PCB Creation Software

[papasan]

is there any free or relatively inexpensive PCB creation software that's fairly easy to use (i'm super novice at this)?...i've found a couple, but they either cost $500+ or they are way too simple...

basically, i just want to put some part/pin measurements in and jump em together and print it out and etch it...i don't need to test the circuit or throw meters, analyzers, etc. on it...preferably, something i can just put like "SOT23-5" in and it'll have the specs...

perhaps i've already found something like this (i've downloaded like 6 so far) and didn't realize it/know how to use it...perhaps i'm asking for too much (it's been known to happen =))...

so what do you guys use to make PCBs?...

 

[ElektroLumens]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by papasan:
is there any free or relatively inexpensive PCB creation software that's fairly easy to use (i'm super novice at this)?...i've found a couple, but they either cost $500+ or they are way too simple...

basically, i just want to put some part/pin measurements in and jump em together and print it out and etch it...i don't need to test the circuit or throw meters, analyzers, etc. on it...preferably, something i can just put like "SOT23-5" in and it'll have the specs...

perhaps i've already found something like this (i've downloaded like 6 so far) and didn't realize it/know how to use it...perhaps i'm asking for too much (it's been known to happen =))...

so what do you guys use to make PCBs?...<HR></BLOCKQUOTE>

papasan,


I use the Eagle PCB design software. It's free, and it works great. It has a lot of components to drop onto the design. Lot's of options.

I have tried a number of free software packages, and so far, for the free version, this is the best I've found. It has some limitations in the free version, like you are limited to one layer, but I don't need that anyway, for now.

Here's the web site:
http://www.cadsoft.de/

For etching, I have found that Liquid Paper works great. I have also heard of finger nail polish, but I have not tried it. I use 2oz. copper PCB board. It takes longer to etch, and etch resist ink doesn't quite do the job. Good luck with your designs and etching.

 

[papasan]

it's a good program...very good for free =)...i made this up pretty quick, it's a layout for the ZXSC3x0 regulator...less than an inch by an inch...could be made much smaller with a small inductor and cap, but with a sacrifice of efficiency...

 

[ElektroLumens]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by papasan:
it's a good program...very good for free =)...i made this up pretty quick, it's a layout for the ZXSC3x0 regulator...less than an inch by an inch...could be made much smaller with a small inductor and cap, but with a sacrifice of efficiency...

<HR></BLOCKQUOTE>

Hey papasan,
You're a fast learner! Nice job!

I like this program a lot. For a free program, it's really good. I think the free program has limited functionality, but it does what I need it to do. I haven't tried to make a board for the Zetex chip yet, but I plan to shortly (I have received the parts for this chip, but have not had the time to build it yet).

 

[papasan]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Wayne Johnson:
Here's the web site:
http://www.cadsoft.de/<HR></BLOCKQUOTE>

excellent link...seems like very close to what i had in mind and a lot more...now i just gotta figure out how to use it =)...thanks...

 

[**DONOTDELETE**]

Papasan,

That is very cool! It's really neat to see the circuit board taking shape.

At 19x19mm, you got size down to the footprint of the Luxeon Star without optics (20x20mm)!

Your design also appears flexible enough to accept a variety of capacitors, which allows the end user to choose the best combination of efficiency, size, and price.

 

[papasan]

yeah, that's one thing i had in mind, actually mounting the board to the bottom of the LS, maybe after adding some heatsinking...but for some projects this won't work because the LS board has to be trimmed down...small as we can make it the beter =)...i think it could be tightened up a little, but i just wanted to try something kinda quick to see what it would look like...

btw, since we're kinda on the subject, i did a littel research on transfer to a copper clad board...i came up with this page that explains how to use high-gloss heavy weight laser paper, probably much cheaper than the $1.50+ for the press and peel stuff...

http://www.fullnet.com/u/tomg/gooteepc.htm

after i try it out i'll post some results, but it sounds good...

and this link for soldering SMD parts in a toaster oven =)...

http://www.seattlerobotics.org/encoder/200006/oven_art.htm

just thought i would share =)...

 

[Mercator]

papasan...

It was interesting reading about how someone else discovered using glossy inkjst paper for the toner transfer method. I wish I had come across that link before I figured it out and tried it myself. I too had pretty good results once I found a paper that worked. I tried several different papers and finally settled on one that performed OK. I made several baords with that "regular" paper. Then decided to try the comerical paper. The results and ease of use were so much better than messing around with the glossy paper. Even the writer in your link talks about pinholes and other problems. With the commercial paper these are not a problem. Once the print is ironed on the copper the transfer backing sheet comes right off the copper and the image is perfect. Your link referred to problems in the larger copper areas. I do see some minor staining by the etchant on these areas, but no pinholes. After having used both the regular paper and the commercial products I think that the problems with the large copper fill areas is related to a gassing effect of the molten plastic that is trapped in the larger areas. Fine lines and solder pads are not effected.
You can go ahead and try the glossy regular paper if you chose, but take it from some one that has been there.... You'll be much happier with the product designed for this purpose. I use the Press 'N Peel Blue product. In addition to transferring the toner the release agent on the transfer paper also gets layed down on the copper making for a better resist covering. The commercial paper comes in 8.5 x 11 sheets. I normally cut those sheets in thirds and tape one section onto the leading edge of a regular sheet of bond paper. Then feed that through the printer of photo copier. So, the "$1.50" per sheet is streached out a long ways. You can also fill in the section with as many of the designs to be transfered depending on the size of your PCB.

