check this out from national semicon...

[Bionic-Badger]

How do you get copyright on a product?

It strikes me that you could probably get copyright on a PCB layout, or the trade dress (visual appearance of your product), but I question whether you hold a valid copyright on a schematic or circuit.

In the US and most other places, the very fact that something is created implies a copyright. You have to explicitly relinquish rights to your creation to make it open (hence license agreements). In fact, in some places in Europe, you can't even place something in the public domain, but rather need to explicitly grant free reign over a design.

I also highly doubt most of these schematic designs can even be patented, as they're based on generic designs, for which prior art would be easy to claim. The graphics used to draw the schematic can be claimed for copyright purposes, but just transcribing them would be sufficient to not need licensing, etc. as they contain little that is actually unique.

 

[benmwv]

Uhm, sorry, but i think this must be said, also if it can blow some dreams

When reading datasheets and designing switching mode drivers, there are some facts that must be kept in consideration .....

First, when you read "Switching current 2.1A", it does NOT mean that the driver can do 2.1A ..... that is only the INTERNAL switching current ..... and, indicatively, for buck drivers, the maximum safe current that you can obtain is half of this current, where instead for boost/sepic drivers, the maximum safe current is usually 1/3 of this value (i said "safe" current intentionally ..... overdriving the chip you can reach some more current, also 2/3 for buck and half for boost, but this stress a lot the chip, putting it in a non-safe working condition ..... this can end in a failure, and if this is acceptable for a LED, is not so much happy when it happen with your high-cost new LD connected )

Same is for the inductor saturation current, that must be choosen usually AT LEAST the double of the expected output current, sometimes also 3 times higher.

As example ..... i got one prototypes with LM3410X that reached 1.2A and worked for long time (heating a lot, heatsinked, and using a 10uH 4A coil), but also burned another one with the same current ..... the normal output current for a LM3410 chip used as boost/sepic driver, is around 800 / 900mA, and this chip is marked as "2.8A" switching current ..... it can be increased a little, using larger output capacitors, but then the driver becomes "a bit too big"

Well Im sad to see you were only able to get 1.2A, but no dreams were crushed I assumed that only boosting from 4.2v to ~4.5v it would be safe to say it could do 1.4A (half the switch current). But maybe I was wrong. And as the battery drains it will get harder on the driver to boost...

But I am fairly positive I can get it to do 900ma at least, and that will make it perfect for paralleling to 1.8A

Also my inductor is 6.8uh 4A. Do you think this is adequate?

 

[rhd]

I also highly doubt most of these schematic designs can even be patented, as they're based on generic designs, for which prior art would be easy to claim. The graphics used to draw the schematic can be claimed for copyright purposes, but just transcribing them would be sufficient to not need licensing, etc. as they contain little that is actually unique.

Exactly. Regardless of the whole generic design issue, even with a completely original schematic, that circuit is an idea. That's not something that you can copyright. Like you've said, the graphics (and even the exact representation of the circuit on the diagram) could possibly be argued the subject of copyright. However, what is actually going on in the schematic cannot. I could look at your schematic, and draw the same circuit in my own schematic layout software, and there would presumably be no copyright violation. This isn't legal advice.

 

[HIMNL9]

@benmwv: for the 1.6MHz version (ends in X, opposite to the lower frequency one that ends in Y) and for 900mA, 6.8uH are probably enough, maybe at the limit (they suggest usually 10uH) ..... not for the low frequency version, that usually needs 15uH or more .....

 

[foulmist]

I was playing with the national semiconductor webbench(c) designer
LINK HERE on their website...

actually that's a pretty good program, you type the INs the OUTs Voltage/Current and it gives you the best solutions for your needs...

like IC/Schematics/PCB Layout/External Components/Oscillations and everything you can ever dream of I think you should try it if you haven't already..

so I put a query for a boost driver for say 1 Li-On max and Iout 2A, at 6V and I got results which require some external components... not much but some
are pretty big for a handheld driver.. and we are talking about very stable driver.... the results were LM3478/LM3481/LM3488, the caps and inductors are actually the biggest components but so is the IC

input some driver requirements and see what you will get, it's actually pretty interesting program :beer:

 

[benmwv]

@benmwv: for the 1.6MHz version (ends in X, opposite to the lower frequency one that ends in Y) and for 900mA, 6.8uH are probably enough, maybe at the limit (they suggest usually 10uH) ..... not for the low frequency version, that usually needs 15uH or more .....

I am using the high frequency one, thanks. I have a question, on your driver did you connect the DIM pin directly to Vin or use a resistor?

 

[HIMNL9]

I used a 10Kohm resistor ..... i was not planning to use the dimming function, cause it's not analogic, only a PWM input, but left open the possibility (and after all, one more smd resistor don't add any complication to the circuit )

I also tried once to use it as "soft-start" in one prototype (using a resistor to Vin and a capacitor to GND), but it does not work in this way, it only add a delay to the turning-on time, so i took that modification away.

