LED drivers and Accessories you want, but don’t exist

488 posts / 0 new
Last post
Schoki
Schoki's picture
Offline
Last seen: 5 days 16 hours ago
Joined: 03/27/2017 - 15:34
Posts: 215
Location: Germany

I now made a PCB for a 1S 2A max buck-boost converter, but somebody needs to write some code for it. It works pretty similar to loneocean’s gxb20: Oshpark Link

Lexel
Lexel's picture
Offline
Last seen: 11 hours 40 min ago
Joined: 11/01/2016 - 08:00
Posts: 5357
Location: Germany
Schoki wrote:
I now made a PCB for a 1S 2A max buck-boost converter, but somebody needs to write some code for it. It works pretty similar to loneocean’s gxb20: Oshpark Link

would take one for my UV light, a high, medium, low without memory
with tailswitch would be enough

Schoki
Schoki's picture
Offline
Last seen: 5 days 16 hours ago
Joined: 03/27/2017 - 15:34
Posts: 215
Location: Germany

Current range is from 14mA to 2A. Regulator IC is TPS63027, MCU is an ATTiny1634. To build the driver, you need a hot air soldering station or an oven, because the regulator IC is a BGA chip.

Lexel
Lexel's picture
Offline
Last seen: 11 hours 40 min ago
Joined: 11/01/2016 - 08:00
Posts: 5357
Location: Germany

to build the driver is the smallest problem

more is which parts it needs and which firmware it runs

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 29 min 49 sec ago
Joined: 03/24/2016 - 07:44
Posts: 8517
Location: Everything is brighter in Texas

This is why I have not pursued making any next gen drivers. We don’t have any firmware for the latest MCU’s. Until a new MCU is selected and firmware is ported over to it, we can’t really improve past what we have been ding sadly. The Attiny’s we have been using are simply too large.

Once some firmware for the 1617 or 1616 mcu comes out, then we can really start moving things forward.

Schoki
Schoki's picture
Offline
Last seen: 5 days 16 hours ago
Joined: 03/27/2017 - 15:34
Posts: 215
Location: Germany

But still, we can make some schematics and collect new ideas

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 29 min 49 sec ago
Joined: 03/24/2016 - 07:44
Posts: 8517
Location: Everything is brighter in Texas
Schoki wrote:
But still, we can make some schematics and collect new ideas

Indeed you are doing good work. It is just sad that it can’t be used due to lack of firmware.

CrashOne
Offline
Last seen: 2 hours 24 min ago
Joined: 02/18/2014 - 13:29
Posts: 309
Location: The Netherlands

A 22 mm boost driver (2.6-4.3V input, 6V output) with high efficiency (98%+) and high output (6V, 7~8A).

The Miller
The Miller's picture
Offline
Last seen: 1 year 6 months ago
Joined: 12/14/2015 - 12:08
Posts: 9630
Location: Charente France

So the real discussion should be what ia the best MCU to adopt?

If people like Mike C, Tom E and Toykeeper can come to a MCU to use with more pins then the AT85 this opens the way for next gen schematics right?

From my non technical and non programmer way of looking at things I hope the costs and availability of the MCU can weigh in for it will make it easier to implement in BLF specials if it is good to source and affordable, each cent helps here.

CrashOne
Offline
Last seen: 2 hours 24 min ago
Joined: 02/18/2014 - 13:29
Posts: 309
Location: The Netherlands

If that’s the next step and it’s feasible from a costs perspective, make it happen Smile

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 29 min 49 sec ago
Joined: 03/24/2016 - 07:44
Posts: 8517
Location: Everything is brighter in Texas

The Miller wrote:
So the real discussion should be what ia the best MCU to adopt?

If people like Mike C, Tom E and Toykeeper can come to a MCU to use with more pins then the AT85 this opens the way for next gen schematics right?

From my non technical and non programmer way of looking at things I hope the costs and availability of the MCU can weigh in for it will make it easier to implement in BLF specials if it is good to source and affordable, each cent helps here.

