Homebrew PCBs

Any suggestions for starting up a little hobby setup for making PCBs?
I was thinking of getting the MG Chemicals 416-K kit. Comes with 3
sensitized copper-clad (single-sided boards), developer, ferric
chloride, foam brushes, rubber gloves. Any reason to not use ferric
chloride, and maybe sodium persulfate instead? I'm thinking of
something economical to make little boards for PIC microcontrollers
and some peripheral circuitry.

And for cheap DIY stuff, what is the smallest package types that are
feasible? Can you do smaller than SOIC? like SSOP or TSSOP?

If you don't know answers but know websites, those are appreciated as
well.

Thanks,
David

On Wed, 02 Feb 2005 07:59:54 -0000, dgrant_79 <david.grant@...>  
wrote:

>
> Any suggestions for starting up a little hobby setup for making PCBs?
> I was thinking of getting the MG Chemicals 416-K kit. Comes with 3
> sensitized copper-clad (single-sided boards), developer, ferric
> chloride, foam brushes, rubber gloves. Any reason to not use ferric
> chloride, and maybe sodium persulfate instead? I'm thinking of
> something economical to make little boards for PIC microcontrollers
> and some peripheral circuitry.

I would not advise buying a kit, because IMO none of the kits has _any_  
components you will use later  (when you get more experience). I still  
have the ferric chloride and some photoboard around which i won't use  
anymore.

Best is to inform yourself as much as possible. Investigate all methods of  
etching (chemicals, tanks/tools), also investigate all methods of resist  
transfer. You might decide you do not want to use what comes with the kit.

I do believe CuCl and toner transfer is possible to start with, even with  
zero experience.

> And for cheap DIY stuff, what is the smallest package types that are
> feasible? Can you do smaller than SOIC? like SSOP or TSSOP?
> If you don't know answers but know websites, those are appreciated as
> well.
> Thanks,
> David

The soldering is the limit, not the board, usually.
If you want to use toner transfer instead of the more complicated  
photoprocess you might be limited to 8mil tracks or so.

You can find loads of good links in the links section of the group.

ST

CuCl????

Stefan Trethan <stefan_trethan@...> wrote:
I do believe CuCl and toner transfer is possible to start with, even with 
zero experience.



		
---------------------------------
Do you Yahoo!?
 Yahoo! Search presents - Jib Jab's 'Second Term'

[Non-text portions of this message have been removed]

--- In Homebrew_PCBs@yahoogroups.com, guja <guja2001bg@y...> wrote:
> 
> CuCl????
> 
> Stefan Trethan <stefan_trethan@g...> wrote:
> I do believe CuCl and toner transfer is possible to start with, even
with 
> zero experience.

CuCl is Cupric Chloride etchant. It is renewable so you don't have to
keep tossing out etchant and buying more.

As Stefan said:

> You can find loads of good links in the links section of the group.

There are links to info on CuCl etchant there, and I archived a few
messages on making CuCl etchant in the Files section.

Steve Greenfield

On Wed, 2 Feb 2005 09:01:03 -0800 (PST), guja <guja2001bg@...> wrote:

>
> CuCl????


copper/cupric cloride, made using HCl, H2O2 and the PCB copper. see links  
section, chemical etching, CuCl.
You do not need additional copper (or CuCl of some form) to start with if  
you use H2O2.

Advantages: cheap, can be regenerated infinitely.

ST

--- In Homebrew_PCBs@yahoogroups.com, "Stefan Trethan" 
> 
> The soldering is the limit, not the board, usually.
> If you want to use toner transfer instead of the more complicated  
> photoprocess you might be limited to 8mil tracks or so.
> 

Well, 8mil sounds like plenty! Plenty as in, very small.

I think CuCl for a novice is too much.  It requires more fussing,
effort and chemistry skill to get going.  Its not impossible but AP or
FeCl are much easier to start with.  Given all the other things they
are learning, anything that makes the first board simpler to produce
is a good thing.

I've used all three and recommend AP for the beginner as it is
non-staining and you can see the progress of the board.  Both CuCl and
FeCl are pretty opaque and cause stains.  