Regarding your PCB lay out shown above. Looks nice. But.... That "Eagle" ain't gonna fly ;-) You had better check out the pin asignments of the components. I see several errors. Now perhaps this was just a quick experiment using the new software, I don't know. I'd hate to see you make such a board and not have it work.

Mercator

 

[**DONOTDELETE**]

It's hard to settle on the "correct" board size, because there are now four versions of the Luxeon Star, and three sizes:

<UL TYPE=SQUARE><LI>The 25x25mm heatsink attached to the Star/O and Star/C
<LI>The 20x20mm heatsink attached to the hexagonal Star
<LI>The 10x10mm Luxeon Emitter without a heatsink
[/list]

For flashlight purposes, it would be ideal to use the LS/O, and trim its heatsink to the collimator diameter, 21.5mm, although I don't know practical that would be. Too bad you can't just buy the Star and collimator separately, and then mate them...

List of proposed component sizes

M9711 12.5 x 10.2 x 5.2mm
ZXSC300 3 x 3 x 1.5mm
FMMT617 3 x 3 x 1.5mm
ZHCS2000 3 x 2.5 x 1.5mm
Rsense 6.4 x 3.2 x 0.6mm
Tantalum 7.3 x 4.3 x 2.8mm

You would almost have to use a circular board of diameter 21.5mm in order to keep intercomponent spacing wide enough to get a soldering tip in.

A smaller square or rectangular board might be easier to cut and more cost effective, but the high component density would require use of the toaster oven soldering technique.

 

[papasan]

thanks for sharing your experience with the toner transfer method...the guy did say that he looked long and hard to find the right paper...i love the ingenuity of it, that's what attracted me to it...i'll probably still try it out, but i'll probably get the press'n'peel stuff for the actual finished board when the time comes...any of these sounds ten times better than the UV solutions...

yes, i meant to put a warning in there for folks not to actually try the circuit, as there are two errors that i see...the power in the NPN should be post coil and the resistor pad is jumping ground to the output of the NPN and to Isense...

i didn't actually set up much beyond the package sizes and layouts, the numbers on the pins have no bearing on real life...still, not bad for a first try, i don't think...

the beautiful thing about doing these boards ourselves is that we can personalize it however we wish...some may like em long and skinny, some short and fat, some perfectly square...veriety is the spice of life =)...

the footprint of the tantalum cap, if i got my package measurements right, is actually bigger (by a good deal) than the electrolytic...plus the electrolytic can stay off the board allowing you to put components 'underneath' it...just a thought...

and now my second layout attempt...the board is 14mm accross and is designed to just pop into the top of a 2AA mini-mag light (after removing the reflector)...uses the zetex 3x0, the 617 transistor, 0805 package resistor, and a 1210 delevan coil...based on the 'longest battery life' zetex solution, but a small enough resistor should still overdrive the 3 LEDs sufficiently =)...this may be the first actualy board i build since it's easier done and only has 2 variables, R and coil...

 

[Mercator]

papasan.....

I'm sure you'll find that the PNP Blue transfer paper is far better than anything else you might try. If you have some inkjet paper, then by all means give it a try just to see how the method works. With the right paper and some luck, you'll end up with a useable board. With either the commercial product or regular paper the trick is getting the temp. and ironing times worked out. Start with the iron temp at aprox. 275F and vary the time on several samples. It's hard to tell on the PNP paper when you have it worked out correctly. With regular gloosy paper you can pretty much tell after removing the paper. ON the PNP stuff you might have to etch your test samples to see a difference. It seems that more time is better than too little even though the image on th copper looks good. About all I can say is that it does take some playing around with. Far less trial and error with the commercial product. I wouldn't suggest that you strickly follow the ironing techniques layed out in that link you provided, it's not all that critical. Just make sure your iron has a smooth even surface and you heat the board evenly for the proper time.

Regarding you newest PCB design for three LED's and the simple Z300 ckt. I still see some errors in your pin assignemnts. You might want to take another look at the Zetex datasheets. Here is a simple parts drawing I created for the Zetex topic awhile back. I think it's on page 2 or 3.

I had, like you tried several PCB programs and found most of them way too complicated and that they all had just too many "bells & whistles". I did try the Eagle program, but for some reason passed it up. I've downd loaded it again and will play with it some more. What I'm curious to see is just the copper area of the finished design. Also, if it is easy to modify the image. In particular some of the solder pads for the SMD's. Making them alittle larger than what the standard programs gives, is a great help in hand soldering the parts onto you finished board. The regular solder pads created in these programs assumes that you will be using some type of wave or re-flow soldering process.

