The Texas Buck driver series, Q8 / Skyray King 2S/4S buck driver RELEASED!

467 posts / 0 new
Last post
Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

 

The input capacitor however is showing up as a very the large majority of the remaining 5% innefficiency in for near 1:1 operation.  I'll double check there, but this can be useful info.  

 

Edited commentary.  I first increased the capacitance and it got better, but then I realized that's because I calculate ESR from the capacitance, frequency and dissipation factor, so maybe that's not right.  Then I realized many of the bigger 0805 caps are just stacks of 2 or three smaller ones.  And it seems to me two parallel caps have the same phase angle as one, so the original conclusion is right. More capacitance, less ESR.  Also checking some spec sheets dissipation factor doesn't seem to vary so much across capacitance values.

 Anyway, I'll see if I can find some cap or combined stack that works a little better.  5% isn't a big deal, but it's probably easy to make it 1.5% instead, which would just be a nice boast.

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 20 hours 58 min ago
Joined: 03/24/2016 - 07:44
Posts: 8492
Location: Everything is brighter in Texas

That is very interesting indeed.

If we kept to a 17mm inductor then we should be able to fit that without much trouble, including the mosfet. With a 22mm it would be real tight but just might fit, although not sure if it would compromise the overall EMF interference due to cramming things closer together.

Although looking at existing buck driver designs, they cram components right next to each other with no apparent side effects, although they also put out a fraction of the current we plan to.

We could try to use the LTC4412 if nothing better presents itself in the Q8 driver setup. For the smaller setups I doubt we will need more then 10 amps of current ability so it should not be as big of a deal anyways.

I kinda figured that things would get harder as we decreased the output voltage. Circuits pretty much always handle voltage better then amps, Ideally it would be something like a 4S input to a 3S output, just enough to keep it in regulation till the batteries are depleted. but thats just not practical in the real world. That is unless you swapped the Q8 out for 4x triples and built a DIY M43 (or M34?).

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Well here are a ton of numbers:

Caps are 10 uF 7.5% dissipation factor.
Inductor is the 15uf one previously (lower resistance than the 22)
Diode is 0.5V Vf Mosfet numbers are a bit vague, but I think at least Rdson is about right, wouldn’t trust the switching losses much at this point though. Input voltage is 16.8, output is in multiples of 3.5 for now (I may add a correction for current, especially since it matters in 1:1). The output power at 4:1 is less because I limited it to 15A for now.

I just noticed, the 5% mode on this light is going to brighter than most pocket lights (not BLF ones). I’m pretty sure we’re going to need that PWM for moonlight and even just an indoor low.

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

So for a true synchronous buck the freewheel mosfet is opened late and closed early to prevent reverse current, using the body diode during the transition.  But I realized for this thing that works during opening, but not closing. It's already open and it closes too late.. and I guess there should be reverse current briefly then. I don't know how ideally that thing really works.  A very smart IC could predict closing based on the last cycle and only have issues then during load changes.

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 20 hours 58 min ago
Joined: 03/24/2016 - 07:44
Posts: 8492
Location: Everything is brighter in Texas

Good data, the specs look good. Ripple is actually better then I expected. I was thinking we would have to settle for 2%+ due to space constants. Less then 2% will be a big step up from what is currently available at 10% ripple.

What is the input cap ripple exactly?

3W of loss through the diode while obviously not good could be dealt with I think. If nothing else you could stack silicon cubes on it to contact the upper shelf. The synchronous IC would be a great option if it works but we might want to stick with a simple setup to start out with and make sure it works as we desire first.

Overall an efficiency of ~90% is what I was hoping for, supposedly it would beat that for 1:1 and match that for 2:1, which is all we really care about. 4:1 is simply for outlying cases that are good to consider in the early stages.