After the novice has gotten TT or photomask techniques figured out,
they should then consider CuCl.  Given that it has a positive
implication for the environment, its worth moving to.

Phil


--- In Homebrew_PCBs@yahoogroups.com, "Stefan Trethan"
<stefan_trethan@g...> wrote:
> On Wed, 2 Feb 2005 09:01:03 -0800 (PST), guja <guja2001bg@y...> wrote:
> 
> >
> > CuCl????
> 
> 
> copper/cupric cloride, made using HCl, H2O2 and the PCB copper. see
links  
> section, chemical etching, CuCl.
> You do not need additional copper (or CuCl of some form) to start
with if  

> you use H2O2.
> 
> Advantages: cheap, can be regenerated infinitely.
> 
> ST

On Thu, 03 Feb 2005 05:58:34 -0000, Phil <phil1960us@...> wrote:

>
> I think CuCl for a novice is too much.  It requires more fussing,
> effort and chemistry skill to get going.  Its not impossible but AP or
> FeCl are much easier to start with.  Given all the other things they
> are learning, anything that makes the first board simpler to produce
> is a good thing.
> I've used all three and recommend AP for the beginner as it is
> non-staining and you can see the progress of the board.  Both CuCl and
> FeCl are pretty opaque and cause stains.
> After the novice has gotten TT or photomask techniques figured out,
> they should then consider CuCl.  Given that it has a positive
> implication for the environment, its worth moving to.
> Phil

Well, i do assume everyone who should be trusted with any chemical can  
dilute one with water and add a second one. The CuCl recepies in the links  
section are a bit complicated at times, because they use some form of  
copper to get started, while it isn't required to do that if you use the  
copper from your boards and H2O2, over a longer period of time.

AP must be mixed with water, right? that involves the same steps as  
diluting HCl. I agree adding H2O2 adds a certain amount of complexity, but  
it doesn't seem that much more complicated than pouring a glass of orange  
juice.

The whole analysis stuff is nice, but you don't really need it, i used  
CuCl for years without, just going by color and etching speed. Of course,  
knowing the molarity takes away someting you need to guess otherwise, so  
i'd recommend it for beginners, but it only involves using 2 eyedroppers.

Sure AP can be used, no problems there, but i don't see much advantage.

ST

Arguments about chemistry aside, I'd like to address the original 
question (novel approach, I know).
Mention was made of buying the MG kits.  I teach high school and 
middle school students part time, and currently they are building a 
robot.  We are about to make some PC boards in class, and needing a 
quick way to put together a small setup for the class I bought the MG 
stuff.  This afternoon I started to set things up, and I was NOT 
impressed.

I bought the exposure kit, expecting to get a UV lamp - I got a $10 
under-cabinet white fluorescent  fixture instead.  The "tanks" were 
just plastic kitchen containers.  Nothing in any of the kits was 
"purpose made".  A trip to Wal-Mart or Target would provide 
everything in the kits other than boards and chemistry. 
Incidentally, the included boards are all single-sided.

So, I suggest buying the items on your own - unless, like me, you 
just need a quick and easy solution for a class.  Even so, if i had 
it to do all over again, I'd probably go the ad hoc route.

Chemistry and boards can be had at All Electronics, Digi-Key, Mouser, 
or your local Fry's.

YMMV

People can do what ever they feel comfortable doing but CuCl can be
touchy stuff.  I think people starting out should know this.  I recall
a lot of people having trouble with slow etching and such.    Conc
H2O2 isn't super common (and definitely NOT like pouring orange juice
- I saw a guy get some on his skin, nasty burn).  Pouring/diluting
acid needs to be done with care - more so than AP which is pretty
benign.  CuCl is a bunch of extra steps and precautions that aren't
necessary.  I use CuCl - I'm not saying its bad.

In my opinion, AP is really the way to go for a first-timer.  It
doesn't burn on contact and doesn't stain like FeCl.  Very easy to use
and clear so you can see the board.  I think it really helps the
novice to be able to see the progress of the etch with out lifting the
board out.  But, hey, the beauty of information is that you can do
with it what you please!

