NejTech Evolution - a new BOOST driver - Nearly finished!

[Eudaimonium]

Re: NejTech Evolution - a new BOOST driver - preparing to build prototypes

My comment was directed at Johnyz, not you Eudaimonium. (btw, you need a shorter, less hard to type when half asleep name. )

How about "Eud" or "Z"? You're not obliged to type out entire nickname, you know.

They actually are 0.27 ohm, it's just Eagle3D not rendering it correctly. Thanks for the capacitance tip.
EDIT: Darn, 22uF is the biggest 0603 cap I can get... might need to switch to 0805 for some caps.

Oh, cool.

So it is LM3410 or sorts of.

I am using 10, 10, 20 uF for input, schottky and output capacitance, respectively. It does very well at 1.15 A output.

Make sure your schottky can handle at least 1.5A of current, and that your coil is precisely calculated. To be honest, I picked mine based solely on inductance and saturation current, and they are doing well. If you precision pick yours, the final product should be exeptionally well made and will do well in handheld hosts.

Respect.

 

[Kmor2004]

Re: NejTech Evolution - a new BOOST driver - preparing to build prototypes

This is interesting indeed Johny and yours looks pretty good also Eud.

If I may make a comment, Johny using eagle 3d it may help to select user defined parts if you didn't already you'll have to select the proper items but if the end result looks something like

and not looking like you forgot important parts on a board., other then that looks great better then my attempt at designing a <17mm round driver board.

 

[Johnyz]

Re: NejTech Evolution - a new BOOST driver - preparing to build prototypes

Eudaimonium: Okay... Let's do some math then. My schottky? It can handle TWICE that!
Kmor: You know, there's simply no model for the inductor, and the rest is good.

EDIT: The datasheet suggests ferrite core inductor, but these are, at the appropriate current rating HUGE!
EDIT2: Layout slimmed down even more. Inductor decision made. I'm really looking forward to the prototypes.
EDIT3: Prototypes will be ordered tomorrow!

 

[Kmor2004]

Re: NejTech Evolution - a new BOOST driver - preparing to build prototypes

There are 2 inductors, but ones PTH which would be too big, and I think I've used teh the otehr one in design before and its also too big, forgive me not looking in to what I was saying.

 

[Johnyz]

Re: NejTech Evolution - a new BOOST driver - preparing to build prototypes

Boards ordered, parts ordered. Let's do this!

 

[random person]

Re: NejTech Evolution - a new BOOST driver - preparing to build prototypes

Haha, looks making your own boost driver is the craze now. =P Should be getting the boards for my driver in a week now. Been on order for awhile. 3A output, 3V-8V input, 1% load line regulation. >4 total bom cost (numbers are ideal and not tested yet, more practically speaking i'd peak it at 2.2A to be safe)

Real question is what you plan on selling them for in the long run. The issue I have with the 1.4A drivers is that you would want 2 of them to in order to full power a 445nm blue. Jib boost looks like the way to go now because you would be able to full power a 445 nm for 21 bucks I think...? So unless you price yours below 21 / 2 , most people would go with jib boost.

 

[Johnyz]

Re: NejTech Evolution - a new BOOST driver - preparing to build prototypes

Personally, I will be happy if the driver market that used to contain one or two available drivers will expand. We have lots of diodes to select from, even more hosts, and nearly all usually contain the same type of driver. It will only be good for the community if there are lots of good drivers available.

 

[qumefox]

Re: NejTech Evolution - a new BOOST driver - preparing to build prototypes

The reason there are more choices in the hosts and diodes is because their development in general is driven by much larger markets outside of the laser hobbyist market. The only people interested in laser diode drivers that suit our applications is us. And most hobbyists are 'consumers' not 'producers'.

 

[Johnyz]

Re: NejTech Evolution - a new BOOST driver - preparing to build prototypes

Exactly. And I'm very happy to see more producers pop up. You know, a good part of this hobby is DIY and we are helping the community by making more of these rare products.

AAAAAH: Farnell 22uF caps out of stock, will be stocked in 86 days... I need to contact them and pick different ones.

 

[random person]

Re: NejTech Evolution - a new BOOST driver - preparing to build prototypes

Exactly. And I'm very happy to see more producers pop up. You know, a good part of this hobby is DIY and we are helping the community by making more of these rare products.

AAAAAH: Farnell 22uF caps out of stock, will be stocked in 86 days... I need to contact them and pick different ones.

