Hi All. Am I right in thinking that MPPT cannot be achieved by PWM control lers ? PWM will have massive current and voltage ripple, effectively makin g the incoming voltage switch between the o/c voltage of the panels, and sl ightly higher than battery voltage. IMO, an MPPT controller would have to utilise a switching DC-DC buck circuit to minimise ripple on both input and output sides.
Thanks
That's what capacitors are for. and slightly higher than battery voltage. IMO, an MPPT controller would have to utilise a switching DC-DC buck circuit to minimise ripple on both input and output sides. A MPPT IS a dc-dc buck converter that stabilizes current, not voltage. Current is adjusted for max...as long as your battery isn't full.
[ snip ]
Low-power MPPT controllers may be able to dispense with the need for a low-impedance source and destination, but higher-power MPPT controllers absolutely do need high-value input capacitors. In fact, that's true for any switching converter connected to a solar panel. All switching converters use PWM. The MPPT part simply adds extra schemes to help determine the optimum PWM value.
--
Thanks,
- Win
I'd like to hear more about that. How does low-power matter?
There is no relationship with power level. I have done MPPT for kW space solar arrays
Just connect the buck without any cap directly to the array. The frequency or inductor must guarantee CCM with small perturbation so you will stay on the MPPT point. Use outer loop to sweep to find MPPT
Cheers
Klaus
It's a strong function of the type of PWM chip you're using. If the chip is especially designed for MPPT, yes I imaging it'll be very careful with its maximum inductor charging current, but ordinary buck converter ICs trying to get a high output from the panel, will create short-term excessive current draw from the panel, quickly pulling its voltage below the UVLO shutoff threshold for the IC, which then goes into soft-start mode. Adding capacitance across the solar panel can solve the problem. At least enough to handle 10 or 20 switch cycles.**
If the PWM chip's power output goal is well below the panel's capability for its solar level, the UVLO fault won't happen and no capacitor will be needed.
The issue here is that most buck-converter ICs will not let you fine tune the input current limit,** let alone with a control loop. Yes, I agree, this is no problem for a specially designed MPPT system.
By "simple common" ICs, I mean low-pin-count small- package parts. This includes most of the ICs in AoE's three tables of switching-converters. Hey, I regret not adding a programmable input current- limit to the ten feature list columns. I could say I forgot, because datasheets don't mention it, since they don't have that feature! However, I had repeatedly run into this issue when selecting ICs, over the years before I made the tables five years ago. Yes, many parts include a fixed switch current limit, but this is of little use in designs where the value is not matched to the application.
--
Thanks,
- Win
trollers ? PWM will have massive current and voltage ripple, effectively m aking the incoming voltage switch between the o/c voltage of the panels,
A big capacitor, esp when using a 1kW array. There is a huge difference be tween PWM and buck switcher. In a switcher, current ripple is small, in a PWM, it switches between zero and charging, whatever the direct-connection charging current would be. I asked the question because there are a lot of PWM units around that claim to be MPPT, but couldn't be. They just keep t he average panel voltage at MPPT while swinging wildly each side of it. Ca pacitors on the panel side do help, but only if there are a lot of low ESR ones, and low impedance between them. That would need screw-terminal ones, and quite a bit of wiring.
Buck converters use inductors to store energy. The chopper is on the low side of the inductor, so that there is always current flowing on the input side.
Thinking of boost converters?
--
Thanks,
- Win
Both of them use inductors, but I was wrong about input current in a buck c ircuit, it is on/off rather than just ripple, unless you have 2 or more, in tereaved. I am working on a 3-phase interleaved circuit for another projec t. That would seem to be the way to go, and I have seen MPPT units which h ave 3 inductors, so I presume that is what they do.
For MPPT you need to run preferably without any capacitance
The MPPT function sweeps to find the MPP, and any capacitance will phase shift the voltage which may lead to operation at the wrong point
Cheers
Klaus
I need some detailed education here. Got a schematic?
For any inductive topology I can imagine, you're either 'charging' or discharging the inductor.
