2xAA blue ray driver (fits in aixis)

[IgorT]

the datasheet limits Rin to 500ohms max...

Indeed it does. So what does that mean for this circuit?

 

[woop]

the datasheet limits Rin to 500ohms max...

my bad :-[
I just realized i screwed up that equation, it should be: (current) I = 0.19 * (Rin / Rout) / Rsense
so as Rout gets smaller, the output current will get larger.

ok just change the resistors.
what current do you want?
should the circuit be adjustable via a pot?

this should work (not using a pot)
Rsense=0.1ohm
Rin=500ohm
Rout=10kohm
will give you 95mA

i attached a schematic.

 

[woop]

So you finished this thing like an hour after we first talked about it?!?  Shocked
I'm still shocked. I thought it was just meant as proof of concept (that it fits). So it would just work?!? With the LTC6106?

lol yeah. it should just work. current mode smps's are not as hard as they look, although i have not worked out the accuracy due to the extra gain of the amplifier, it probably doesn't matter.

what, did you think that i just threw some components together blindly?  it doesn't take much longer to arrange them properly

I am definitely going to order some of the 6106's, i have some other uses for them too

 

[IgorT]

Isn't the one you drew yesterday already adjustable? I thought you said there was a pot on it.

I don't really NEED a pot. I'm happy with changing resistors. BUT!

When weird value resistors have to be used or i can't get SMD in small qantities, a design that could use the same few resistors for all applications and a have pot, would actually be superior. Especially, if i would end up having to use a through hole resistor, that would take up more space than an SMD pot.


I did think you just threw the components together "blindly", because you said "just to prove that it fits".

 

[woop]

what current do you need?

and to make it adjustable, a pot can be put in series, it would lower the current. a 10k pot would vary the current from 40-95mA

and that board layout i put up before needs an extra resistor (in series with the pot) and another cap on the high side of the current measuring resistors

 

[phenol]

ok, suppose Rin=100R and Rout=10k, adjustable. We have a gain of 100. For Rsense=0.1R and Vref=190mV, the min I=1.9mV/0.1=19mA. The max current 6106 is capable of sourcing is 1mA, for 190mV this would be 190R for Rout min. If Rout min is 330R (Resistance of trimmer pot=0 + a fixed 330-R resistor), gain=3.3 and Imax=57.58mV/0.1 = 576mA.
Sounds good.
Transient response and stability have to be verified under all conditions, though.

 

[adgmeijin]

hey woop where did you get the parts of your boost driver? and how do you test the output and current  using DMM?

 

[danq]

Re: prices, availability


@ Newark:
LT6106 - $1.42, or 25 for $1.17 each
LM3410 - $2.80 each (there is a SOT-23 version, same price)
ZXCT1009 - $1.00 each, but $15 special handling fee! (this part shipped from UK)

coils: @Newark - 10uH .066ohm $0.22 each #NP04SB100M
@Nightfire Electronics (eBay) - 3.3uH 0.018ohm $1.50 for two

Schottky: STPS2L25U 2A 25V Vf=0.325 100 for $6.00 @ eBay RFBasic

also have good cap sources but gotta get to work
;-)
DanQ

 

[IgorT]

hey woop where did you get the parts of your boost driver? and how do you test the output and current using DMM?

I hope you realize, this is not like putting a 317 together. If you want a laser fast, i suggest going with Dr_Lava's driver. If you want to spend a month or two sampling components and looking for them in old or broken electronic devices, etch your own PCBs and solder ICs without legs, well....


We got the main components as samples from the manufacturers. But they can sometimes be picky about whom they send free samples to. The rest was scavanged out of DVD drives and so on. You need a DMM capable of measuring capacitance, because SMD caps are not marked.

And the soldering can really go wrong, if it's the first time you're doing something like this. The IC is really easy to destroy. So are the other components. Just removing them from the original devices can destroy them if not done right.


In the end you won't really save any money, compared to buying Dr_Lava's circuit. But you may learn a thing or two, so...

 

[IgorT]

what current do you need?

For the boost version more than 200mA wouldn't be required for a long time (405nm diodes). But for the sepic version, it would be nice to have it up to 500mA or more if the IC is capable of it.

Still, the pot shouldn't be adjusting too big of a range. The range should still be set with resistors, and the pot would only adjust the little in between. Otherwise, it could be dangerous, since it's a tiny single turn pot.


But since all this can be solved by selecting correct values resistors, it's really not important right now. I'm just happy i have something i can build and use, once i have some time on my hands again. Don't want my ICs and inductors to go to waste...

 

[adgmeijin]

thanks Igort , actually i set an order for the sample of Dr. Lava's microflex boost driver, i think this may be the solution to make a pen laser with PHR803T LD on it.

 

[IgorT]

thanks Igort , actually i set an order for the sample of Dr. Lava's microflex boost driver, i think this may be the solution to make a pen laser with PHR803T LD on it.

If you grind the sides away a little so it's only 10x12mm, then yes, you can power any 405nm diode with it from two alkalines or even Ni-MHs in a pen laser housing. Something, that is impossible with linear regulators.


You will like it very much. Just don't try to measure the output voltage with no load! Nor power it on with no load! It's explained in the instructions.

 

[woop]

ok, suppose Rin=100R and Rout=10k, adjustable. We have a gain of 100. For Rsense=0.1R and Vref=190mV, the min I=1.9mV/0.1=19mA. The max current 6106 is capable of sourcing is 1mA, for 190mV this would be 190R for Rout min. If Rout min is 330R (Resistance of trimmer pot=0 + a fixed 330-R resistor), gain=3.3 and Imax=57.58mV/0.1 = 576mA.
Sounds good.
Transient response and stability have to be verified under all conditions, though.

that sounds like a good combination.
the only thing is that the current is proportional to 1/Rout, so as the current gets higher, smaller turns of the pot will have a bigger effect.
the only way to overcome this would be to setthe max required current and only use the pot to decrease the current

 

[phenol]

yes, by using different values for Rmin, Imax could be limited accordingly

 

[woop]

yes, by using different values for Rmin, Imax could be limited accordingly

thats what i was getting at.

well here are some updated board layouts, top views, and overlayed (with the red on top)
and phenol, do you think C3 in my schematic is necessary? i included it because it is the supply for the amp, and what value should it be?

 

[phenol]

yes, C3 would be a good idea. 2.2uF ceramic and above should perform fine suppressing HF spikes. For C2, in my experience with SEPIC configuration, a 10u tantalum (10V or better yet 20V) should be sufficient. C1 should be smth like C3, ceramic, connected to pins 1 and 2 using the shortest possible tracks.
Also, to further suppress any HF spikes that may have made it to the LD pins, a small SMD ceramic cap (10-100nF) directly across anode and cathode would also be a good idea.