Originally Posted by hellosailor
1- Why pwm charging
hasn't taken over the world: I suspect because battery chargers fall into two markets. The main one, which is based on "CHEAP" with no other criteria. Batteries are way above the comfort level of most consumers, they can't even figure out which end goes up in a remote
control. Witness Apple, who make many friends by requiring them to replace a phone
instead of changing the battery in it.
And the second market, where performance counts and price
is not a criteria, but that's the tail which will never wag the dog. And that's where MPPT
is for now.
, ( ie proper PWM, not solar
PWM ) hasn't taken over because there are little quantifiable gains , greater design challenges and often prohibitions from battery manufacturers. In today's electronics cheap
or dear is often set by marketing
departments not the price
of the components.
isn't PWM chargers. Which mppt controllers are using PWM charging. Often they do so because you have switched mode conversion anyway and its cheap
to pulse charge. , but I don't think any of the consumer mppt units are going anything other then 2-3 stage DC charging.
Apple do not allow replaceable batteries , because of Li technology. Firstly the battery lasts long " enough" secondly Li charging means the battery needs to tightly coupled to the electronics
. This is a trend in all Li products. Its nothing to do with the issue of changing batteries ( seen am apple mouse, trackpad etc recently !)
2-A data logger or scope will NOT tell you the algorithm really being used, until you have made many traces and analyzed much data. Sure, you can see the conditions at which point something changed, but that doesn't tell you the why of it, or show you how the entire logic table is populated. And these days, a chip with a logic table in it is what, two cents? five? Who's going to spend the day analyzing it?
The only reason I know what one system was actually doing, was because I set up the monitors and took notes on a day while I was there doing other things. Giving me enough information to ask "WTF?" but not enough to figure out the whole program.
can only use current and voltage , you can monitor
that as well, equally the science of LA charging is well documented. You can determine enough to know if you are looking at 2-stage , 3-stage , pulse , absorption cutoffs , safety
timers etc. sure you can't read the code, but you can see the method the device is using to charge the battery , after that who cares.
3-"How is the front end charging algorithm any different for any three stage ( or 2 stage ) charger." From watching dual voltmeters and ammeters and noting the readings every 15 minutes for about four hours, I saw that the charging VOLTAGE to the batteries was being ramped up about 1/10th of a volt every time I turned around, while the amperage was coming slightly down. There is no 3-stage charger that continuously adjusts the voltage and amperage during the bulk charge phase, is there?
AFAIK every 3-stage charger is applying one set voltage during each stage, and any variation in that voltage is simply the result of something being saturated. Whether that's an analog charger or a digital one, whether the "control" is from regulation or from battery condition, the design of a 3-stage charger, even a "smart" one, is not going to keep changing the charging voltage suring each stage. That's different.
Do all MPPTs do this? I don't know, you'd have to ask them.
There are many methods to implement constant voltage and constant current systems, some may step , others may be continuous analog, often the " step" you saw is that the regulator
is digitally controlled by a micro and the voltage is set in preset steps rather then an analog constant current. At the end of the day , its just CC charging
In practice battery resistance sets voltages not charger circuits. But what you describe is again just a variant of CC CV charging , what you described is also a function of simple mppt design where the charging voltage control is used to regulate the input to the mppt so as to move the panel operating point. ( ie simply letting max current flow would drag the controller away from the panel Vmp ) , so in mppt design , you must control current through every part of the charging process so that so have the ability to actually load the panel correctly to achieve Vmp. This is because some low cost mppt implementations don't have true DC DC converter isolation between the input mppt process and the charging process
Again there no magic. The charge method is a function of the mppt design but its just 2stage charging
Whether a charger is going to use "pure" DC, or chopped DC, is partly going to be intentional and partly just the result of what design is cheap. Transformers, rectifiers, a cheap 3-pin regulator chip and some bootstrap transistors were the cheap way to build a charger 20 years ago. Today? The price of the iron and copper in the transformer makes them unaffordable, and those expensive IC-controlled chargers using logic tables become cheaper, way cheaper, so a whole range of options is there. Either by default or intent.
The costs of toroids today from China
is cheaper then I was buying
them from the UK 30 years ago. The change over has nothing to do with transformers
Today switched mode is popular because its far more thermally efficient ( and hence more efficient ) , lower weight and more controllable. Modern ICs means the considerable circuits needed to support good SMPS can be easily design in as opposed to numerous discrete components otherwise needed. SMPS is simply a technically better way , especially in high powered designs., that's why it's prevalent. This has nothing to do with pure DC or pulsed DC charging , the output of most SMPS is smoothed to DC, proper high power pulse DC charging of big LA banks is quite uncommon and not something in see regularly.
But the MPPT market? Catering to folks who have probably already spent a thousand dollars on solar panels and anothousand on batteries and interested in spending more as needed to max out their performance? In Bill Gate's immortal words "Hardware is cheap" and sloppy programming even cheaper.<G>
Sure , there is lots of snake oil
in the mppt market. How many users actually validate that the controller actually does Vmp tracking, how many actually verify charge controller performance. But then again this is true in many areas , so what , I'd retort.