Kinda, we have already figured that the latest 1616 or 1617 MCU is the best candidate for the next gen MCU. The issue is the firmware to make it work and being able to program it. Last I knew the programmers had not been updated to flash the new MCU’s yet. Only a matter of time until that happens.

The extra pins are important but the extra space and other features of the latest MCU’s is just as important. Particularly with all the features we are wanting to cram into the firmware now days.

Price is not a big deal, even the latest 1617 mcu is cheaper then the attiny85 IIRC.

Schoki
Schoki's picture
Offline
Last seen: 5 days 16 hours ago
Joined: 03/27/2017 - 15:34
Posts: 215
Location: Germany

Yeah the problem with the 1616/1617 is to program it. It uses the new UPDI interface, and the only programmer right now is the Atmel-ICE.

Jerommel
Jerommel's picture
Offline
Last seen: 4 months 1 week ago
Joined: 01/04/2014 - 13:18
Posts: 5721
Location: the Hague, Netherlands

Meh, i’m probably ‘a stick in the mud’ but who really needs more than 3 or 4 modes spaced like 3% – 10% – 30% – 100% ?
Maybe add a 1% mode and a ‘long press’ to go to the previous mode for lumen canons, but what more do you really need?
You only need low voltage protection to be added, and that’s it as far as i’m concerned.

CrashOne wrote:
A 22 mm boost driver (2.6-4.3V input, 6V output) with high efficiency (98%+) and high output (6V, 7~8A).

+1
It would be nice if it could be easily adapted for 3 Amperes though (for Nichia 144). This could be done with the firmware or by changing the ‘sense resitor’
And no PWM ! please…
A 17mm version would be awesome too.

2Q19

Schoki
Schoki's picture
Offline
Last seen: 5 days 16 hours ago
Joined: 03/27/2017 - 15:34
Posts: 215
Location: Germany

You can change the modes how you like them in the firmware (which doesn’t exist). On my experimental buck-boost board (you shouldn’t order that board, forgot some things), you could have 1024 brightness levels to choose from. But the current difference at the high levels is much bigger (2A -> 1,88A ->…) than on the low end (14mA -> 16mA ->…). That’s just how the circuit works (it’s a voltage divider with an OpAmp, forming a noninverting amplifier).

CrashOne wrote:
A 22 mm boost driver (2.6-4.3V input, 6V output) with high efficiency (98%+) and high output (6V, 7~8A).

We probably have to wait for these drivers until there is a public datasheet for the MP3429/MP3431 (thanks Jensen567, nice chip you found). But if someone finds a similar chip, just post it.
Texas_Ace
Texas_Ace's picture
Offline
Last seen: 29 min 49 sec ago
Joined: 03/24/2016 - 07:44
Posts: 8517
Location: Everything is brighter in Texas
Jerommel wrote:
Meh, i’m probably ‘a stick in the mud’ but who really needs more than 3 or 4 modes spaced like 3% – 10% – 30% – 100% ? Maybe add a 1% mode and a ‘long press’ to go to the previous mode for lumen canons, but what more do you really need? You only need low voltage protection to be added, and that’s it as far as i’m concerned.

Right now we can’t even do this, we have zero firmware for these MCU’s. Once we have some firmware adding all the extra features will be just a matter of porting them from the firmware we already have.

Jerommel
Jerommel's picture
Offline
Last seen: 4 months 1 week ago
Joined: 01/04/2014 - 13:18
Posts: 5721
Location: the Hague, Netherlands

Then maybe stick with the older (cheaper?) MCU for now?
I mean it would be cool if BLF came up with a line of drivers “you want but don’t exist”, like aforementioned 1S to 6 Volts (3 to 6 Amperes) w/o PWM. (i keep mentioning that because i really need one).