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

It took me awhile to come to terms with this input cap thing.  It seems to assume that 100% of the AC/ripple current component is provided by an input capacitor not the battery.  The size of the cap then determines the translation of that current ripple into a voltage ripple and, as there's ripple current through the cap, of course there are losses there too.  Supposedly this cap should be connected as close to the high side mosfet as possible, and I guess should also be grounded pretty well to avoid inductive pickup on the ground side of the cap. 

 

Yeah, I agree the synchronous stuff is a bit elaborate.  However it takes time to get boards.  If we really think this diode replacement circuit could work, could layout pads for both in a prototype board.  Might need to get some breadboard and scopes involved for that kind of development though.  I do have access to good scopes but not sure I would find time. 

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 20 hours 58 min ago
Joined: 03/24/2016 - 07:44
Posts: 8492
Location: Everything is brighter in Texas

It only needs a 10uf cap to handle that kind of power? Seems kinda small?

Far as the synchronous side of things, I am not sure it is worth it, I think we would come out ahead with a larger inductor over that personally and smaller drivers sure won’t have room for it. So it kinda leaves us at square one and doing all that work for a single driver / setup. It can be done, just not sure it is worth the time and effort.

So now that we know the specs we just need to settle on exact components and finalize a design so we can try it out to see how it works.

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Just discussion for fun.  I think we'll have a traditional diode this round and 4% diode loss in 2:1 is ok with me, but the larger inductor doesn't actually help efficiency until ripple current saturates, so below 3.5W as 200% ripple means you hit zero current.   Just means you use PWM from a little higher mark.  

 

As for the cap, it's only handling the ripple current, and at very high frequency, so low charge  (I*t) per cycle.  This isn't 60hz AC.  I'll check the numbers again by hand though.  I did just kind of bash in the equation from that article for that one without much thought, other than following the idea about at the detail I just stated.  Still I think there's a tendancy to think like the output cap is supporting power during off time, but the inductor is doing the bulk of that work, at least under optimal conditions.

 

 

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Oh, one problem with diode loss, as they hot they start behaving more like resistors.  That means Vf drops, that's good, but reverse leakage current also increases.  That's bad. I have not even tried to estimate how much this matters if at all, but the point is yes, heat sinking the diodes well is probably needed.

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 20 hours 58 min ago
Joined: 03/24/2016 - 07:44
Posts: 8492
Location: Everything is brighter in Texas

Yeah, I figured they would need a large thermal pad, although since it is positive and not ground it is hard to actually transfer that heat directly to the host via copper pour.

Some silicone thermal cubes are quick and easy way of doing it but also a bit more involved for most and I would prefer that not be a requirement, just optional.

A small heat sink could also be glued onto it but with no contact to anything I doubt it would make a large difference.

A synchronous buck IC would be great but I sure have not found anything suitable, have you?

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

So I did check the cap ripple equation and the bottom line is those numbers were actually too high. Basically what I scribbled down looks like equations 24 and 25 on page 6 here:

http://www.ti.com/lit/an/slva630a/slva630a.pdf  (fixed the link)

For output ripple, really only the capacitive part ends up mattering, so what phase angle to add the resistive part in (which creates the other detail in that paper really)  is irrelevant.  The cap term is just Q/C where Q is average I*t   and the rising time is haf of 1/f and the average current is half of the max current and the max current is half of the peak to peak, so with three halves you get v_p2p = 1/8 * I_p2p/(f*C) as per above link.

But the powerlectronics equation doesn’t really agree with that for a couple of reasons, which seem to me to just be wrong, and wrong usually actually too high.    I’ll use equation 25 from Ti now which you can see still errs on the high side compared to their simulation.

 

Components:

Ok I looked at caps a little again.  I’m sticking to my 1210 recommendation.  There are no 10uF caps on digikey in x7r with ratings 35V or higher in anything smaller (and the 50V ones have the lowest ESR).  Furthermore, I looked at larger caps and they get REALLY expensive in those kinds of specs:

http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail&itemSeq=209394405

That’s a cheap one.  The one’s in between cost twice as much.

