Help Needed: Using pcb cad software to design board for this micro ic I have

[wirednuts]

ok, so basically-

i have this for my phone

then i bought some of these to upgrade that phone charger
http://www.onsemi.com/pub_link/Collateral/NCP1835-D.PDF

and i want to use these guys to make me some custom pcb's
http://www.futurlec.com/PCBService.shtml

and this is the software they say i can choose from. just send them the final project file and they can make the pcb

all im asking right now is what the best software i could use in my case. i have never done this before, and im not familiar with most cad stuff. im pretty green, but i do understand electric circuits and can solder, just not something this small. thats why i have to get custom pcb's made, there is just no other way. (im not making them myself, not when for $30 i can have 12 of these made).

 

[CycloWizard]

ExpressPCB is very easy to learn and has always worked well for me. I started with it a few years ago and have never bothered to try any of the others... I've heard that Altium is bloated and tries to do everything, which in turn makes it difficult to use for anything.

 

[wirednuts]

ok sounds good. ill try to get a design done tonite then, and ill post it here in case anyone wants to show me what i might have done wrong.

i doubt ill even do the whole circuit, all i need is a place to solder this chip too that has channels extending out to nice big solder pads i can use. after that its like 3 resistors, a couple of light diodes and a couple caps. if i find the design phase easy though, ill do the whole circuit. maybe even add a high/low switch too, who knows

(ive been pondering this now for quite some time. would it ever be beneficial to say charge my batts at 250ma compared to what will be the normal 800ma? i can only think maybe if i know the battery will be left overnight, i might as well charge it at a low current so i dont prematurely wear out the battery... but at the same time these batts cost a whopping $4 a piece so its not like i care much. i just know my phone can drain its load in about 1.5hrs if im streaming music over 3g, and thats simply not long enough for one of these 350ma chargers to finish charging. having a switch so i could select between 800ma and 250ma would be very easy to make according to the datasheet... i just dont know when i would realistically use it.)

 

[Howard]

Eagle and ExpressPCB would probably be the most popular out of that list.

 

[William Gaatjes]

If you are only going to make one pcb, this is also an option if you are willing to try :

I bought an experimental pcb specially made for soldering smd components.
For example :

PCB type : RE450-LF, ROTH ELEKTRONIK.
This can be ordered for example at the shop : www.farnell.com

I sawed the pcb to pieces for a very small part :
It is not visible but i cut the traces down the middle in order to have 2 * 5 traces.
I glued and soldered a small smt connector on it.
I soldered singe strands of wire on the pins of the connector.

I used a very fine tip. For reference, the drill bit is 1 mm.

A single strand of wire. For reference, the drill bit is 1 mm.

This is the pcb with connector glued to a normal 2.54mm experimental pcb (single square holes)
The wire strands are guided through the holes and soldered on the otherside to smt components.
The glue is to fixate the wire strands.

 

[William Gaatjes]

I do not have pictures with me, but what i usually do with the ic as you want to use in dfn package, is i glue them with the topside on the experimental pcb and i solder the same strands to the pins. This is easy because the bottomside is now exposed. The strands i solder to normal 2.54mm square pads. After that, i test if it works. If it works, i use a little superglue to fixate the strands to the pcb.

However, perhaps you can also find an experimental roth elektronik pcb with the dfn pad distance you seek.

http://uk.farnell.com/jsp/search/re...&isGoback=false&displaytext=&isRedirect=false

 

[wirednuts]

this is what i had originally imagined doing.

and im sure i probably still could... but once i found that website that claims to make custom pcb's for next to nothing, i then figured i might as well go that route. i am planning on making at least 5 or 6 of these, and even if i could solder tiny wires to it and get it all to work, making half a dozen of them would be a pain

thank you though, i think i still should get ahold of one of these blank generic pcb test boards, i would have to imagine the pin layout of the 1835 ic is somewhat standard

 

[Jeff7]

How well-acquainted are you with electronics in general, and making circuitry like this?

The "lithium ion" part caught my attention. I'm also not exactly what I'd consider to be an "expert" in some of this stuff, though that may be subjective.
However, I do know that lithium ion batteries can turn ugly if not properly charged, as in fire-or-small-explosion ugly.

 

[Modelworks]

A few things :
The package type for the part is DFN. That is a VERY tricky part to solder without the correct equipment. The problem comes from soldering the center pad. That pad has to have a good connection to a large enough area on the PCB to act as a heatsink for the part. If the contact is poor or the PCB layout is wrong the part will overheat and quit.
Make sure you understand page 13 of the datasheet.

I would choose a different chip that is both easier to solder and handle. Since this is going to charge phone batts makes sure there isn't a micro inside the battery that has to be dealt with too. Some batteries contain internal charge limiters that will not allow any charging without a proper string being sent over the connection pins.Others will disconnect if the charge rate is higher than the factory charger. Hopefully it is just a bunch of cells in a package without any circuitry in the battery.