Phil

--- In Homebrew_PCBs@yahoogroups.com, "Stefan Trethan"
<stefan_trethan@g...> wrote:
> On Thu, 03 Feb 2005 05:58:34 -0000, Phil <phil1960us@y...> wrote:
> 
> >
> > I think CuCl for a novice is too much.  It requires more fussing,
> > effort and chemistry skill to get going.  Its not impossible but AP or
> > FeCl are much easier to start with.  Given all the other things they
> > are learning, anything that makes the first board simpler to produce
> > is a good thing.
> > I've used all three and recommend AP for the beginner as it is
> > non-staining and you can see the progress of the board.  Both CuCl and
> > FeCl are pretty opaque and cause stains.
> > After the novice has gotten TT or photomask techniques figured out,
> > they should then consider CuCl.  Given that it has a positive
> > implication for the environment, its worth moving to.
> > Phil
> 
> Well, i do assume everyone who should be trusted with any chemical can  
> dilute one with water and add a second one. The CuCl recepies in the
links  
> section are a bit complicated at times, because they use some form of  
> copper to get started, while it isn't required to do that if you use
the  
> copper from your boards and H2O2, over a longer period of time.
> 
> AP must be mixed with water, right? that involves the same steps as  
> diluting HCl. I agree adding H2O2 adds a certain amount of
complexity, but  
> it doesn't seem that much more complicated than pouring a glass of
orange  
> juice.
> 
> The whole analysis stuff is nice, but you don't really need it, i used  
> CuCl for years without, just going by color and etching speed. Of
course,  
> knowing the molarity takes away someting you need to guess
otherwise, so  
> i'd recommend it for beginners, but it only involves using 2
eyedroppers.

> 
> Sure AP can be used, no problems there, but i don't see much advantage.
> 
> ST

Hi just wanted to put in my 2 cents worth on this subject.
I use a version of CUCL that is cheap and easy to get the stuff for. I
use regular 3% Peroxide available at any drug store. Cheap in larger
sizes ( I actually get it at Costco, 1.78 for 2 qts. ) I then use pool
acid (muratic acid, available at pool supplies and most grocery
stores) about 3-4 bucks a gallon. Then get a clear plastic box made
from the soft milky colored polyrthelene usually sold as shoe boxes.
Now pour 2/3 cup of peroxide in container. I recommend doing this
outside as a few fumes may be given off when mixing in the acid. After
the peroxide is in add 1/3 cup pool acid. Be carefull pouring this
into the peroxide. It may fume slightly. I then take it into the
kitchen sink. It is clear at this time. Place the board into the
solution. At first it seems nothing is happening. In less than a
minute the solution starts turning a very pretty bright green thas is
still transparent. In about 3-5 minutes the board is completely
etched. I can do at least a 6x6 board is just one cup total etchant
doublesided. Rinse the board good. The downside is the solution will
go bad within a couple of hours. You can regen with concentrated
peroxide (tried it didn't like it) so I just make fresh each time.
When you get done pour it down the drain with running water. This is
appropriate for occasional use. If I was doing large quanities I would
go full blown CUCL. As far as temps go I made it one time with my room
temp peroxide and some 26 degree pool acid (it was cold out) and it
worked just as well. Use plastic tongs for the board. And last but not
least I just rock the container back and forth for the etching time.
With the small amount of etchant you are unlikely to slop any out. BTW
I have a 15/16 full gallon of new FECL sitting on the shelf gathering
dust. I'll never use it, this is just to easy.



