Buck and Boost Drivers, Testing, Modding, and Discussion (Pic Heavy)

Wellp, as I said before I can't assess the optimization needs for that driver to operate nicely with slightly larger input to output voltage difference, although I can read and pinpoint some interesting tutorial/article like How to Choose the Right Inductor for DC-DC Buck Applications @ Passive Components Blog.

I bought some Murata FDA1254-H-4R7M (half an inch square, 4.7 µH, rated current Isat/Itemp: 11.2/10.2 A) inductors for a crazy mod project more than a year ago, do you think these would do the job any better?

No, I did not expect You to tell the Highest Wisdom. :innocent:
Just talking about modding.

The working theoretics are clear for me, but the calculations are not easy since the major vast of required data is not available.
Though, it is clear that increasing the voltage difference between Vin and Vout reguires higher inductance value for the same ripple.

In the meantime I got remembered that I have some spare inductors in my loot trunk. :sunglasses:
And I’ve found some 4.7uH inductors and also some 10uHs. different sizes, different qualities. Some of them are Coilcrafts.
I removed the original inductor and will test the other ones. :smiley:

Thanks for the tips!

Looking more closely the stock inductor looks about ½" × ½" (more or less like the Murata I mentioned above) so, if its inductance is 4.7 µH, its current specs should be high enough. But of course, very possible it is that whoever is making these drivers may not be purchasing top spec inductors. Just a little bit more inductance should be fine.

You might try the xal7070 series coilcraft inductors. I have used a few and never heard a whine.

Has a low DCR so you should gain a little current without the whine.
The New XGL series looks interesting.

Actually, it’s a 1040 size inductor. On my driver it’s not covered by glue and the 4u7 mark is clearly readable.
The size is not a big constraint, bigger inductors can be easily installed.

I’ve found in 4.7uH some MSS1048 Coilcrafts, some SER1360s and a noname 1080 one. Also have some 10uH SER1360.
I will test whether the inductace value, or the current handling capability is the bottleneck factor.
Thanks for the tips!

Wellp Lacerta, according to what you are saying the stock 10 × 10 × 4 mm inductor in the Manta Ray driver is likely doing its best already, or very close. I mean, were they had chosen a lower specified inductor the driver would have malfunctioned.

The way metal alloy inductors work, their inductance diminishes as they approach their current limits. This basically means a proper quality, higher current rating bigger inductor quite likely will work right even if the Vin to Vout delta is somewhat bigger.

Now that I think of it, the FDA1254s I have lying around will find home in some of these drivers sometime, hope I can raise the current output a nice bit with them. ;-)

Considering the 1040 data sheets ( on Aliexpress ), 4.7uH coil saturates at 12A, and can handle 9A RMS. Now we have 4.8A RMS. We should have.
Of course, maybe these should be interpreted as chinese ampers. :smiling_imp:
I will try other inductances and will feedback about the results.

The FDA1254-H-4R7M (12.6 × 12.5 × 5.4 mm including terminals, 4.7 µH, rated current Isat/Itemp: 11.2/10.2 A) I spoke about earlier is noticeably bigger and does not reach your above figures. So something's wrong, and I guess Murata is not.

Yes, in same size Coilcraft inductors are not such good either. :slight_smile:
Should be chinese ampers, as they invented chinese lumens too. :smiley:

No success. Mission Failed. :frowning:

I’ve tried several inductors, all of them had higher current handling capacity than the original 1040/4.7uH.
Some had same inductance, some had higher ( up to 10uH ).
Still no stabilzaton on high and turbo, and all the coils were saturating.

It means that the saturation is not the rootcause, only the consequence. Consequence of the converter runs with false current sensing or regulation.
I don’t know the reason.

I should look for another solution ( driver ).

Hey guys;
Since we are here in a boost driver modding discussion…I’d like to get some help in modding this xhp35 2.3A driver from convoy;

https://www.aliexpress.com/item/4000046041955.html?spm=a2g0o.productlist.0.0.1eaf4bc2QymTBc&algo_pvid=98a46962-2d49-49c0-91c8-36bb56be05f7&algo_expid=98a46962-2d49-49c0-91c8-36bb56be05f7-0&btsid=0ab6f83115874007092831047eed4b&ws_ab_test=searchweb0_0,searchweb201602*,searchweb201603*

Can anyone tell from the pictures of the board, or by experience if it is possible to make it Single , turbo 100% Mode only? The led is just perfect for my project
except for that reason
thanks alot

Hey Fabio, could I ask, do you want to use it with an XHP35? And with single battery supply?

yep exactly, do you think it is possible??

It's certainly possible; it's pretty simple. Just connect a piece of wire like this:

I posted more detailed information on this Convoy driver here: https://budgetlightforum.com/t/-/61809

It uses the MP3431 chip, and works really well.

I include schematic, board layout, and testing results. I invite anyone in this thread to check it out if you’re interested.

Hey man, thanks alot for the answer;
Since Im a total newbie, I have here a magnified picture just to make sure I got it right, please confirm!
Very impressive and deep analysis of the driver in the link…congratulations;
Well I wasnt expecting so much info …but since you are here,please tell me… is it possible to boost the amps keeping 12V, instead of 6v like you modded in the provided link??
I mean, possible for a normal person with poor smd soldering skills.
thanks



Imgur

Yep, that’s right. That’s the simplest thing to do to make it fixed 100% output.

As for increasing the current at 12V, it is definitely possible with this driver, but I’d be cautious about why you need to.

If you want to drive a 12V XHP70, then yes, I could give more advice, but for an XHP35, there is no need to increase output current. Any more will kill that LED.

Finally, I solved the case.
Not the most elegant solution, but it works.
I skipped the 3V XHP50.2 and returned to a 6V version, and replaced the 10mOhm current sensing resistor to a 20mOhm one.
The output current is halved now in each mode and stable in the entire battery discharge range. No more inductor whining and current rush.
However, since 6V and not 3V LED is used, the LED Vf is very close now to the battery voltage in high and turbo mode when the batteries are almost discharged, so the current path in the whole flashlight should provide the lowest possible voltage drop.

Has anyone managed to mod any single cell boost driver (for 6v leds) with an e-switch?
The only commercial model I found was this one on taobao.
But it’s UI is pretty bad, and it is too big for my build.