Also it might be wise to test your first pcb with as small an order as possible so you do not end up with a bunch of boards with a bad route.

 

[wirednuts]

A few things :
The package type for the part is DFN. That is a VERY tricky part to solder without the correct equipment. The problem comes from soldering the center pad. That pad has to have a good connection to a large enough area on the PCB to act as a heatsink for the part. If the contact is poor or the PCB layout is wrong the part will overheat and quit.
Make sure you understand page 13 of the datasheet.

I would choose a different chip that is both easier to solder and handle. Since this is going to charge phone batts makes sure there isn't a micro inside the battery that has to be dealt with too. Some batteries contain internal charge limiters that will not allow any charging without a proper string being sent over the connection pins.Others will disconnect if the charge rate is higher than the factory charger. Hopefully it is just a bunch of cells in a package without any circuitry in the battery.

Also it might be wise to test your first pcb with as small an order as possible so you do not end up with a bunch of boards with a bad route.

thanks. i knew about the center pad, thats why i opted to have the pcb's built for me because i knew thats the only real chance i have at getting it to work.

the batteries have basic protection chips in them as far as i know. the external chargers just use the + and - pins though, as they work like any basic li-ion charger.

i can verify this by charging the batteries myself manually. if you just hook straight 5v usb power to the + and - pins, when the battery is not in the phone, the battery will charge, and when it gets to 4.2v the charging stops because the battery itself cut off voltage.

so i really dont even need a charger circuit if i just wanted something REALLY dirty, but using protection circuits as a means to charge is not safe.

 

[Modelworks]

The problem with DFN packages is you will need to use solder paste and hot air to do it right because you can't solder the pad other ways, it will bridge the pins if you try.

I would use the LTC1731 chip. It works well for what you want and it is an easier to use package type and requires just a small amount of external parts. You set the charge rate with just a resistor value.
http://www.linear.com/product/LTC1731-4.2

 

[wirednuts]

got a chip that can do 1000ma? the chargers i have now are already 350ma, and they take 3+ hours to charge. im looking to get that closer to an hour and a half or less.

i think the reason you dont see cell chargers over 500ma is because the smallest cell phone batteries are 1000mah, so to stay universal they keep the charging rates lower. im not afraid to charge these close to 1C rate though, it will be temperature protected plus i will of course monitor it myself regularly- at least until i know its not too hard on the cells.

also, that was what was so hard to find on mouser. all these li-ion charger ic's seem to be in DFN package. i just didnt realize how small that actually was.

if i have a custom made pcb with the right center pad and pins, wouldnt this be feasible to do? i understand i need to use solder paste and a heat gun, but that doesnt seem too hard

 

[Howard]

The problem with DFN packages is you will need to use solder paste and hot air to do it right because you can't solder the pad other ways, it will bridge the pins if you try.

I would use the LTC1731 chip. It works well for what you want and it is an easier to use package type and requires just a small amount of external parts. You set the charge rate with just a resistor value.
http://www.linear.com/product/LTC1731-4.2

What if you had a large via right in the middle? Then perhaps you could lay down a big blob on the underside of the chip, and heat it through from the back. With enough flux it would stick on its own.

 

[wirednuts]

What if you had a large via right in the middle? Then perhaps you could lay down a big blob on the underside of the chip, and heat it through from the back. With enough flux it would stick on its own.

yeah i think thats what the engineers who made the chip were thinking-

 

[Howard]

You will have trouble aligning the other pins though. How will you even solder them? DFN is SMT plain and simple.

http://electronics.stackexchange.com/questions/6983/is-hand-soldering-a-dfn-package-possible

 

[wirednuts]

maybe i should just buy these at 2 bucks a pop-

http://www.ti.com/lit/ds/symlink/bq2057c.pdf


or, what other package types should i look for? thats the only MOSP in stock at mouser

 

[Modelworks]

yeah i think thats what the engineers who made the chip were thinking-

It is really risky to do it without the proper setup. The pads are really close together and if you start pushing .5A + through the chip and the contact is poor then the chip will fail. Something others have done but I haven't tried myself is to drill through the board where the pad will be and solder it from the underside.

Charging at 1C isn't something I would recommend without a very good temperature monitoring setup. The risk of fire is much higher at those rates. The reason you don't see 1A chargers for phones is because it is too risky and it also destroys the cell.

Give this a read if you get time, it talks about charge rates and also a chip from ti that is a QFN-TEP package which makes the pad exposed from the top.

http://www.ti.com/lit/an/slyt269/slyt269.pdf

 

[wirednuts]

Charging at 1C isn't something I would recommend without a very good temperature monitoring setup. The risk of fire is much higher at those rates. The reason you don't see 1A chargers for phones is because it is too risky and it also destroys the cell.

in my experience its not that big of a deal. these batters and discharge at greater then 1c rate without issue, so its not hard to see why then can take a 1c charge. some phones do charge them at close to 1000ma too, especially when the phone is off and plugged into an a/c adapter. i think the reason you dont see them in phone chargers is because most phone chargers are generic and made to be universal.

