H6CC 6.7A LED driver

georges80

Flashlight Enthusiast
Joined
Oct 23, 2002
Messages
1,262
Location
Sunnyvale, CA
Apr 23. Below is the original post. Since the prototype stage, the driver has evolved into the H6CC and can output up to 6.7A.
-------------------------------------------------------

Well, until I think of a better name - high five constant current...

5A buck converter prototype. This is actually a hyperbuck driver that I realized already had all the guts to deal with 5A output - if I just changed the inductor, sense resistors and increased the switching frequency a bit from the hyperbuck configuration.

Anyhow, here's a picture that should say it all:

h5cca.jpg


Running just over 81% efficiency with 3.4V output and 14V input - which is not bad given the large voltage in / voltage out delta.

My load is 2 x P7 in parallel to deal with 5A output.

Been running for 30 minutes and totally stable. Given it's based on the proven hyperboost and hyperbuck architecture I'm pretty confident with the 5A output. Lots more tests to do to characterize efficiency over various input/output combinations.

I also need to re-lay the PCB into a round shape to make it CPF friendly :)

cheers,
george.
 
Last edited:

Th232

Flashlight Enthusiast
Joined
Dec 25, 2008
Messages
1,064
Location
Sydney, Australia
Re: H5CC driver

Toasty!
This'll make a lot of SST-50 builds a lot easier.

Any thoughts on the price difference between this and the regular hyperbuck/hyperboost?

I also need to re-lay the PCB into a round shape to make it CPF friendly :)

Lol, just last night I was trying to figure how to fit a hyperboost into a Creebar (channel is 32 mm, you can see where this is going) and wishing it was rectangular instead!
 

Codiak

Enlightened
Joined
Dec 7, 2009
Messages
515
Re: H5CC driver

Well, until I think of a better name - high five constant current...

5A buck converter prototype. This is actually a hyperbuck driver that I realized already had all the guts to deal with 5A output - if I just changed the inductor, sense resistors and increased the switching frequency a bit from the hyperbuck configuration.

Anyhow, here's a picture that should say it all:

h5cca.jpg


Running just over 81% efficiency with 3.4V output and 14V input - which is not bad given the large voltage in / voltage out delta.

My load is 2 x P7 in parallel to deal with 5A output.

Been running for 30 minutes and totally stable. Given it's based on the proven hyperboost and hyperbuck architecture I'm pretty confident with the 5A output. Lots more tests to do to characterize efficiency over various input/output combinations.

I also need to re-lay the PCB into a round shape to make it CPF friendly :)

cheers,
george.


Nice!

I'll start the chants.... production, production, production :)
 

darkzero

Flashaholic* ,
Joined
Oct 7, 2003
Messages
4,459
Location
SoCal
Re: H5CC driver

:twothumbs

Can't wait for these! Keep up the great work George!
 

Greg G

Enlightened
Joined
Oct 17, 2007
Messages
772
Re: H5CC driver

I need one of these (along with a new SST-50 emitter).
 

SUBjohan

Newly Enlightened
Joined
Dec 5, 2007
Messages
155
Location
Barchem, The Netherlands
Re: H5CC driver

Will it be possible to put 2 of these in parallel to drive 1 or more SST90's ?

Greetz Johan

Or are you secetly building a dedicated SST90 driver??
 

Hill

Enlightened
Joined
Apr 11, 2008
Messages
318
Location
MA
Re: H5CC driver

Congrats George!! Way to come through! Looking forward to getting one of these!

Hill
 

georges80

Flashlight Enthusiast
Joined
Oct 23, 2002
Messages
1,262
Location
Sunnyvale, CA
Re: H5CC driver

2 drivers in parallel is not feasible due to the topology of the driver.

Price - it'll have to wait until I've finished the new layout to see where it ends up in terms of component choices. Certainly won't be more than a hyperbuck or hyperboost.

Tests of a sample size of 1 shows it remained in regulation down to about 6.5V driving 2 parallel P7's (total Vf around 3.4V). There's about 0.4V drop/loss across the polarity protection diode. I "may" replace the protection driode with a FET - we'll see.

The spec for the switcher controller IC is 8V when configured for wide input voltage range. Its minimum operating voltage is 6V, but only when wired for Vin always less than 7V which obviously would not be the case with 2 or more li-ion cells.

It's pretty clear if you want to drive a single LED that you really need to consider a minimum series stack of 3 li-ion/li-poly/lifepo4 cells or 8 nimh cells.

Harry - there is NO way that you'll find a current regulated switching driver IC that can run from 5V at substantial output amps with a LED load. The switching element (FET) needs a high enough gate drive to fully enhance the device for efficiency purposes. At 5V you may as well use a linear shunt regulator and just burn the extra VA.

cheers,
george.
 

Mettee

Enlightened
Joined
Aug 5, 2008
Messages
678
Location
Arizona
Re: H5CC driver

wow George, great to see this...I knew you would come out with something for these new leds. Nice Work!
 

georges80

Flashlight Enthusiast
Joined
Oct 23, 2002
Messages
1,262
Location
Sunnyvale, CA
Re: H5CC driver

Ok, tested 2s2p P7's (total Vf around 6.6V or the equivalent of 2 SST-50's in series) at 5A and all looks good - driver is stable and running well thermally. Input voltage was 8.1V, below that it would go out of regulation.

This is with a schottky diode for polarity protection - so that's eating up 0.5V (4.6A input current). The H5CC driver will utilize a low Rdson FET that will drop less than 0.1V while providing reverse polarity protection to the driver.

Measured efficiency with the 2s2p P7's at 5A and 8.1V input was 88% (with the protection diode). If it was a FET the efficiency would have been around 93%.

Next test will be with 3p2p P7's to emulate 3 x SST-50 at 5A, that'll have to wait till tomorrow or Monday.

cheers,
george.
 

georges80

Flashlight Enthusiast
Joined
Oct 23, 2002
Messages
1,262
Location
Sunnyvale, CA
Re: H5CC driver

That would be sweet!!
I have a lime green 3D mag 4x 26500 Li-ion and 3x SST50 just waiting here for a driver :D

Greetz Johan

Just tested 3p2s P7's (equivalent to 3 series connected SST50's) running at 5A. Ran the test for 20 minutes (fan on the poor P7 heatsink - 50Watts going to the LEDs) and the driver was very happy on its chunk of aluminium. Hottest area of the driver was about 20C hotter than the heatsink it is mounted to (thermal pad material).

With output at 10V (total Vf of my 3p2s P7 stack) input voltage at 11.7 kept the driver in regulation - again with the polarity protection schottky diode in place.

Measured efficiency was around 93% (with the schottky in place), so it'll actually be a couple or more % higher with a FET in place.

Looking good and the next step is some cleanup of the round 1.3" diameter layout (components all on one side) and then I'll order production boards. So, by the end of March I'll have H5CC drivers available.

cheers,
george.
 

ma_sha1

Flashlight Enthusiast
Joined
Jan 10, 2009
Messages
3,042
Location
CT, USA
Re: H5CC driver

George, can you explain what you mean by "topology" here?

Thanks
 

georges80

Flashlight Enthusiast
Joined
Oct 23, 2002
Messages
1,262
Location
Sunnyvale, CA
Re: H5CC driver

George, can you explain what you mean by "topology" here?

Thanks

The way the driver is implemented as a buck driver - i.e. connection of sense resistor, power FET, inductor, etc. It can not be paralleled.

cheers,
george.
 
Top