 

and it’s not clear you can do any better.  For output 10uF is enough.  For input ESR is really the problem anyway and above 10uF the caps are pretty much all measured only at low frequency.  Even for 10uF where they’re measured at 1khz, I’m not sure I trust extrapolation to 1Mhz well, but at least I can find some cheap ($.30)

http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail&itemSeq=209396138

 and some individually specced (rare, reading through these cap sheets is a pain) up to high frequency, even if not amazingly low ESR (average).

http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail&itemSeq=209397082

I would recommend room for two or three in parallel on input though. 

So I’m done with caps.  If you can find something with more capacitance, less ESR, rated better for high frequency or in a smaller package, have at it, but this is the best I got.  I am probably about done with diodes too once I post what I have, so a power Mosfet will wrap up the main bits.

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Texas_Ace wrote:
Yeah, I figured they would need a large thermal pad, although since it is positive and not ground it is hard to actually transfer that heat directly to the host via copper pour. Some silicone thermal cubes are quick and easy way of doing it but also a bit more involved for most and I would prefer that not be a requirement, just optional. A small heat sink could also be glued onto it but with no contact to anything I doubt it would make a large difference. A synchronous buck IC would be great but I sure have not found anything suitable, have you?

 

Wait, the diode doesn't connect to ground? (or maybe just the casing doesn't?). I better review the circuit.  I've seen one Ti example where everything is upside down.   Suitable syncrhonous?  Not really.  I found some IC with like 50 pins and a billion features that yes would work in principle and could handle boost circuits too, but...  And it turns out thse "Or'ing" IC's are pretty slow and yes do leak during the 1/2 microsecond when they're closing.

 

I'm tracking components in post 96: http://budgetlightforum.com/comment/1013662#comment-1013662

But your links here don't work: http://budgetlightforum.com/comment/1012189#comment-1012189 post 69.

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 20 hours 58 min ago
Joined: 03/24/2016 - 07:44
Posts: 8492
Location: Everything is brighter in Texas

Odd, my links work fine for me but none of your digikey links for me.

We can put a cap in the input, just goes against my principle of large traces for large current as it all has to pass through the small cap. Although paralleling them is an option as well. I doubt I could find anything better, the issue wasn’t the exact cap to use, it was wether it was truly needed and would make a noticeable difference in the final product.

I will look at the parts list and update the PCB when I get a chance.

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

OK, I'll try to understand the digikey thing (probably will have to use two browsers to figure it out, sounds like the links might be linked to cookies or something).  I better add model numbers.

 

The input current doesn't pass through the input cap.  It's in parallel  to the battery contact, and the traces for the cap only need to handle the ripple current, not full DC current.  I think it will help protect from all the stray field stuff you're worried about, among other things.

 

As for what's really needed, you thought 10uF was small, so I don't know what to say.  x7r seems required, and I think 2X actual voltage is standard minimum and those specs lead you to 1210 or at least 1206.  There's just nothing else.  Plus that samsung one is really cheap.

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Wow,  I see, yes, paste the link in a different browser, and they don't work!

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

I updated the digikey links in post 96.  To get a good link you have to find the share button on the digikey product page. For my own links.. sometimes I have to re-enter the product number in the search to get the share button back.

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 20 hours 58 min ago
Joined: 03/24/2016 - 07:44
Posts: 8492
Location: Everything is brighter in Texas
Flintrock wrote:

OK, I’ll try to understand the digikey thing (probably will have to use two browsers to figure it out, sounds like the links might be linked to cookies or something).  I better add model numbers.


 


The input current doesn’t pass through the input cap.  It’s in parallel  to the battery contact, and the traces for the cap only need to handle the ripple current, not full DC current.  I think it will help protect from all the stray field stuff you’re worried about, among other things.


 


As for what’s really needed, you thought 10uF was small, so I don’t know what to say.  x7r seems required, and I think 2X actual voltage is standard minimum and those specs lead you to 1210 or at least 1206.  There’s just nothing else.  Plus that samsung one is really cheap.

