Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
isochronic Guru Joined: 21/01/2012 Location: AustraliaPosts: 689
Posted: 01:55pm 29 Jan 2016
Copy link to clipboard
Print this post
I don't understand something about MPPT controllers. (and battery chargers)
MPPT controllers are supposed to hold power draw from a solar cell bank
at the maximum power amps/volts combination. So why do they often
use a modulated pulse width of management? It seems to me although the
average is no doubt accurately close to the MPPT, the actual draw is
usually a higher value repeated pulse interspersed with zero or
lower values - ie the instantaneous values are only roughly near the MPPT.
So the actual power draw is probably not the MPPT but just looks like it.
Similarly for battery chargers - although no doubt the average charge rate
is right, it is often a pulsed higher current, and the instantaneous
charge rate is a lot higher. Discharge as well is usually a rapid switched load.
So the actual current experienced by the cell would be higher.
Has anyone actually run some tests on these things ? I guess you would need
a bank of PVs and cells, connect it up to run near the MPPT naturally, then
switch in a MPPT controller and see if made much difference.
(ed) maybe someone could put a cro on the pv bank and check the waveform ?Edited by chronic 2016-01-31
Downwind Guru Joined: 09/09/2009 Location: AustraliaPosts: 2333
Posted: 04:25am 30 Jan 2016
Copy link to clipboard
Print this post
I understand your miss concept, But you really need to get your head around how batteries and current loading works.
Regardless of what voltage is supplied to a battery it will soak it up as current, but there needs to be a controlled limit of voltage to suit the battery voltage.
That is where your MPPT comes in to control current Verses voltage.Sometimes it just works
isochronic Guru Joined: 21/01/2012 Location: AustraliaPosts: 689
Posted: 01:52pm 30 Jan 2016
Copy link to clipboard
Print this post
I thought the idea of MPPT was not so much to limit the output voltage
maximum (eg to prevent battery overvoltage) - but to operate the solar cells
at their maximum power output ?
ed- eg when the solar cells are connected to the gridEdited by chronic 2016-01-31
oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686
Posted: 12:42am 31 Jan 2016
Copy link to clipboard
Print this post
You have the general idea correct.
The cells ( solar) are really constant current devices... not constant power.
This means ( simplified) that the current will not increase as the impedance of the load decreases... unlike say batteries.
They are diodes. As each photon hits the diode substrate, it displaces an electron, sweeps it across the PN barrier, and goes through the load to return to the other side again..... this means more photons.. more electrons.
The electric field is a constant ...... depending on the materials... but usually around .5v ( internal resistances and temerature changes this, but essentially thats it )
So to make a 12v panel we use 36 cells for a nominal 18v.
If we connect to a battery, we get say 12v... the panel gets pulled down to the battery voltage as it has a low impedance, and a 5X5 cell will give about 5 amps or more.... but does not markedly increase as the load impedance goes down... short circuit is still about 5 amps mark.
So we have lost 6v@5amps...and we want it back.
If it were AC, we could use a transformer, and the 18v x the 5 amps would mean we had a source power of 90w, so we could transform the 90 watts @ 18v to 90watts at 12v... thats what transformers do... but it's DC, so we need a buck converter to do it instead.
Here we use a switch and an inductor. We configure so that we dump the current of the cell into the inductor, it stores it as a magnetic field, and then we turn off the switch.... and the field collapses, the energy is released there is your quasi pwm..... we have a DC transformer... it's all about time and energy
The pulse spacing gives you the new lower voltage but at a higher current to keep the power equation healthy ( some losses involved)....but acts like a transformer. Mark space ratio gives the voltage differential from in to out.... but power should remain about the same.... so current rises.
We can use this to keep the cell voltage up in the right part of the graph.. ie near the end of the horizontal curve.
Knowing that the voltage is essentially fixed, we need to keep it where it will be in the best zone... ie near the 16-18v mark in this experiment.... not the battery voltage.
As the photons get less (less light) we can change the mark space ratio to keep the voltage high, but still the product of the falling current and higher voltage in the buck converter gets the maximum power out of the panel.... ie is the current drops from 5 amps to 4 amps, if we keep the voltage high, the product ( power ExI) is still higher than letting the current sag with the same battery voltage.
So yes, we use pwm. The resultant current pulses add up to a higher average than we would achieve without the buck conversion and constant current, as the power in = power out, rather than just original current in = original current out.... at a lower voltage.
