Improving both the output capacitance and its ESR equals to less ripple on the load.
Cheers ^:)
Posted on: Fri, 09/23/2016 - 14:42; edited for a little fixup.
Improving both the output capacitance and its ESR equals to less ripple on the load.
While true it will not affect the ripple if the buck converter module itself has a ripple of 10. It will still wait for voltage to drop 5 below target and raise it back up to 5% over target no matter what you do to the circuit.
In order to improve that you would need a better buck converter IC that has less ripple and a suitably large enough inductor to be able to support it.
Unless of course you are saying to put a large capacitor after the sense resistor which would have a minor effect on ripple but due to limited space the size of the cap would be very limited and as such it would only mildly reduce ripple on high current driven LED’s. Still a good thing to add though if you have the space.
Although that in a simple addition to the end of the circuit and has nothing to do with the PWM?
You are inviting bicycle.
The only known way to improve buck efficincy is using IC with PSM mode.
But I suppose that 1% mode in real buck driver on 17-20mm board all the same will be less effective than 10% PWM to 1x7135.
https://www.startpage.com/do/search?q=%22inviting+bicycle%22 -> click/tap on images (lewdness warning 
). O:)
Cheers ^:)
Ok, so nickleflipper that's interesting. SO that setup seems to actually be exactly what I described in the OP. If that's the best there is, then bad news. I'm optimistic that FPC is right and there are much better. I mean there is no reason there can't be better ones.
TA no, I didn't talk about the PWM inside the buck because it wasn't the point. I'll try to summarize the short, or at least more direct, version. A buck converter is fine, but if it's just regulating off one fixed sense resistor to either a fixed current or voltage, that doesn't give you modes. That just gives you your 100% level. Now how do you get modes? That's the question, and how to do it efficiently.
The MTN-MAX drivers actually turn the buck on and off to get low modes. This is nothing like the internal switching and it's at a much lower frequency. This results in square pulses to the LED, just like an FET driver and has the same inefficiencies as an FET driver due to driving the LED at higher peak current than needed.
You could instead just linear regulate down to the modes and you get the linear regulation inefficiency. All this on top of the buck losses, adds up.
So instead you could try to get the buck to output less current for lower modes, after all it's a buck converter. That's what we're talking about. The last use of PWM here is the one that still comes from the microcontroller, but doesn't operate an FET. It instead gets filtered into an approximately DC level to feed into the bucks feedback controls. This is shown in the image nickleflipper referenced above, but it seems it didn't work all that great in that regulator.
Since buck regulators can have trouble going to low very low currents it was proposed that the buck could bring say anywhere between 20% and 100% and to get to 5% you use buck throttling plus PWM.
I think you are misunderstanding what is happening. It is already doing what you describe internally in the buck IC. It outputs a constant current, it does NOT PWM the current past the 10% ripple.
The output from the MCU goes to the dimming pin on the buck IC. The buck IC sees this signal as a matter of duty cycle. Whatever duty cycle it sees in that signal, it then cuts the output accordingly. The MCU DOES NOT turn the buck IC off or on at any point.
So if you have it setup for a 1A output and input a 50% duty into the buck IC, then it will output a .5A constant current output.
The PWM out from the MCU is simply to tell the buck IC what current you want relative to the max. The buck you listed in the OP has a dimming ratio of 5000:1 IIRC. Which means if setup for a 1A output, it can dim all the way down to 0.2ma, far less then most moon modes.
Efficiency wise, it will be just as efficient at that mode as it will running 1A in a theoretical world (It will actually be slightly different due to the lower VF of the LED at that current but that is balanced somewhat by the larger losses due to higher current at 1A).
While it is good to think outside the box, in this case though the box already does what you are trying to do.
I don't know what buck implementation you're talking about. Some do that already. Nickleflippers buck setup did that. The MTN-MAX buck is PWM'd by the microcontroller through its enable pin. The max buck typical application diagram makes this clear and unless RMM badly misunderstood me, or I him, he made that clear. We can head back to that thread and ask again. He seems to monitor that one.
From all the looking at docs and discussion here, actually none of it has suggested than any of the IC's discussed so far can "interpret" or internalize an external PWM signal to modify the internal switching cycle. What nickleflippers setup did was pass the PWM to a cap and resistor attached to the voltage sense, effecitvely inputting a near DC control level. This is also what I discussed in the OP.
Well I did most of my research on buck IC’s for things other then LED driving so maybe these are different, it doesn’t make a lot of sense for that to be the case though. It destroys the entire point of using a dimmable buck driver in the first place. What is dimmable about it?
Looking at HKJ’s driver tests:
http://lygte-info.dk/info/indexLedDrivers%20UK.html
Most of the buck drivers use PWM to get lower modes, but some use constant current for lower modes.
I agree with that. edit: was responding to TA.
Let's be REAL clear about these wordings. It's difficult. PWM from the microntroller is still what nickelflipper was testing, but that produce constant current from the buck in that implementation. I'd prefer to refer to "LED PWM", as in the current through the LED is pulsed (to/from zero) to achieve an average LED output. I assume this is what you are saying.
Yes, LED PWM is what I meant.
Very odd, but then the buck IC in the mtnmax is a cheap Chinese piece, so I guess it is not real surprising.
As easyb pointed out though, there are already buck drivers that don’t use PWM (why on earth would you if it is a buck driver??), the trick is finding which buck IC’s they are using and simply use those. A lot easier then trying to shoehorn it in.
I agree that constant current dimming is ideal, but using LED PWM to get lower modes doesn’t necessarily defeat the purpose of the buck driver. The buck driver allows a large input voltage range to be used efficiently.
Well are there any that actually interperet/ internalize a PWM input though? And if they do are they not just a resistor and cap in the IC? The Ti and the one NF showed have analog setpoint inputs. I'm not sure inputting PWM with a cap and resistor is so shoe-horned. It seemed a little that way until finding it in the standard application diagram. Are you sure the ones you used didn't call for a cap and resitor on the control input. Maybe you just didn't take much notice?
That's half of what it does obviously. But it seems strange to not use the other half.
Can you sent link to him as blf member? I hope he would make test if Ill sent him one driver.
Like I said, I only skimmed the thread, didn’t really catch what the proposed solution was, only that it seemed to fix an issue that is already not an issue in a lot of cases.
I am sure there are both ones that sense PWM and that use resistors and caps be it internal or external.
Really it is not that hard if that is all you are talking about. The issue is driver space. Except for something like a SRK driver space is already silly tight for a buck driver that can handle high currents. Adding anything at all to that would be interesting.
At this point, I would not be sure that the ones using constant current aren't doing some secondary linear regulation. I mean they shouldn't be, but I don't know anything until I know.