 

[William Gaatjes]

this is what i had originally imagined doing.

and im sure i probably still could... but once i found that website that claims to make custom pcb's for next to nothing, i then figured i might as well go that route. i am planning on making at least 5 or 6 of these, and even if i could solder tiny wires to it and get it all to work, making half a dozen of them would be a pain

thank you though, i think i still should get ahold of one of these blank generic pcb test boards, i would have to imagine the pin layout of the 1835 ic is somewhat standard

That is ok. I will just post the pictures as a generic possibility for anyone interested.

I should note that the weakness of using this technique is that it is difficult to create a proper cooling solution. Also for rf purposes this is not really advised. But for general prototyping it is very usable. :

The other side of the pcb. Here i just solder wires and smd components. For 2.54mm square pads, 0603 components are ideal to place. Also SO-8, SO-14 and SO16 can be placed(with a little cutting of pads into 2 seperate islands for the pins. Sot23 packaged components fit nicely. Also components in packages : SO23-5 and SOT23-6 can be soldered easily with a little cutting of the pads.

 

[William Gaatjes]

The problem with DFN packages is you will need to use solder paste and hot air to do it right because you can't solder the pad other ways, it will bridge the pins if you try.

I would use the LTC1731 chip. It works well for what you want and it is an easier to use package type and requires just a small amount of external parts. You set the charge rate with just a resistor value.
http://www.linear.com/product/LTC1731-4.2

I agree. Linear makes nice chips. But the real advantage is these chips are in msop and SO-8 available thus easy to solder.

 

[Howard]

That is ok. I will just post the pictures as a generic possibility for anyone interested.

I should note that the weakness of using this technique is that it is difficult to create a proper cooling solution. Also for rf purposes this is not really advised. But for general prototyping it is very usable. :

I would imagine that a thin copper strip could be soldered on, narrowed to reduce the heat input required for fusion.

 

[wirednuts]

i figured it out! i had everything i needed.,... just some solder and hot glue

no seriously, i tied the a/c mains together on a cut-off extension cord, and hot-glued the cases together for now (ill epoxy this all up later). it works! took 20min and was free! (doh!)

i figured, why am i trying to cut charging time in half when i could just charge two batteries at once? i usually over think things and this time was another case. as you can see, they work independently, blue is charged, red is charging, purple is almost done! easier to plug into outlets now too. more durable as well. and i still get 2x800ma usb ports to charge any usb device with. total cost- ~$8 and 1/2 hour of life.

this thread is still very helpful. i likely will be making custom circuits in the future for something else, and i will look back to this thread for reference.

 

[William Gaatjes]

I would imagine that a thin copper strip could be soldered on, narrowed to reduce the heat input required for fusion.

?

I do not fully understand your sentence. with respect to fusion...
But i have used copper tape as a means to create a ground plane once on a dfn package. However, for cooling purposes it is not better then using etched copper traces for cooling.

http://uk.farnell.com/jsp/search/re...&isGoback=false&displaytext=&isRedirect=false

 

[William Gaatjes]

i figured it out! i had everything i needed.,... just some solder and hot glue

no seriously, i tied the a/c mains together on a cut-off extension cord, and hot-glued the cases together for now (ill epoxy this all up later). it works! took 20min and was free! (doh!)


i figured, why am i trying to cut charging time in half when i could just charge two batteries at once? i usually over think things and this time was another case. as you can see, they work independently, blue is charged, red is charging, purple is almost done! easier to plug into outlets now too. more durable as well. and i still get 2x800ma usb ports to charge any usb device with. total cost- ~$8 and 1/2 hour of life.

this thread is still very helpful. i likely will be making custom circuits in the future for something else, and i will look back to this thread for reference.

Good for you.
An old saying is : "There are many ways to Rome..."
^_^

 

[davejo]

It is really risky to do it without the proper setup. The pads are really close together and if you start pushing .5A + through the chip and the contact is poor then the chip will fail. Something others have done but I haven't tried myself is to drill through the board where the pad will be and solder it from the underside.

Charging at 1C isn't something I would recommend without a very good temperature monitoring setup. The risk of fire is much higher at those rates. The reason you don't see 1A chargers for phones is because it is too risky and it also destroys the cell.

Give this a read if you get time, it talks about charge rates and also a chip from ti that is a QFN-TEP package which makes the pad exposed from the top.

http://www.ti.com/lit/an/slyt269/slyt269.pdf

It's better to buy an experimental pcb for soldering smd components.

pcb assembly