I'm guessing your using the LM2733?

National is a decent choice because most of there ICs have a nice high switch frequency. This means your inductor and caps can be smaller. Make sure your SW frequency is above 1MHz.

Disadvantages is if your buying from digikey, odds are your IC is going to be 3usd-4 usd.... which is abit too pricey. Especially if you will need 2 to run a 445nm.

Also, disadvantage of those sot23-5 based drivers is they don't have a soft start option. If you take a look in LT spice of the turn on characteristics of small drivers like that, they have pretty hideous spikes. How do you plan to compensate for that? Its not ideal sensitive load.

As far as the cap goes.. This your output cap? 22uFs are a dime a 1000... Making sure you have the correct cap is far more important. I assume ceramic if its an output cap...? Whats your ESR level? You should select your cap such that the min peak impedance is at the native switching frequency of your supply. Lower ESR, less ripple... Decent datasheets have a plot of the ESR or ESI versus freq.

Best of luck to you with your driver

 

[Wolfman29]

Re: NejTech Evolution - a new BOOST driver - preparing to build prototypes

I hate how none of the boost OR buck ICs show how to make them current regulators on their datasheet in the examples. Where do you people figure out how to go from CV to CC with these ICs?

 

[qumefox]

Re: NejTech Evolution - a new BOOST driver - preparing to build prototypes

Um, that's where 'electronics theory' comes in.. When you know how parts work together, you don't *need* reference examples.. Do you think the people who put the examples in the datasheet, just looked up how on another datasheet? heh

 

[random person]

Re: NejTech Evolution - a new BOOST driver - preparing to build prototypes

I hate how none of the boost OR buck ICs show how to make them current regulators on their datasheet in the examples. Where do you people figure out how to go from CV to CC with these ICs?

a suupperrr seecrreettttt! :shhh:

measure the voltage across a resistor and you get a voltage that is proportional to current

one way to do this is use a op amp circuit and boost the voltage gain over a current sense resistor in series with your output. The advantage of this is that your dropout voltage is as low as possible. You can also put a pot on the gain and then you don't have to worry about putting your pot inline with the current.

Or you can get a regulator with a low reference voltage in which case you can feed the voltage across the Rsense resistor directly to the FB pin. In my opinon this is a slightly worse idea because your dropout voltage is
is increased and your using more power the higher resistance you use. Uses less parts though. However, in both cases this is dangerous unless you take the appropriate filter considerations since the high freq boost ICs are meant to run with a compensation network.
Take a look at this circuit again by Eudaimonium to see how he current limits it

watch this video... its verrry helpful! http://www.youtube.com/watch?v=cM7t1Mpu7s4

Um, that's where 'electronics theory' comes in.. When you know how parts work together, you don't *need* reference examples.. Do you think the people who put the examples in the datasheet, just looked up how on another datasheet? heh

thy has the capacity to be of more help then that... =P

 

[Wolfman29]

Re: NejTech Evolution - a new BOOST driver - preparing to build prototypes

@Qumefox: Yeah, I wish I had electronics theory. But my question is more related to the fact that I don't necessarily understand how each of these ICs work, with all of their quirks, etc. I understand the basics of switching buck or boost regulators. It's just going from that to the ICs that I don't get. And... voltage regulation is a lot easier than current regulation -.-

Is it true that if I essentially just stuck a resistor in series with the load with a given voltage, i.e. set the boost to boost to 10V, and then just stuck a 10Ohm resistor in series with that output, would it give me 1A of current at the load, max?

@random person: Hmm. I haven't delved much into op-amp circuits, so I am not that familiar with them. But I think I understood what you said regarding the linear regulator. Anyway, I am going to check out that video!

 

[qumefox]

Re: NejTech Evolution - a new BOOST driver - preparing to build prototypes

thy has the capacity to be of more help then that... =P

Yep. However people don't learn anything when you tell them answers without them working to find it themselves. Though I seriously suggest studying up on op-amps wolf. Give me an op-amp and i'll rule the world!

 

[Johnyz]

Re: NejTech Evolution - a new BOOST driver - preparing to build prototypes

Oh wow. Now that LPF is back up I need to share one amazing experience with Farnell. I ordered my IC's, coils, passives etc. at Farnell yesterday. Shipping was 130 CZK (about $8) for UPS (!) and it arrived overnight! Amazing! Now I just need my boards...