--- In Homebrew_PCBs@yahoogroups.com, "Phil" <phil1960us@y...> wrote:

> 
> People can do what ever they feel comfortable doing but CuCl can be
> touchy stuff.  I think people starting out should know this.  I recall
> a lot of people having trouble with slow etching and such.    Conc
> H2O2 isn't super common (and definitely NOT like pouring orange juice
> - I saw a guy get some on his skin, nasty burn).  Pouring/diluting
> acid needs to be done with care - more so than AP which is pretty
> benign.  CuCl is a bunch of extra steps and precautions that aren't
> necessary.  I use CuCl - I'm not saying its bad.
> 
> In my opinion, AP is really the way to go for a first-timer.  It
> doesn't burn on contact and doesn't stain like FeCl.  Very easy to use
> and clear so you can see the board.  I think it really helps the
> novice to be able to see the progress of the etch with out lifting the
> board out.  But, hey, the beauty of information is that you can do
> with it what you please!
> 
> Phil
>

=========ORIGINAL MESSAGE:

Date: Thu, 03 Feb 2005 02:19:23 -0000

   From: "dgrant_79" <david.grant@...>
Subject: Re: Home PCB kit


--- In Homebrew_PCBs@yahoogroups.com, "Stefan Trethan" 
> > 
> > The soldering is the limit, not the board, usually.
> > If you want to use toner transfer instead of the more complicated  
> > photoprocess you might be limited to 8mil tracks or so.
> > 

> Well, 8mil sounds like plenty! Plenty as in, very small.

Hmmm...  Well...  OK.   But I routinely use 6.67/1000ths-inch (2/300ths-inch) tracks and spacing, with toner transfer, with very close to a 100% success rate. I have also routinely used 6.67/1000ths-inch tracks (2/300ths-inch) with 3.33/1000ths-inch (1/300th-inch) spacing, with no problems (with an HP LJ 4 @ 600DPI, in both cases).

And, in very-limited tests, I have even had almost 100% success with 1.67/1000ths-inch tracks (i.e. 1 dot wide printed with an HP LaserJet 4 at 600 dots per inch (i.e. 1/600th-inch wide, on screen), with HP toner on Staples' "Picture Paper", as before) (but unfortunately didn't test the same size-range for the spacing limit).

Tom Gootee

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

----------------------------------------------------


[Non-text portions of this message have been removed]

On Fri, 4 Feb 2005 03:28:39 -0500, Thomas P. Gootee <tomg@...>  
wrote:

>
> Hmmm...  Well...  OK.   But I routinely use 6.67/1000ths-inch  
> (2/300ths-inch) tracks and spacing, with toner
> transfer, with very close to a 100% success rate. I have also routinely  
> used 6.67/1000ths-inch tracks (2/300ths-
> inch) with 3.33/1000ths-inch (1/300th-inch) spacing, with no problems  
> (with an HP LJ 4 @ 600DPI, in both cases).
> And, in very-limited tests, I have even had almost 100% success with  
> 1.67/1000ths-inch tracks (i.e. 1 dot wide
> printed with an HP LaserJet 4 at 600 dots per inch (i.e. 1/600th-inch  
> wide, on screen), with HP toner on
> Staples' "Picture Paper", as before) (but unfortunately didn't test the  
> same size-range for the spacing limit).
> Tom Gootee


Good to hear there is still some room in it. My HPIIID just isn't up to it.

Don't you have problems with toner spreading? if i make a 10mil track and  
10mil space they turn out not exactly the same width.

ST

I archived an earlier post on this mix by dkesterline in the Files
section under Acid Etching, same proportions. Cheap and easy.

I also have lots of FeCl, several pounds of it dry. What a waste, even
heated it etches only very slowly. Stains everything.

I'm in Tacoma, someone wants to come pick up this FeCl you can have it.

Steve Greenfield

--- In Homebrew_PCBs@yahoogroups.com, "klmjr22" <keithlmartin@c...> wrote:

