H6Flex 6.6Amp Adjustable Buck

[PCC]

Any updates as to when these will be available?

[georges80]

Soon, soon

I sent out for another protopcb run last week, should have boards in hand in the next few days (likely Tuesday). I already have the parts on hand to assemble the prototype.

So, by the end of this week I'll have tested the new layout and uController and at that point will kick off a production PCB run that will take about 2 weeks to deliver.

I'll keep this thread updated as I run the v2 prototype through its paces.

If all goes to plan (and it should, given the testing so far), the h6flex should be available by the end of this month (June).

cheers,
george.

[georges80]

Ok, prototype boards arrived yesterday. Got one assembled and had a bit of time to port the code to the new uController I'm using for this board.

Little bit of debugging to fix a few porting errors and I've got the board running on the lowest current table (1.4A). Low runs around 20mA. More testing to do to verify the standard flex features (voltage monitoring/warning, temperature monitoring etc).

First picture shows the new prototype and two jigs (top left is the test jig, top right is programming jig).



Closeup of the lastest prototype sporting the new uController. The main reason for going with the new uC (still an Atmel AVR device) was that I needed one more i/o pin and this unit provides lots of i/o's and in a smaller physical package. The extra i/o was needed to provide PWM control for reliably achieving the lowest current levels (<200mA) with decent tint control. On my previous drivers I could use a single pin for power management and PWM control, but the switcher IC I'm using on the h6flex (and h6cc, hyperboost and hyperbuck) has a dedicated PWM pin that provides some impressive control. So, now I needed one more i/o pin - so that's the story...

Bit more of a challenge to solder this tiny package, but fortunately a hot air tool and quality solder paste makes the job reasonably easy even without the soldermask of a production PCB.



The bottom of the PCB has the thermal interface (no components) as per my other high power drivers.

cheers,
george.

[dan1million]

Wow H6flex is really starting to look good !

Will this have all the features of the Hipflex ?

-Dan

[wquiles]

Looks awesome George - loved that close-up picture

[georges80]

Wow H6flex is really starting to look good !

Will this have all the features of the Hipflex ?

-Dan

Yes, it wouldn't be called a h6FLEX if it didn't have all the standard FLEX features of my drivers. i.e. it will have the exact same UI functionality, menu structure etc etc.

That imho is the one of the beauties of my flex drivers, doesn't matter whether it's a bflex,d2flex,hipflex or h6flex, the end user sees the exact same interface, nothing new to learn.

cheers,
george.

[georges80]

Have done more testing on all current tables and all is looking good on the proto. Tested the voltage monitoring hardware and that seems to be running within 1-2% setting accuracy which is in spec.

I've taken some video that shows the driver in action on the 6.6A current table. I'm controlling the driver via a small microswitch soldered to some flying leads to the h6flex proto.

First part shows 2 parallel P7's with the h6flex set to UIP (5 levels). I'm switching through various levels from 20mA up through the levels to high (6.6A on the 6.6A table). You can see the output current on the meter (in series with the LED output).

The second part shows 3 series sets of 2xP7 (emulating a total Vf of around 11V) and doing similar tests to the first part. That's 11V x 6.6A = 72.6W running to the LEDs

I think my camcorder was screaming for some sunglasses during the test

http://www.youtube.com/watch?v=7xssW3AI1qk

cheers,
george.

[wquiles]

Very cool video

[georges80]

I have 7 current tables implemented on the h6flex.

These are the output currents measured on the proto for the 7 tables for UIP and 5 level UIB:

L1_____L2_____L3____L4_____ L5
22mA 220mA 362mA 732mA 1430mA
22mA 280mA 534mA 1018mA 2030mA
22mA 364mA 734mA 1440mA 2825mA
22mA 477mA 932mA 1835mA 3625mA
22mA 621mA 1120mA 2364mA 5040mA
22mA 734mA 1418mA 2350mA 5654mA
22mA 818mA 1618mA 3374mA 6580mA

I've set L1 to a very low level (1 in 10 duty cycle of 200mA nominal so 20mA average). All levels above 200mA are fully current regulated (no PWM).