About the frying pan or toaster oven method of soldering SMD's. It requires the use of special solder paste which is rather costly. Also, I'd think that it would require one heck of a lot of trial and error before you got everything worked out. It certainly is worth a try though.

Good luck with your efforts,

Mercator

 

[papasan]

merc,

thanks very much for the drawing, i see the error of my ways...D'OH!...i gues i just took the schematics (which have the transistor's pins attached as i did) to be too much based on reality...

and again, i assumed the pin layout for the 3x0 on the block diagram was the same for the actual package...oops...what would happen with the STDN hooked up to ground?...would the IC stay in operation always?...

here's an update...

question...how would i figure out the Isense R value to use?...say i want to push the LEDs to .12A @ 3.6V (40mA/LED) from a 2 cell power source (say 2.4V) with the 3x0...do the transistor and coil figrue into the equation?...thanks..

 

[**DONOTDELETE**]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by papasan:
question...how would i figure out the Isense R value to use?...say i want to push the LEDs to .12A @ 3.6V (40mA/LED) from a 2 cell power source (say 2.4V) with the 3x0...do the transistor and coil figrue into the equation?...thanks..

It's not easy to come up with the right value for Rsense, because a lot has to do with the efficiency of the components you're using.

One thing you might consider is putting the LEDs in series, not parallel. That way, all the LEDs get the same current. Such would likely not be the case when connected in parallel, because some LEDs (even from the same lot) draw more current than others.

In series, output current is just 40mA, so I^2R losses are 1/9th what they would be at 120mA, and you might be able to find some tinier inductors.

But keep in mind that output voltage is now at least 10-12 volts, so be sure to use capacitors and transistors that can handle that.<HR></BLOCKQUOTE>

 

[**DONOTDELETE**]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by papasan:
The footprint of the tantalum cap, if i got my package measurements right, is actually bigger (by a good deal) than the electrolytic...plus the electrolytic can stay off the board allowing you to put components 'underneath' it...just a thought...<HR></BLOCKQUOTE>

Beware of electrolytics with such small footprints, as they often have very high ESR, and ESR can waste a lot of battery power.

Unfortunately, electrolytics are usually much larger than tantalums of the same ESR, and such electrolytics often tower over the other components. In applications where height is not a concern, electrolytics are much less expensive and usually easier to obtain.

 

[Mercator]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by papasan:
merc,
oops...what would happen with the STDN hooked up to ground?...would the IC stay in operation always?...
<HR></BLOCKQUOTE>

papasan,

On the X300 IC there is no physical connection within the chip for the "STDN" pin. I went ahead and soldered the ground and STDN together on the same solder pad on the proto-type board we are using. Was much easier ;-) This wouldn't be the case with the Z310 or, if and when we ever see it, the Z320. BTW - I think that the Z310's are back ordered on Digi-Key, so you might have to settle for the Z300's

Figuring R-Sense values....
If you look at the datasheet for the Z310 you'll find a chart that lists various components for different voltages and number of LED's. This might help answer your questions. I have the Zetex eval. board for the Z310. It has four LED's and I think they are running at 20mA each. There is a jumper header on the board that either uses the STDN or not. From what I gather using that pin to turn off the ckt, still draws some power. Not much, but again, not something you'd want in a flashlight application.

Mercator

 

[**DONOTDELETE**]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by papasan:
What would happen with the STDN hooked up to ground?...would the IC stay in operation always?...<HR></BLOCKQUOTE>

For the ZXSC300, it doesn't matter.

For the ZXSC310, tying STDN to ground will disable the IC. Tying STDN to Vcc will enable the IC until Vcc falls to 0.1 volts. Leaving STDN not connected will permanently enable the 310.

Frankly I don't see much value in the STDN feature, so if I got a 310, I would simply snip off that pin.

 

[Mercator]

Copied from tha Zetex 310 dataSheet....
Shutdown Control
The ZXSC310 offers a shutdown mode that produces a standby current of less than 5uA when in operation. When the voltage at the STDN pin is 0.7V or higher the ZXSC310 is enabled, hence the driver is in normal operation. When the voltage at the STDN pin is 0.1V or lower the ZXSC310 is disabled, hence the driver is in shutdown mode. If the STDN pin is open circuit the ZXSC310 is also enabled.

I just put my meter in series with the two AA cells on my eval board. and placed the STDN pin shorted to ground (dis-able) Zetex is right about less than 5uA. I read 4.7uA. Not much current by any means, but still a current draw.

The advantage of using the STDN feature shorted to Vcc and having it shut down the ckt at 1 volt, would be to prevent the ckt from causing your re-chargeable batteries to go into a deep discharged state. Not a good thing to happen. Also, with regular alk. batts it would prevent the risk of having those batts start leaking. So... There is a reason for using the STDN feature. One of my LS/Z300 applications is based around the Li-Ion 3.6v rechargeables. I most likely will consider a Z310 for that use.

Mercator