> 
> Hi just wanted to put in my 2 cents worth on this subject.
> I use a version of CUCL that is cheap and easy to get the stuff for. I
> use regular 3% Peroxide available at any drug store. Cheap in larger
> sizes ( I actually get it at Costco, 1.78 for 2 qts. ) I then use pool
> acid (muratic acid, available at pool supplies and most grocery
> stores) about 3-4 bucks a gallon. Then get a clear plastic box made
> from the soft milky colored polyrthelene usually sold as shoe boxes.
> Now pour 2/3 cup of peroxide in container. I recommend doing this
> outside as a few fumes may be given off when mixing in the acid. After
> the peroxide is in add 1/3 cup pool acid. Be carefull pouring this
> into the peroxide. It may fume slightly. I then take it into the
> kitchen sink. It is clear at this time. Place the board into the
> solution. At first it seems nothing is happening. In less than a
> minute the solution starts turning a very pretty bright green thas is
> still transparent. In about 3-5 minutes the board is completely
> etched. I can do at least a 6x6 board is just one cup total etchant
> doublesided. Rinse the board good. The downside is the solution will
> go bad within a couple of hours. You can regen with concentrated
> peroxide (tried it didn't like it) so I just make fresh each time.
> When you get done pour it down the drain with running water. This is
> appropriate for occasional use. If I was doing large quanities I would
> go full blown CUCL. As far as temps go I made it one time with my room
> temp peroxide and some 26 degree pool acid (it was cold out) and it
> worked just as well. Use plastic tongs for the board. And last but not
> least I just rock the container back and forth for the etching time.
> With the small amount of etchant you are unlikely to slop any out. BTW
> I have a 15/16 full gallon of new FECL sitting on the shelf gathering
> dust. I'll never use it, this is just to easy.
> 
>

I'll take it!

I can pick it up this weekend.  Give me a time and place.

--- In Homebrew_PCBs@yahoogroups.com, "Steve" <alienrelics@y...> 
wrote:
> 
> I archived an earlier post on this mix by dkesterline in the Files
> section under Acid Etching, same proportions. Cheap and easy.
> 
> I also have lots of FeCl, several pounds of it dry. What a waste, 
even
> heated it etches only very slowly. Stains everything.
> 
> I'm in Tacoma, someone wants to come pick up this FeCl you can have 
it.

> 
> Steve Greenfield
> 
>

-------- ORIGINAL MESSAGES:   

Date: Fri, 04 Feb 2005 09:46:09 +0100

   From: "Stefan Trethan" <stefan_trethan@...>
Subject: Re: Re: Home PCB kit

On Fri, 4 Feb 2005 03:28:39 -0500, Thomas P. Gootee <tomg@...>  
wrote:

>
> Hmmm...  Well...  OK.   But I routinely use 6.67/1000ths-inch  
> (2/300ths-inch) tracks and spacing, with toner
> transfer, with very close to a 100% success rate. I have also routinely  
> used 6.67/1000ths-inch tracks (2/300ths-
> inch) with 3.33/1000ths-inch (1/300th-inch) spacing, with no problems  
> (with an HP LJ 4 @ 600DPI, in both cases).
> And, in very-limited tests, I have even had almost 100% success with  
> 1.67/1000ths-inch tracks (i.e. 1 dot wide
> printed with an HP LaserJet 4 at 600 dots per inch (i.e. 1/600th-inch  
> wide, on screen), with HP toner on
> Staples' "Picture Paper", as before) (but unfortunately didn't test the  
> same size-range for the spacing limit).
> Tom Gootee


- Good to hear there is still some room in it. My HPIIID just isn't up to it.

- Don't you have problems with toner spreading? if i make a 10mil track and  
- 10mil space they turn out not exactly the same width.

- ST

----------Current RESPONSE:

Stephan,

Yes.  It does look like there's still room for better toner transfer performance, i.e. smaller and closer traces and pads.

But yes, I DO have problems with toner spreading.  But I usually EITHER use large-enough spacings for trace separation (2/300ths-inch seems like about the minimum, to have basically no worries about developing a short between two 2/300ths-inch-wide traces), OR only use 1/300th-inch spacing where it won't really matter too much if a short develops, such as when I have several parallel traces that all return to a common ground point, which might have small (1/300th-inch) spacing when they are getting close to their common connection point.  (Gee, I almost hope that someone asks me WHY I would want to run multiple ground traces, in parallel, when they're all going to the exact same place, anyway; heh heh.)

I remember trying some surface-mount types/sizes of patterns, once, just as a test, and having some very close-together pads that I tried to run traces between, which really had problems from toner-spreading.  I think I was using pads that were 7x14 dots, at 600 dpi, with 7 and 8 dots spacing tried, between some of them, and tried putting one and two traces between them (e.g. two one-dot-wide lines at dots 3 and 6 of an 8-dot space).  I guess a solution might be to allow some space for the toner spreading, i.e. draw the pads narrower.

