P7 or MC-E buck driver (hipCC)

[georges80]

Hi Folks,
if some of you have been frequenting the bike subforum here on CPF you would have seen the hipFlex development that is ongoing.

I have taken the hipFlex buck driver core and made a 'dumb' driver - the hipCC. I received production PCB's today to build up a test unit and it all works nicely just as it does integrated into hipFlex.

This is a buck driver that can run from 4.5V to 24V input and drive from 1 to 5 P7 or MC-E (in 4P) LEDs. It will be set to a nominal 2.8A output but by changing sense resistors it can be adjusted up to 3A as can be seen in the pictures below.

The board is 1.1" in diameter and single sided so the bottom can be mounted to a heatsink. I will be supplying a piece of thermal pad material (nominal 1mm thick) to use to interface from the PCB to the heatsink. There are 2 mounting holes to enable screwing the board down to a heatsink and 'squeezing' the thermal pad material.

Efficiency will be similar to that posted for hipFlex and the curves are posted below.

Reverse polarity is provided on board. Total headroom at 2.8A from battery in to Vf out is 0.5V (so very low voltage drop across the driver).

A PWM input is provided and I'll be documenting that on my website as I get the driver ready for shipping within a week or so.

I have bench tested the driver for 2+ hours with 3 P7's at 2.8A and with thermal pad material to a heatsink the driver reached around 30C (~20C ambient). I have tested 5 P7's for over an hour and it was around 35C.

So, picture of the top of the board:

and of the bottom of the board. The silvery areas with lots of vias is to optimize the heat transfer through the board from the main power components.

and a picture of the driver fired up and pumping 3A through 3 series connected P7's (with some paper to 'shade' the camera).

and a measured efficiency plot of hipFlex that is also applicable to hipCC. Note that these curves are from the hipFlex proto that used a schottky for reverse polarity protection so is a hair lower than the production hipFlex and hipCC.

cheers,
george.

 

[ICUDoc]

Great work georges80- many people are waiting for a reliable driver for these new LEDs. Thanks, this looks very exciting.

 

[ktronik]

:rock::bow::bow::bow::bow::bow::bow::rock:

we love you George!!!

K

 

[Teego]

Amazing stuff. EXACTLY what I need for my new project.

 

[phantom23]

Amazing stuff. EXACTLY what I need for my new project.

Still looking for boost driver for mine...

 

[Aircraft800]

Darn, I was just going to place an order, now I'll have to ad to it
I'll have to wait till these are in stock..

 

[georges80]

Ok, ran some efficiency tests on hipCC. It is very close to the hipFlex efficiency, just a bit better due to this being a production board with a FET for reverse polarity protection and also due to one less item in the DC path versus hipFlex.

Oh, a boost for P7, not going to happen (certainly not in a small size driver), you would need >>2.8A at the input to put 2.8A at the output.

cheers,
george.

 

[phantom23]

Don't know much about electronics but 3xMC-E 2s2p (so like 6 separate emitters) + 2xLi-Ion + MaxFlex may work?

 

[georges80]

Don't know much about electronics but 3xMC-E 2s2p (so like 6 separate emitters) + 2xLi-Ion + MaxFlex may work?

Depends on current. That's a LOT of heat and inefficiency due to the large delta from Vin to Vout.

6 series LEDs at say 3.5V each is 21V and only being fed with nominal 7.4V is not a good recipe for efficient boost. Lots of heat will be the byproduct and you'll have a challenge cooling maxflex or any compact boost driver faced with that scenario.

There is a reason I developed a 2.8A BUCK driver...

cheers,
george.

 

[phantom23]

3x4p - three leds will be better but MaxFlex delivers up to 1,2A only.

Buck means I need 4 Li-Ions to three emitters which means huuuge flashlight.

 

[Aircraft800]

Can this be used in conjunction with the D2Flex for multilevel?

 

[georges80]

Can this be used in conjunction with the D2Flex for multilevel?