PWM battery chargers use pwm instead of linear voltage regulators... thats the crux of it. If they are online chargers they will smooth the ripple( caps), so you won't see the pulses at the battery.
Solar pwm controllers ( not MPPT) will not bother with the ripple filter, and the CRO will see quite high voltage spikes ( very short before the battery damps it).
As an aside, I dont ever use mppt controllers, as I find it is actually better to use 60 cell panels instead, and a pwm controller.... much better performance for the same money.. ie 100 amp controller I can make for $50.... the other 800-1000 dollars I don't have to spend on mppt controllers of that power, can buy more panel anyway ( about 1500w ).
But the difference of mppt with 60 cell panels direct connected to batteries is barely useful ( used 2 CSL 60A mppt as test http://www.ebay.com.au/itm/GSL-MPPT-60-2-Solar-Regulator-Controller-Prog-60A-12V-24V-48V-PV-Battery-Charger-/26130608424 4?hash=item3cd70e6b94:g:U0QAAOSwYHxWQWzD)
The 72 cell panels can usefully make more power with mppt, but the differentials with 60 cells is not enough to warrant it I think.
For solar pumping units I use pwm to keep the voltage at a set point, and we can then use the brush dc motor as the inductor, as we can get quite high currents from a low current panel group.. have seen 40 amps from 10 amp panels... this helps starting too... it is slightly different to the technique used for current boosters in the same application.
Solar boosters rely on discharging caps at a certain voltage plateau. The stored energy is released in a short time so the power surge is substantial... enough to break inertia, and start the motor rotating
.............oztulesEdited by oztules 2016-02-01Village idiot...or... just another hack out of his depth
isochronic Guru Joined: 21/01/2012 Location: AustraliaPosts: 689
Posted: 12:13pm 31 Jan 2016
Copy link to clipboard
Print this post
Thanks oztules - that was very useful !
clears it up a lot
I have a remaining question, why is 18v used ? I guess the 18v
is too useful for power otherwise (?)
My thought is,
if the solar cells are used to charge batteries, they could
be configured to hold voltage just at the battery-charged
point, eg 60 cells could be connected as 2x30 and twice the current
at about 15v
oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686
Posted: 03:54pm 31 Jan 2016
Copy link to clipboard
Print this post
The typical boron phosphor doping of silicon yields the .5v.... but that semiconductor is not a pefect conductor, so introduces it's own problems.
( to fully understand it, you need to brush up on quantum mechanics and band gap theory to see why they are frequency dependant, and why the field is as it is etc)
In weak light the terminal voltage of the panel seems to decrease, as so makes everything I said about constant electric field across the PN junction wrong.... but it isn't......
We need to push our current through the substrate, and this resistance starts to become a bigger part of the equation... so E=IR... and as I drops ( less photon action ) R remains the same(ish), so E must drop too.... and as R is a constant we can say E is proportional to I.
Thats why we use the PWM to control the current coming from the panel.... to keep E higher, as we know it is a constant current device with resistance.... ExI will be better with mppt.
But in days of old, when no buck converters were common, we off set the voltage drop by putting in more cells to try to keep some current available above the 14v needed to fully charge a battery ( and in line diode drop too )
60 cell panels ( for 24v systems ) or 30 for 12v, means we will probably not require a controller on small systems.... electric fence manufacturers did this, so if you see an old Gallagher fence panel, it will have only 30 cells in it,( and some even less) and no controller for the unit... as the 15v is not much above the fully charged battery voltage, and as it is cycling, 15v is fine, as there will be very little current flow at 15v.
With the advent of buck and boost technology, and grid tie particularly, any number is fine, and I see far more 6x6 8-9 amp cells in 60 cell packages, than I see 72 cell 5x5 stuff like we did before.. we go from 5-6amps to 8-9 amps with no other changes, and so we really don't need mppt as much as we did for this range, just pwm to control fully charged.... and the voltage differential is not enough to bother to use the mppt current converting capabilities, unless the battery is flat.... other wise charging is between 52-60v for my system...( or 13-15v) never see it below 52v while there is any light at all about.
Hope that helps.
There may be another reason, but I don't know what that might be.