My thinking is a rough halving of current per drop in level, with L1 being the ultra low for extended runtime. 22mA is still quite a lot of light from a high power LED. Halving current at each level from L5 to L2 gives reasonably nice linear intensity steps, but of course that means quite a huge jump from L2 to L1. I could make the L2/L3 current steps have a larger delta to more 'even out' the jump from L2 to L1 - I'm open to suggestions.

The one thing I won't change though is UI scheme, so no programmable current values etc - the steps will be fixed based on a consensus and of course my own thoughts on the step sizes.

cheers,
george.

[aurum]

The one thing I won't change though is UI scheme, so no programmable current values

[netprince]

Hey George,

About how tall is the driver? I'm trying to work out the spacing with some heatsink ideas... (sorry if I missed it)

Thanks!

[georges80]

Hey George,

About how tall is the driver? I'm trying to work out the spacing with some heatsink ideas... (sorry if I missed it)

Thanks!

Same as the hipflex (basically the PCB thickness + inductor height) - info on that driver on my website.

I sent off for production boards yesterday, so should be available around the end of June.

cheers,
george.

[dan1million]

nice work !!!! be interested to see this one for sure !

[greencardigan]

July is here. Any production boards yet?

[SUBjohan]

Hi George,

Any updates??

Because I want 4 I want 4 I want 4 I want 4 I want 4 I want 4 I want 4 I want 4 I want 4 I want 4....

Greetz Johan

[georges80]

Yes, I have production boards on hand. Just haven't had time to run all the verification tests on the firmware etc - YET...

Just returned from a camping trip with my sons, so still digging myself out of a backlog of stuff...

I'll post pics of the production boards - maybe later today.

cheers,
george.

[georges80]

So, finally, here's a picture of a h6flex production board. I have to verify the firmware that was running on the prototype before I start shipping drivers, but that should happen within the next few days.

Top view of the driver, components only on the top side to enable the bottom to be mounted with thermal tape or pad material to a heatsink.



cheers,
george.

[Mettee]

yes!

[wquiles]

Very nice looking board - nice work George

[dan1million]

that is a work of art george
cant wait for the release of this one into the wild.

[Al Combs]

It didn't occur to me until I just looked at your h6flex production board (a real beauty BTW), but how do you mount it with no mounting holes? I was under the impression the TGF120K thermal pad was just thermally conductive rubber. Does it have adhesive properties as well?

[georges80]

It didn't occur to me until I just looked at your h6flex production board (a real beauty BTW), but how do you mount it with no mounting holes? I was under the impression the TGF120K thermal pad was just thermally conductive rubber. Does it have adhesive properties as well?

That's why I mentioned thermal tape....

cheers,
george.

[SUBjohan]

Any updates?

Greetz Johan

[georges80]

Yep, it's working well. Just back from a short business trip so over the next couple of days I'll finalize the last of the testing and open up for sales by mid week - but heading out on business again towards the end of the week - so realistically it will be one more week before it's really available for general purchase.

When it appears in the order page of my website, it's ready to order

cheers,
george.

[greencardigan]

George, can we in Australia order these through your website? Or do we have to wait until Cutter stock them?

[Lumens Industry]

Excellent work George!! Cant wait to get my hands on some

[georges80]

George, can we in Australia order these through your website? Or do we have to wait until Cutter stock them?

For the initial while I'll ship direct since I need to get them out to the field and handle any teething issues directly (like the new manual that I'll have to put together).

After that it'll depend on whether Cutter stocks them or not.

cheers,
george.

[dan1million]

Keep us posted on price and availability please

we are all looking forward to this one !!!!!!!!!!!!!!!!!!!

[georges80]

Ok, I have a batch programmed & tested and ready to ship.

I'll put tech info up later today onto my website, but basically all the features of the standard Flex drivers with various current tables up to 6.6A.

Pricing will be up on my website order info page in the next little while.

cheers,
george.

[SUBjohan]

Pricing will be up on my website order info page in the next little while.

You have got my order in your mail box
Thank you for making this great driver possible!

Greetz Johan