Yes, I will develop a firmware tweak to d2Flex to allow it to control hipCC, but if you have room the hipFlex has it all in one driver....

I have added a PWM input to hipCC and the V1.1 d2Flex has a PWM output - so the hardware is in place, just need to tweak the firmware.

I figured that the hipCC at 1.1" diameter with a d2Flex may fit in places where the 1.4" diameter hipFlex may not.

cheers,
george.

 

[Aircraft800]

Yes, I will develop a firmware tweak to d2Flex to allow it to control hipCC, but if you have room the hipFlex has it all in one driver....

I have added a PWM input to hipCC and the V1.1 d2Flex has a PWM output - so the hardware is in place, just need to tweak the firmware.

I figured that the hipCC at 1.1" diameter with a d2Flex may fit in places where the 1.4" diameter hipFlex may not.

cheers,
george.

All would be nice if it fit in a standard "C" Mag tube, looks like I have a problem

 

[TexLite]

Hi Folks,
if some of you have been frequenting the bike subforum here on CPF you would have seen the hipFlex development that is ongoing.

Looks great!

And yes,I really should check that forum more often.I probably would have missed this if you hadn't posted here.

On an off topic question,how do you like the Extech?Looks like a very useful tool.

-Michael

 

[georges80]

Looks great!

And yes,I really should check that forum more often.I probably would have missed this if you hadn't posted here.

On an off topic question,how do you like the Extech?Looks like a very useful tool.

-Michael

The Extech is a pretty useful tool. I use it only for the current clamp aspect since I already have a decent multimeter. Extech also has another model that has 1 mA resolution, the one I have is 10mA. So, in retrospect I would have gone the 1mA model, but it wasn't on the shelf at Fry's a year or so ago, so I bought the 10mA model.

Regardless of model/brand, make sure that the clamp meter can read DC current. A lot of the models (especially the cheaper ones) only do AC current.

Also, at those low current levels, you want to put the clamp meter over the wire (with power off), then turn the meter on, let it settle for a few seconds, then push the 'zero' button to remove any offsets etc. Don't move the meter once it has been zeroed and you'll get pretty accurate readings.

I also have a fancy smancy current probe that I hook to my scope when I'm looking at current transients and pulses etc - but that's in a totally different $tratum.

cheers,
george.

 

[TexLite]

The Extech is a pretty useful tool. I use it only for the current clamp aspect since I already have a decent multimeter. Extech also has another model that has 1 mA resolution, the one I have is 10mA. So, in retrospect I would have gone the 1mA model, but it wasn't on the shelf at Fry's a year or so ago, so I bought the 10mA model.

Regardless of model/brand, make sure that the clamp meter can read DC current. A lot of the models (especially the cheaper ones) only do AC current.

Also, at those low current levels, you want to put the clamp meter over the wire (with power off), then turn the meter on, let it settle for a few seconds, then push the 'zero' button to remove any offsets etc. Don't move the meter once it has been zeroed and you'll get pretty accurate readings.

I also have a fancy smancy current probe that I hook to my scope when I'm looking at current transients and pulses etc - but that's in a totally different $tratum.

cheers,
george.

Wow,thanks for taking the time to share those tips,I appreciate it.

One of these is now tops on my want list.

Thanks,
Michael

 

[Fluor]

Nice driver!

When can i place a order and what will the price tag be?

 

[9x23]

George,

I am excited to see your new drivers for the P7 and MC-E coming soon. Thanks for offering these. Paypal is standing by!

9x23

 

[georges80]

I'm hoping to have them available within 2 weeks. I'm just waiting for a supplier to provide one of the IC's that goes onto the board, the rest of the components are on hand.

Pricing, I'm still working out the details. Given hipCC is a 'dumb' driver it will be priced in line with my other fixed output drivers.

cheers,
george.

 

[wquiles]

Awesome George!!!

Not only the availability of the driver itself, but also the fact that this is a high efficiency driver :twothumbs

Will