Ah, ok now things clicked (been distracted recently and never really took the time to think this aspect through). For some reason I was picturing the input cap inline with the mosfet instead of parallel. Now everything makes perfect sense and 10uf should indeed work fine.

Thumbs Up
Texas_Ace
Texas_Ace's picture
Offline
Last seen: 20 hours 58 min ago
Joined: 03/24/2016 - 07:44
Posts: 8492
Location: Everything is brighter in Texas
Flintrock wrote:

I updated the digikey links in post 96.  To get a good link you have to find the share button on the digikey product page. For my own links.. sometimes I have to re-enter the product number in the search to get the share button back.

Interesting, I will keep this in mind.

I will be placing an order with arrow in the next few days due to the free shipping they have going right now, so I might see if they carry these items and grab a set to have on hand.

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Texas_Ace wrote:
Flintrock wrote:

Ah, ok now things clicked (been distracted recently and never really took the time to think this aspect through). For some reason I was picturing the input cap inline with the mosfet instead of parallel. Now everything makes perfect sense and 10uf should indeed work fine. :THUMBS-UP:

 

It will work.  Fine, well, you can get about a 3% efficiency boost for using 3 instead of 1 maybe more if real ESR at 1MHz is worse than predicted, or less if it's better.  That's why it wouldn't hurt to add places for a few if there's no downside.  $.30 for a percent more efficent?  Enh, you now, I might do it, just because.

leroycp
Offline
Last seen: 1 hour 1 min ago
Joined: 12/27/2015 - 19:24
Posts: 731
Location: Stewart TN

LOL………..just do it!

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

leroycp wrote:
LOL...........just do it!
 

 

Not if you do it first.

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 20 hours 58 min ago
Joined: 03/24/2016 - 07:44
Posts: 8492
Location: Everything is brighter in Texas
Flintrock wrote:

Texas_Ace wrote:
Flintrock wrote:

Ah, ok now things clicked (been distracted recently and never really took the time to think this aspect through). For some reason I was picturing the input cap inline with the mosfet instead of parallel. Now everything makes perfect sense and 10uf should indeed work fine. Thumbs Up

 


It will work.  Fine, well, you can get about a 3% efficiency boost for using 3 instead of 1 maybe more if real ESR at 1MHz is worse than predicted, or less if it’s better.  That’s why it wouldn’t hurt to add places for a few if there’s no downside.  $.30 for a percent more efficent?  Enh, you now, I might do it, just because.

Yeah, in parallel is easy and there should be plenty of room, could even go with a larger form factor if there was a better option vs paralleling a few 10uf.

I will rearrange everything with the latest component list and see what the space looks like in the next few days.

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Send a working link to that mosfet and diode you found if you get a chance.

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 20 hours 58 min ago
Joined: 03/24/2016 - 07:44
Posts: 8492
Location: Everything is brighter in Texas

remember these were just the first things I found and I didn’t peruse them much at all.

The P-channel version of the SIR800: http://www.digikey.com/short/39nh5h

This diode was literally the first thing I clicked on that had specs that should work, simply to see the form factor: http://www.digikey.com/short/39nh50

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Great. 

I still want to look over diode choice more soon. There might be things in more standard packaging without losing much.

 

Haven't searched for MOSFETs but on first glance that one looks pretty good.  Still need to run reasonable numbers on switching losses when I get a minute for it. 

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Ok, I updated post 96 with a bunch of diode research.  Basically there are good options in POWERDI5060-8, and in TO-263 (D2PAK).   These can probably fit on the same pads if done right.  Tiny heat sinks exist for the TO-263 which also need footprint and space consideration if used.  

 

TO-220 through hole also has many options, in 2 or 3 pin.  Those would certainly require some kind of heat sink, but are also easy to attach to one. If the surface mounts can handle the heat, it's probably better to use those instead.  I see your point about the thermal pad being on the wrong side of the diode.  

 

I think your SO-8 mosfet is almost certainly a fine footprint. So I think we're about all set for major components.

