both haha

I find this interesting "...admit i dont often see Float but SG readings after the voltage in drops to less than the battery bank are 1.28 - 1.285 " Something doesn't sound right. The float thing is no big deal im told due to the way the M* ts60 CC makes the calculations . I take SG readings after the charging has finished at night and before it starts in the morning as i understand the SG is a better indication of SOC than voltage alone .

This definitely sounds wrong...lol "...that the sun comes up in the west and , Good to see your awake and spotted that deliberate mistake , of course as everyone knows it actually comes up in the East . hehe

I normally see close to max charging a couple hours either side of "solar noon . I will only when the circumcisions are right , ie: the sun is shinning and i am still in bulk or not long into Absorption , unfortunately by the ideal time for max current in i have been in Absorption for a while so dont often get to see the higher inputs , if i were to disconnect my array until around 11am then i would see more amps coming in as Bulk would allow it all through but for me its already started tapering down in Absorption by that time . I am considering adjusting my Float cycle limit to 40% . I still have it at its default factory setting of 30% ,  duty cycle reaches 30% or less for a cumulative hour so raising it to 40% should help .

Estragon

I think it's pretty normal for most of us off-grid to not see full current most days. As long as batteries are getting charged I don't worry about it much.

Could it be the controller isn't going into float because loads are taking part of the duty cycles?

Ako

Could it be the controller isn't going into float because loads are taking part of the duty cycles? Yes im told by Morning star its common never to go into float for that as well as other things , for instance if the voltage at night drops below 25 volts then the duty cycle is extended for an hour making 2 hours so in my case would mean everyday would need duty cycle of 30% of less for 2 hours and taking into account that there are also loads it will be rare to see it floating .but im not concerned , voltages and SG  give me confidence batteries are getting the charge they need within the charging period and not being cycled very deep . The indications that voltage and SG  give lead me to feel that my batteries are getting a full charge or as near as makes no difference .

From everything im picking up on this forum im becoming more confident through knowing and understanding more .

Ako

Starting to feel like pulling my hair out here . Thought it would be a good idea to take actual voltages reading from the arrays at the CC terminals while in Adsorption , msview showed 34v , voltmeter on array terminals in CC in showed 38v . I have 2 sets of panels and each has its own cable into the building where CC , inverters etc are , once in the building they join into a single larger cable for the last 3 meter run into the CC so i unplugged one of the cables and took an open circuit reading on one bank of panels , was 39.9v , the remaining cable from the second set of panels left connected showed 34v at CC array terminal , msview 33.9v  Plugged that cable back in and unplugged the other cable , 40.1v open circuit as both sets of panels are identical there is never more than half a volt between them so that was fine , array voltage at the CC terminal of the first set now plugged back in and showed 34v , msview 33.8v, then plugged the second cable in so both back online , 38v again with voltmeter , msview showing 34v again so back to where i was before i disconnected anything . Dont understand why with both sets of panels plugged in the voltage at array terminals in CC and at the point they connect from 2 individual cables into 1 is 38v but each individual set only 34v and also why msview shows 34v whatever is connected , 1 set or 2 . The CC was in absorption at the time showing only about 9 amps coming in as it had been in Absorption for a while and tapering down .  Its my understanding that if you join 2 positives and 2 negatives both with the same voltage and amps running through them then the voltage will remain the same but the amps will double although i know i wouldent see an increase in apms unless i was in Bulk , in Absorption there tapered back .There is no problems with any wiring or cables , inspected them before anything else and they have been in long time already without any issues .

mcgivor

As long as the batteries are happy and receiving a full charge, the SG readings and voltages  are good etc etc leave your hair alone.....luckily you still have hair you could pull out if needed, don't sweat the small stuff. 

Ako

Thing is i do worry , i like to understand how and why things are happening or working a certain way especially when it feels wrong . Ignore something and it comes back t bite you in the butt when something goes wrong and your told its because of something that was happening that you knew about but dismissed it as not important . Recognising signs should be a good thing , reminds me something i was told a long time ago , " death is natures way of telling you to take it easy "

BB.

Sometimes (don't know the MS internal circuitry) the numbers displayed are estimates made from other numbers--(i.e., they may not measure voltage everywhere, but use equations like P=V*I*assumed efficiency). Many charge controllers seem to be ~5% accuracy for voltage and current readings, and 10% accurate for power (i.e., 5% voltage error + 5% current error = ~ 10% error in power).

Or measurements made across shunts that are not all that precise. I.e., a good, high current, shunt may cost $25 to $75 retail... (computer trimmed for accuracy, materials chosen so the resistance does not change over time and temperature, Kelvin contacts to measure the low voltage drop without errors due to resistance and current flow, etc.).

https://www.solar-electric.com/residential/meters-monitoring/shunts.html

There is also a difference in, high current, AC measurements vs DC measurements. AC current measurements can be made with a "current transformer" -- I.e., 100 amps on one side of the transformer can produce 0.10 amps on the other side (1,000:1 turns ratio). A DC current (long term accuracy) is usually done with a precision power resistor (more expensive, more heat loss/less efficient).

In the charge controller, they may just use a copper trace or a 2" wire of some sort (or even the voltage drop across a FET) to get a "useful voltage" that can be compared during MPPT "sweeps" (to find the Vmp/Imp of the panel at that point in time). Long term (repeatable readings over days/weeks/months/time temperature/operating points) accuracy does not matter to the MPPT calculations. Only the A/B comparisons for that second in time.

The manufacturer is not going to put a bunch of high precision components in their meter when it does not make it operate any better (on average).

-Bill

Ako

Thankfully my ts-60 is PWM and im struggling to understand some of the things it does , with MPPT i dont think i would have any chance . Had an invite for a webseminar from Morningstar on March 28th so will sit through it and see if there is any mention of the issues im experiencing . The voltage thing i noticed today is nothing to do with either the CC or msview , it was just something i noticed on the cables coming in from the panels and present not only on the array terminals in the CC but the cables themselves , I know that if you connect 2 identical power sources in series the voltage should remain the same , dont it with pairs of panels and batteries many times , today i noticed they were only identical inputs when connected individually on there own and only after being connected together they increased the voltage by 4 volts .

BB.

PWM measurements of voltage and current is even more "exciting". Basically, you have some waveform that (unless bulk/maximum charging current) is some frequency of on/off current flow. And the (typically digital these days) measuring devices may take measurements during the peak/valley/wondering over peaks and valleys and give you strange readings.

Aliasing (i.e., high frequency wave form, low frequency sampling example):

https://www.dataq.com/data-acquisition/general-education-tutorials/what-you-really-need-to-know-about-sample-rate.html



To measure non DC and non 60 Hz AC wave forms (somewhat) accurately, you need a "True RMS" reading meter:

http://en-us.fluke.com/training/training-library/measurements/electricity/what-is-true-rms.html

Is this the "issue" you are seeing--I am not sure--But it could be part of your measurement errors. With a smaller array, the charge controller is in "bulk" (100% on). And with a larger array, the controller is cycling at 50% (or whatever) "on time" or duty cycle--And somebody's meter (yours, the controller), is not "sampling" the average, but reading peak voltage or some variant of it.

-Bill