 

[benmwv]

I managed to fry the ic on my attempt, and I'm pretty sure it's because I somehow looked over that resistor... Have to get new boards now

 

[HIMNL9]

Uhm, maybe too long tracks from the pins 1 and 5 and the coil ? ..... it must be kept as shorter as possible, and of a discrete size (no thin tracks)

(EDIT: pins 1 and 5, assuming you are using the SOT23 package like mine)

Also, you don't have connected for error the dim pin to the output, instead than to the input, right ? ..... this will fry the chip for sure (dim voltage must NEVER be greater than Vin) .....

 

[benmwv]

I'm pretty sure it fried because I didn't have a resistor between vin and dim.

In the test it drew spike of ~750ma then quickly down to 500ma. It output 23ma and it got really hot so I'm thinking all that power went through the dim pin a converted to heat killing the IC.

 

[HIMNL9]

Uhm, i think not ..... the dim pin is simply a high impedance logic input, like a cmos port input ..... also if connected directly to Vin, it cannot draw more than 1microampere, and this cannot cause any overheating .....

Is more easy an open-output or overvoltage failure, cause the LM3410 does not have internal protection, for this event ..... it have an overcurrent protection, and also a thermal shutdown protection, but there is no internal overvoltage protections, especially for the output ..... in case of no limitation, or insufficent FB voltage (like, as example, if the RSET is too low or if the FB pin go shorted to GND, or if the load get disconnected), the IC simply increase the output voltage til it go over the safety limits and fry.

 

[foulmist]

hey RHD did you come up with some design

 

[rhd]

hey RHD did you come up with some design

I started working on one, but then it became clear that another forum member was already doing much better work than I could have, and had a similar positive intention to open source / release the design.

So I decided not to duplicate efforts. Or rather, I decided not to poorly approximate a far superior effort already underway elsewhere

 

[foulmist]

haha yeah .. you are right..

well I guess we'll have to wait for his to come out :beer:

 

[benmwv]

Uhm, i think not ..... the dim pin is simply a high impedance logic input, like a cmos port input ..... also if connected directly to Vin, it cannot draw more than 1microampere, and this cannot cause any overheating .....

Is more easy an open-output or overvoltage failure, cause the LM3410 does not have internal protection, for this event ..... it have an overcurrent protection, and also a thermal shutdown protection, but there is no internal overvoltage protections, especially for the output ..... in case of no limitation, or insufficent FB voltage (like, as example, if the RSET is too low or if the FB pin go shorted to GND, or if the load get disconnected), the IC simply increase the output voltage til it go over the safety limits and fry.

I originally thought it was high impedance, but now I have found that it needs a resistor (100k on the datasheet examples) between dim and vin so I am thinking that it isn't. Idk

I tried connecting it to a 6, 5, and 4 diode test load. I had one set to 1.4A and one set to 1.2A, both put out 23ma and got really hot. Do you think the real problem was that it couldn't achieve the set current and over-volted? I didn't have resistors for 900ma. I'm going to order sone more parts, including a 10uH inductor and resistors for various currents. Ill start with the resistor for the smallest current then work my way up and see what happens.

 

[HIMNL9]

The 100Kohm resistor suggested from the datasheet is simply for keep it in "high" state, when you plan to use the dim pin as PWM driver pin ..... also, just the fact that they suggest you to use 100K resistor, let you know that the input IS a high impedence input (if not, 100K was not enough for keep it in high state ..... the lower is the input impedance, the lower is the resistor value needed for keep it in a specific state) ..... with 100K pull-up resistor, it must be over 1Mohm impedance, probably around 10Mohm.

If not used as dimming pin, it can be connected directly to Vin without any problem, just remembering to use the shortest track possible, for not catch disturbs from the switching section.

About the problem, i think is possible ..... especially with lower inductance values (this also increase the ripple a lot) and insufficent output capacitor ..... also the input capacitor is important, cause the inductance is directly connected to Vin too (practically, you "short" the coil from Vin to GND at each charge cycle, so you need a decent input capacitor)

Also, remember that the LM3410 is a boost-only chip, in one-coil configuration (for work as sepic, you need to use the 2-coils circuit), so the voltage that is set for the output MUST be higher than the Vin ..... as example, for a 3.7V charged Li-Ion cell, the Vout MUST be over 3.7V (indicatively, at least the Vin + 0.3V), otherwise the chip don't work correctly ..... cause we use it as current driver, the Vout is given from the FV of the circuit connected to the output, so 6 diodes dummy load is ok, 4 diodes dummy load is a bit on the "too low limit" , maybe