..............oztules
Edited by oztules 2016-02-02Village idiot...or... just another hack out of his depth
isochronic Guru Joined: 21/01/2012 Location: AustraliaPosts: 689
Posted: 11:28am 01 Feb 2016
Copy link to clipboard
Print this post
Ok...So the logic is, for a given battery storage capacity, the solar cells
are selected so as to give a reasonable output at say mid-morning and afternoon
not just at peak daylight. They are wired up in series to give higher than battery-charge voltage, and the excess
voltage (at say noon) is downconverted to extra current instead. The series
connection means less current and resistance drop in cables etc.
That is pretty reasonable. BUT the panel doesn't store any energy -
so if the buck converter pulse-connects with say a 80% duty cycle
it can't be more than the 80% useful...I think they should use a design
with a proper transformer and use the ~100% instead.
[ I was going to use "efficacy" instead of "efficiency" - but it leads
to describing systems as "efficated" - which I will reserve for descriptions
of some political systems - which really are completely "efficated"
if you ask me ]
Edit - I see the SC MPPT project includes large low esr capacitors on the input from the panel - so they would store power no doubt, but there will still be momentary low impedance loads to the panel. The output to the battery has small capacitors (ie bypass) so I guess the output charging is choppy (?).
It would be nice to see how much spike/ripple there is on the input and output,
or on similar devices - anyone with a panel MPPT and cro ??
Edited by chronic 2016-02-02
oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686
Posted: 03:36pm 01 Feb 2016
Copy link to clipboard
Print this post
"
"That is pretty reasonable. BUT the panel doesn't store any energy -
so if the buck converter pulse-connects with say a 80% duty cycle
it can't be more than the 80% useful...I think they should use a design
with a proper transformer and use the ~100% instead. "
No.... thats not it..... the capacitors on the input and the inductor stores the energy..as you say panels cannot. The caps get pummelled fairly hard in this application too.
An 80% duty cycle will give a corresponding decrease in voltage, and a corresponding increase in current.... power in = power out... less switch and copper and ferrite losses... maybe around 95%
The transformer you mention does it exactly the same way.... primary stores the energy in a magnetic field, as it collapses it creates a changing field that the secondary is subjected to ( same as in your alternator.. but no moving parts), and this induces an EMF in the secondary.... once again power in = power out less hysteresis, eddy current and copper losses.... maybe around 95% too.
The buck is about as efficient as a transformer. The transformer has no switching losses, but does have iron, and copper losses and leakage. The buck has the same family of losses, but also switching losses.. but eddy, and hysteresis and leakage will be much much less also..... and of course you can't use a transformer on DC.
You need to take into account everything that is happening to get the gist of it, but it does work.... pretty darn well too.
Perhaps google buck converters to get a better understanding of their workings.
The buck is a DC transformer for want of a better word.
............oztulesEdited by oztules 2016-02-03Village idiot...or... just another hack out of his depth
isochronic Guru Joined: 21/01/2012 Location: AustraliaPosts: 689
Posted: 09:48pm 01 Feb 2016
Copy link to clipboard
Print this post
Please measure the current and voltage from a connected panel,
before and after, eg without and with, a mppt connected ?
I'll leave it at that - I need data !!
oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686
Posted: 11:10pm 01 Feb 2016
Copy link to clipboard
Print this post
Sorry, I can't do that for you... I gave those away years ago to people who were running 90varrays to 24v.
Maybe someone else can
........oztulesVillage idiot...or... just another hack out of his depth
Tinker Guru Joined: 07/11/2007 Location: AustraliaPosts: 1904
Posted: 11:44pm 01 Feb 2016
Copy link to clipboard
Print this post
I can tell you that the solar panel current into my MPPT controller (Outback flexmax 60) is *always* smaller than the current this unit feeds into my battery bank.
The solar panel voltage is, of course, always higher than the battery bank voltage during the charging, absorbing or floating cycle.
From memory, today at the bulk charging stage 20A flowed into the MPPT and 25A flowed out from it into the batteries.
I cannot tell you the exact current/voltage at present as the sun is about to go behind nearby houses and the charging for today had been completed.
I am not willing to disconnect my MPPT from the circuit, I value my lithium battery bank too much for exposing it to a direct charge from the panels ( 1.6Kw of 72 cell types).
If you are so desperate for data I suggest you rig up your own MPPT system, you might be pleasantly surprised.Klaus
Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
Print this post
- which really are completely "efficated"