You can cause the battery charge current to be the MPPT current. But, for some part of the cycle, the panel is not stuffing current into the load. That would seem to be a serious efficiency loss something like the ratio of battery voltage to solar cell voltage. You'd be better off with a direct connection. You'd be well below the maximum power point, but still at greater current.
What I want is charge current = MPPT solar current X (vsolar/vbattery).
I can imagine using tapped inductors or transformers for fixed conditions, but, for the wide range of solar voltage and MPPT points, that doesn't look attractive compared to a capacitor.
What am I missing?
Isn't there an optimum voltage and current to get the max power out of the solar array? Won't any voltage or current ripple push it off that point?
Caps are cheap.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
See fig 1. If the inductor is reasonably big, the output is constant-current, with a bit of ripple maybe. The input current is zero when the switch is open.
MPPT of the solar cell requires constant voltage and current; there is only one MPPT point, so stay there. Just put a capacitor on the supply side, across the solar cell.
The load side typically has a cap too. Even a battery will benefit from that.
EMI and spikes would be awful without caps close to the switcher on both ends.
If the array voltage is low, a boost or buck/boost converter could step up into the battery.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
I evaluated a half-dozen MPPT converter ICs, using their Eval boards, etc. None of them worked that way, but then each had a completely different scheme. The scheme Klaus described sounds like a linear-circuit approach. Some MPPT converters turn off the power- conversion process while taking a quick measurement. Some take measurements frequently, others not.
My scheme wasn't even MPPT, but was probably within 10% or so of MPPT, during overcast conditions, which was when it mattered -- days of rainy cloudy weather. During bright sun, the battery would quickly charge, and I had no further use for the extra solar energy.
--
Thanks,
- Win
the difference is one inductor and one diode.
easily fixed as above on the output and a capacitor on the input.
they can easily be MPPT.
capacitor size depends on the PWM freqiency.
-- When I tried casting out nines I made a hash of it.
An MPPT algorithm to control a buck switcher should be simple. Just dither a tiny bit to see where you are.
Of course, if the battery is full, or soon will be for sure, you don't care.
Do you keep bees? We had two hives in the back yard and got great honey all year. We got stung now and then, but one time Mo had a systemic reaction so the bees had to go.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
shift the voltage which may lead to operation at the wrong point
I don't follow this either. I suppose you are suggesting that the ditherin g that is done to sample the power level with varying operating point is an AC on top of the DC. The added capacitance can potentially alter the meas urement of the AC which is what allows you to determine the MPP.
This should be workable by use a rather high rate for the PWM while using a significantly lower rate for the MPPT. Then the capacitance won't impact the AC measurement allowing the MPPT to operate correctly.
What I am confused about is that in a buck circuit, if you don't have capac itance at the input, the current from the supply (the solar cells in this c ase) truly is on/off.
Actually, using the dither created by the PWM on/off action and measuring t he current on the solar cell side of the input capacitor rather than betwee n the capacitor and the switch should provide the AC signal to allow you to find the optimum power point. No? I don't see how that would not work.
Someone pointed out to me once that if you are charging batteries all you n eed to measure is the current really. For small changes in the current the battery voltage does not vary much, so the current maximum is very close t o the power maximum.
Rick C.
A PWM'd switch (with no inductors) between a stiff source (a cap bypassed anything) and a stiff load is no better than a linear regulator.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
controllers ? PWM will have massive current and voltage ripple, effectivel y making the incoming voltage switch between the o/c voltage of the panels,
ld
between PWM and buck switcher. In a switcher, current ripple is small, in a PWM, it switches between zero and charging, whatever the direct-connecti on charging current would be. I asked the question because there are a lot of PWM units around that claim to be MPPT, but couldn't be. They just kee p the average panel voltage at MPPT while swinging wildly each side of it. Capacitors on the panel side do help, but only if there are a lot of low E SR ones, and low impedance between them. That would need screw-terminal on es, and quite a bit of wiring.
Linear regulators produce quite a bit of heat. Switchers and PWMs don't.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.