2Q19

Agro
Agro's picture
Offline
Last seen: 1 hour 59 min ago
Joined: 05/14/2017 - 11:16
Posts: 4558
Location: Ślōnsk

Schoki wrote:
You can change the modes how you like them in the firmware (which doesn’t exist). On my experimental buck-boost board (you shouldn’t order that board, forgot some things), you could have 1024 brightness levels to choose from. But the current difference at the high levels is much bigger (2A -> 1,88A ->…) than on the low end (14mA -> 16mA ->…). That’s just how the circuit works (it’s a voltage divider with an OpAmp, forming a noninverting amplifier).

Could be fixed by making the mode progression non-linear in time. And one wants visual linearity anyway…
Jerommel
Jerommel's picture
Offline
Last seen: 4 months 1 week ago
Joined: 01/04/2014 - 13:18
Posts: 5721
Location: the Hague, Netherlands

Schoki wrote:
You can change the modes how you like them in the firmware (which doesn’t exist). On my experimental buck-boost board (you shouldn’t order that board, forgot some things), you could have 1024 brightness levels to choose from.
But that’s quite pointless. Only makes it more complex (from an end user point of view).
Obviously the high mode is always 100%, and the next usable mode downward should be 30% or 33% of that, the one below that another decrease of the same magnitude.
Or is this a personal thing?
Every light i own that uses smaller or larger steps i like less in that respect than a Nanjg 3 mode driver light.
Ramping is not user friendly either (in my opinion), you always end up where you didn’t really want to end up or it has to be very slow.
It could work well with 2 buttons (up and down) though.

2Q19

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 29 min 49 sec ago
Joined: 03/24/2016 - 07:44
Posts: 8517
Location: Everything is brighter in Texas
Jerommel wrote:
Then maybe stick with the older (cheaper?) MCU for now? I mean it would be cool if BLF came up with a line of drivers “you want but don’t exist”, like aforementioned 1S to 6 Volts (3 to 6 Amperes) w/o PWM. (i keep mentioning that because i really need one).

The new MCU’s are cheaper by about 10-20% from the prices I have seen.

We can’t use the existing MCU’s for new driver designs due to space constraints. A tiny85 takes up almost half the usable space on a 17mm driver, there is no room for anything else except what we have been doing.

The new MCU’s come in tiny 3×3mm packages that allow the space for new driver designs.

For example, that driver that was posted earlier in this thread is impossible with a tiny85, just not enough space.

Jerommel
Jerommel's picture
Offline
Last seen: 4 months 1 week ago
Joined: 01/04/2014 - 13:18
Posts: 5721
Location: the Hague, Netherlands

Okay, thanks for explaining. Thumbs Up

2Q19

The Miller
The Miller's picture
Offline
Last seen: 1 year 6 months ago
Joined: 12/14/2015 - 12:08
Posts: 9630
Location: Charente France

Jerommel, just wait till you have played with Narsil and things change for you Wink
There is a lot I really want added and even deem important like a couple of “safe mode groups” limiting both modes and highest output, so when handing a light to somebody who knows nothing it ia safe for example.
Not only for safety but also pretty useful in small lights, find the best top mode it can run without overheating

But also I want two switches operating two sets of LEDs separately for the Q16. I think we would need two drivers now but with more pins maybe one driver can do it (yeah sorry I am a technical newb)

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 29 min 49 sec ago
Joined: 03/24/2016 - 07:44
Posts: 8517
Location: Everything is brighter in Texas

A single driver can do that, the only issue would be space depending on how the LED were controlled. Although unless you are talking about different color LED’s you are pretty much always better off driving all the LED’s at a lower current then running a few of them harder.

The Miller
The Miller's picture
Offline
Last seen: 1 year 6 months ago
Joined: 12/14/2015 - 12:08
Posts: 9630
Location: Charente France

Well needing two sets of LEDs each set of two having a different function (2 for throw and 2 for flood). Makes a lot of sense to be able to choose either or both for max output.

