LM3410X based boost converter, SPARK driver! PART 2

[Eudaimonium]

It's still a hobby for me too, I enjoy designing and making these performance drivers so that all kinds of tiny builds are possible for all forum members. Every time it's like a new puzzle that needs to be solved as to how much can be crammed into such a small space, and maintain as much flexibility as possible. In the past I've gone the route of hand assembling drivers whenever someone needed one, and after a few months it's just not maintainable. But when you have the ability as you do, making them for yourself and a couple (not wearing yourself out) for others is great.

Well said, dude!
I can sure hand solder a sh't load of drivers, given the etched PCBs are already done.

Oh and I'm purchasing a new soldering station next week... somebody stop me!

Puzzle to solve, sure is, when a driver like Flex has so many components.
But you can easilyfabricate double sided PCBs.

I have to do with single sided, and routing the traces is not easy. Well, for first few times, later you can already do it without having the IC handy , to use it while drawing on PCB - you just kind of learn the spacings and dimensions by heart.

But your drivers are unsurpassed. What is the efficiency of a flex drive when powering GGW (200mA) from one lithium ion (full charge but given the voltage drop down to 4V when powered)?

I calculated mine to be around 67.8% since it's drawing 350 mA.

Thatot exactly ideal, but it's better than cramming 2x CR123 and DDL driver into the thing.

hey Happy man (heheheh) remember how you helped me with that blu ray driver???
i salvages through probably 11 old things and got the parts for yourdriver!!!!! i made one (geez there hard to solder SMD. lol but i hooked it up to 6V and a test load and the voltage reg like craked in half :crackup::crackup: maybe i should ask help before doin it next time i failed misserably

If you are going to insult my nickname, at least do it with style.
You got it wrong.

Discussed via Skype, wrong IC used. Better luck next time.

 

[TTerbo]

lol Euadimonium you easily have the best name on here (i think) and i want paying it out

 

[drlava]

But your drivers are unsurpassed. What is the efficiency of a flex drive when powering GGW (200mA) from one lithium ion (full charge but given the voltage drop down to 4V when powered)?

I calculated mine to be around 67.8% since it's drawing 350 mA.

Hi, thanks for that. I went and measured it for both the FlexDrive and the MicroBoost drive, with 200mA out into about 5.4V with a 4V input.

The FlexDrive operates at 93% efficiency and the MicroBoost at 86% efficiency. However the MicroBoost is still the driver I would suggest for the Blu-ray build due to its new design and components (sturdier potentiometer for those who like to tweak a lot).

 

[Eudaimonium]

Hi, thanks for that. I went and measured it for both the FlexDrive and the MicroBoost drive, with 200mA out into about 5.4V with a 4V input.

The FlexDrive operates at 93% efficiency and the MicroBoost at 86% efficiency. However the MicroBoost is still the driver I would suggest for the Blu-ray build due to its new design and components (sturdier potentiometer for those who like to tweak a lot).

:drool::drool:

Dooode, that means that when powering 200mA bluray diode at 5.4 V (too much ? ) which is 1.08W of electrical power output, the driver will produce no more than 80 mW of heat, so total consumption power is 1.16 W, with 4V input the driver will pull 290 mA !!

Opposed to 350 mA pull of my driver, I'm dwarfed!

Your driver produces 80 mW of heat, mine around 450 mW,

IF my quickie calculation is correct...

Is it?
:bowdown:

 

[drlava]

Yep, this is how I calculated it, all parameters were measured at steady state:

FlexDrive:
293mA in, 4V in =1.172W in
201mA out, 5.42V out = 1.089W out
1-(PIn-POut)/PIn = 0.93 (93%)
driver power dissipation 83mW

MicroBoost:
315mA in, 4V in =1.26W in
201mA out, 5.42V out = 1.089W out
1-(PIn-POut)/PIn = 0.86 (86%)
driver power dissipation 171mW

As an aside, I have an older Sonar build I did with a FlexDrive and PHR diode running at 60mW and 17670 cell over a year ago. Use it occasionally, and have never had to recharge it. Build provides 14 hours continuous runtime.

 

[Eudaimonium]

Yep, this is how I calculated it, all parameters were measured at steady state:

FlexDrive:
293mA in, 4V in =1.172W in
201mA out, 5.42V out = 1.089W out
1-(PIn-POut)/PIn = 0.93 (93%)
driver power dissipation 83mW

MicroBoost:
315mA in, 4V in =1.26W in
201mA out, 5.42V out = 1.089W out
1-(PIn-POut)/PIn = 0.86 (86%)
driver power dissipation 171mW

As an aside, I have an older Sonar build I did with a FlexDrive and PHR diode running at 60mW and 17670 cell over a year ago. Use it occasionally, and have never had to recharge it. Build provides 14 hours continuous runtime.

Is 5.42 V of diode measured or estimated?

Because I think I did my calco with 5V output, not 5,4.

Hmm, is I have an 5.4 V output...

1.2 W input power and 1.045 W out, hence my driver dissipates 155 mW of heat, NOT 350 mW,
Also that brings up the efficiency to 74.51 % which is cool.

I was calculating your 93% flex, I was off by only a couple of mA and mW.


So, 150 mW of heat to take care of, not 350 ...
I though my driver was cooking, but I guess I can allow myself some larger duty cycles without much fear of burning out the driver.

And I was kind of thinking that the gap between current draw and driver heat dissipation was kind of big... :crackup:

 

[charliebruce]

350mW? That's nothing! A driver I made quite some time ago dropped 5V linearly down to 3.2V at 2.8A (targeted) to run a high-power LED. That's a theoretical max of 5W but in reality the thermal limiter kicked that right down. I worked out that in real usage terms, each chip could dissipate 350mA. Here's it with an IR camera...

Really inefficient, but functions perfectly to this day. The power limiter prevents it from killing itself. I didn't have to make it run at full efficiency for the project, only to make it light and be brighter than a standard torch. When the project is fully marked and approved, I might disassemble it and make myself a proper buck driver for it, so I can run the LED at full potential. I'll also release proper pics, but until it's marked I'd prefer to keep it to myself.

The LED itself keeps remarkably cool though - it's heatsinked in a Kryton-like host.

/threadjack.

 

[drlava]

5.42V is measured because I used a 27 ohm power resistor as a dummy load, because of its stability. So if you are using a diode as your load you need to measure its forward voltage and all input parameters after it heats up to calculate your efficiency.

 

[Eudaimonium]

5.42V is measured because I used a 27 ohm power resistor as a dummy load, because of its stability. So if you are using a diode as your load you need to measure its forward voltage and all input parameters after it heats up to calculate your efficiency.

Hmm but what is the real Vdrop of GGW diode at 190-200mA?
If it's 5V, then the driver really does dissipate 350 mA, if it's 5.4 then it's up to 76 % efficiency...

Hmm, I don't think I'd care to dissassemble the laser now, risk breaking it and all, I gotta find somebody to measure the Vdrop of GGW diode at 200mA to get my calculations straight...

 

[morganw]

Hey Eudaimonium, I know this is an old thread but can you post the schematic of your driver, I couldn't really get a straight answer of the component values from the data sheet for the LM3410X.

Also, I have an idea that I used on a different circuit and worked very well as to making a regular trimmer resistor work in your circuit so the current can be easily adjusted!!! Just the current set resistor would set the minimum current. If it works good I will post it.