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 20 hours 58 min ago
Joined: 03/24/2016 - 07:44
Posts: 8492
Location: Everything is brighter in Texas

Great, I am real busy over the next few days but I will get around to updating the PCB before too long.

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Now, about your circuit, I don't know about how C1in and R1Sense are arranged in the top right.

 

Edited:  Never mind... I see Ti is actually intentionally reading ripple current, so it makes sense as is.

 

I'll start reading the manual more now.  The constant off time and peak current control are interesting.  It doesn't seem they change any of the fundamentals, just changes how frequency and current get controlled, so for instance in my table a particular frequency at one voltage output should probably imply a different particular frequency at another.  I'll probably try to get familiar with it and re-arrange my tables instead of by frequency as by offtime RC constant.  That will give a little better picture, but shouldn't impact parts selection I think.

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Learning all the dirty laundry about this IC.

 

So if we use full input range for current adjust, 1.24V max on IADJ corresponds to 248mV of sense voltage.  At 15A, VI is nearly 4 watts across the sense resistor.   We could use half the range for control (lose resolution a little).  Or add an amplifier (seems a bit drastic to save 5% power maybe).

 

However there is also this funny business where the peak current to turn off the switch jumps up and down because they keep flipping the voltage comparator every two cycles.  See figure 24 page 15 and the discussion.  To me this seems  nuts.  They claim it makes a more accurate average current, but accurate compared to what?   It  seems keeping the fixed target would be just as consistent but with less ripple and without all the annoying caveats that come after it.  Actual inductor ripple will be higher than what I calculate because of this effect.

 

Enough grumbling, the point is they claim you should maintain a minimum ripple current (see equation 11), bigger than the variation in their sensing setpoint (that they caused), which they esitmate as maybe 24mV.  Using full scale 248mV sense range that's already 10% at full current, so going to a much smaller sense resistor and smaller voltages is maybe not a great idea. Note this is not output ripple voltage though. Mostly we care about this because if we try to keep ripple high at full current, it's even higher at low current, and goes into discontinuous mode earlier.  We'd rather not have high ripple current.

 

You have to read details there to understand it.  Actually though I think it's ok to have ripple lower than what they say, at high power and I'm not sure I'd worry about the degraded current control they say it causes.  Peak current still won't ever go above the max cuttoff(s) and the net effect seems to be you'll just get an effectively double off-time followed by twice as long of an on time, in other words, higher ripple current anyway (kind of by way of effectively lower frequency), but no need to target higher ripple and damage the low current performance at the same time in my opinion.  Basically I would ignore the constraint in equation 11.  It could make ramping uneven as you cross the threshold for double or tripple off-time, creating jumps in ripple current, since it's peak current that's controlled, not average and higher ripple for same peak lowers the average.  But any given mode should still stay pretty steady. 

 

I suspect when all is said and done, there's about 5 more places with 1W losses that are just too detailed to find, and all this brings 7V output down to 80% efficient at high power.  It's a good reason to fight a percent or two here and there where possible though, like the input caps.

 

 

 

Flintrock
Offline
Last seen: 1 year 8 months ago
Joined: 09/10/2016 - 20:29
Posts: 1544

for footprint, 5W, 7 to 15mohm resistors are available in D2Pak's (like the diodes I listed) or in sizes from 2818 up to 4527.  Bigger pads, lose more heat anyway, so. Could also make it through hole and try to sink it to the case.  The D2PAK's are about $2.00 and the standard flat resistors are about $1.35 for the cheapest.  This driver won't be cheap. The D2Pak's are rated at very high power, like 35W so might stay cooler.

 

In a 4 by xp-l setup this is only heading toward about 1.5W at 6A per led if tuned for 248mV sense at max current give or take (1S or 4S doesn't matter, so long as you tuned the resistor that way for either, duty factor cancels the current change).  It's just driving 4 x 12V xph35's that pushes it close to 4W.

Pages