Agro
Agro's picture
Offline
Last seen: 1 hour 59 min ago
Joined: 05/14/2017 - 11:16
Posts: 4558
Location: Ślōnsk

BTW, how about some waffer level package chips? These can be nearly 3 times smaller than Attiny1616.
Example here
Not sure if it has all that is needed. Not sure if it’s hand-solderable. And it’s slightly more expensive than Attiny. But is significantly more powerful while saving another 5.7 mm².

mrheosuper
Offline
Last seen: 1 day 4 min ago
Joined: 09/30/2016 - 12:44
Posts: 1473
Location: Ho Chi Minh City, Vietnam

is there any driver that i can modify the current limit by changing resitor?( voltage divider, not shunt resitor)

Forgot my pen

Jerommel
Jerommel's picture
Offline
Last seen: 4 months 1 week ago
Joined: 01/04/2014 - 13:18
Posts: 5721
Location: the Hague, Netherlands

I think the LD-3 and the new LD-4 from led4power can, or the LD-25 sold by Simon.
(All FET CC drvers)

2Q19

mrheosuper
Offline
Last seen: 1 day 4 min ago
Joined: 09/30/2016 - 12:44
Posts: 1473
Location: Ho Chi Minh City, Vietnam

Jerommel wrote:
I think the LD-3 and the new LD-4 from led4power can, or the LD-25 sold by Simon.
(All FET CC drvers)

there are some limit, the range of changing current is quite narrow
(2 to 3amp on 3amp version), what i want is wider current selection, like 0.5 to 10A, all determined by a fomular with 2 resitor

Forgot my pen

led4power
led4power's picture
Offline
Last seen: 2 hours 30 min ago
Joined: 12/29/2012 - 09:48
Posts: 1210
Location: Croatia,EU

mrheosuper wrote:
Jerommel wrote:
I think the LD-3 and the new LD-4 from led4power can, or the LD-25 sold by Simon. (All FET CC drvers)
there are some limit, the range of changing current is quite narrow (2 to 3amp on 3amp version), what i want is wider current selection, like 0.5 to 10A, all determined by a fomular with 2 resitor

You can change current down to ~50% WITHOUT any resistor change, just via config. Settings.

By changing one resistor,you can change current to practically any value you want.

 

 

mrheosuper
Offline
Last seen: 1 day 4 min ago
Joined: 09/30/2016 - 12:44
Posts: 1473
Location: Ho Chi Minh City, Vietnam

led4power wrote:

mrheosuper wrote:
Jerommel wrote:
I think the LD-3 and the new LD-4 from led4power can, or the LD-25 sold by Simon. (All FET CC drvers)
there are some limit, the range of changing current is quite narrow (2 to 3amp on 3amp version), what i want is wider current selection, like 0.5 to 10A, all determined by a fomular with 2 resitor

You can change current down to ~50% WITHOUT any resistor change, just via config. Settings.


By changing one resistor,you can change current to practically any value you want.


 


 


Woah, you didnt mention that we can change resitor to get the current we want

Forgot my pen

Jerommel
Jerommel's picture
Offline
Last seen: 4 months 1 week ago
Joined: 01/04/2014 - 13:18
Posts: 5721
Location: the Hague, Netherlands

mrheosuper wrote:
led4power wrote:

mrheosuper wrote:
Jerommel wrote:
I think the LD-3 and the new LD-4 from led4power can, or the LD-25 sold by Simon. (All FET CC drvers)
there are some limit, the range of changing current is quite narrow (2 to 3amp on 3amp version), what i want is wider current selection, like 0.5 to 10A, all determined by a fomular with 2 resitor

You can change current down to ~50% WITHOUT any resistor change, just via config. Settings.


By changing one resistor,you can change current to practically any value you want.


 


 


Woah, you didnt mention that we can change resitor to get the current we want
I have to know which one on the LD3, i’ll PM you (led4power).
because i bought 6 of them, all 12 Amperes, which in hind sight wasn’t all that smart.. Facepalm

2Q19

Pages