-Tom Gootee

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

----------------------------------------------------

[Non-text portions of this message have been removed]

On Fri, 4 Feb 2005 20:45:56 -0500, Thomas P. Gootee <tomg@...>  
wrote:

>
> Stephan,
> Yes.  It does look like there's still room for better toner transfer  
> performance, i.e. smaller and closer traces and pads.
> But yes, I DO have problems with toner spreading.  But I usually EITHER  
> use large-enough spacings for trace separation (2/
> 300ths-inch seems like about the minimum, to have basically no worries  
> about developing a short between two 2/300ths-inch-
> wide traces),

that's 6.66mil

> OR only use 1/300th-inch spacing where it won't really matter too much  
> if a short develops, such as when I
> have several parallel traces that all return to a common ground point,  
> which might have small (1/300th-inch) spacing when
> they are getting close to their common connection point.

Well, i don't see the logic for that, either i have spacing or i have a  
single track, but "dunno" seems of not much advantage to me.

> (Gee, I almost hope that someone asks me WHY I would want to run
> multiple ground traces, in parallel, when they're all going to the exact  
> same place, anyway; heh heh.)

No, nobody wants to ask that as everybody knows it, but some might ask  
what good it is if you are unsure if they are actually separate or not.

> I remember trying some surface-mount types/sizes of patterns, once, just  
> as a test, and having some very close-together
> pads that I tried to run traces between, which really had problems from  
> toner-spreading.  I think I was using pads that
> were 7x14 dots, at 600 dpi, with 7 and 8 dots spacing tried, between  
> some of them, and tried putting one and two traces
> between them (e.g. two one-dot-wide lines at dots 3 and 6 of an 8-dot  
> space).  I guess a solution might be to allow some
> space for the toner spreading, i.e. draw the pads narrower.

Yes, narrower pads help, but you didn't report definitive success here.

So, in the end you are doing 6.66mil traces with 6.66mil spacing reliably,  
just like everybody else.
I bet your traces are closer to 8 mil and your spacing closer to 4mil in  
the end.

Seems to me like this is a rather hard limit of TT.
I bought a 1200DPI printer only to find out it isn't real 1200 but 600  
instead. Well, i hope i can at least get a bit better than the 300dpi  
results.

There is a point where the process (transfer, spreading,...) is the limit  
and not so much the printer resolution. I had hoped i can still gain much  
by using a new printer, but i'm not so sure now...

The components aren't going to get bigger in the future.....

ST

--- In Homebrew_PCBs@yahoogroups.com, "Thomas P. Gootee" <tomg@f...>
wrote:
...
> But yes, I DO have problems with toner spreading.  But I usually
EITHER use large-enough spacings for trace separation (2/300ths-inch
seems like about the minimum, to have basically no worries about
developing a short between two 2/300ths-inch-wide traces), OR only use
1/300th-inch spacing where it won't really matter too much if a short
develops, such as when I have several parallel traces that all return
to a common ground point, which might have small (1/300th-inch)
spacing when they are getting close to their common connection point.
 (Gee, I almost hope that someone asks me WHY I would want to run
multiple ground traces, in parallel, when they're all going to the
exact same place, anyway; heh heh.)


I had toner spreading (AKA trace bloom) problems when using heavily
coated papers.   Using less coated paper such as ink jet paper seemed
to solve the problem for me (or at least, control it).  It appears
that the excess toner is being absorbed by the paper rather than being
squeezed between the clay coating and the copper.  Also, controlling
the amount of toner helps somewhat with the coated papers - I use a
copier so I can tweak the toner level to what ever I want.

On Sat, 05 Feb 2005 21:17:41 -0000, Phil <phil1960us@...> wrote:

>
> I had toner spreading (AKA trace bloom) problems when using heavily
> coated papers.   Using less coated paper such as ink jet paper seemed
> to solve the problem for me (or at least, control it).  It appears
> that the excess toner is being absorbed by the paper rather than being
> squeezed between the clay coating and the copper.  Also, controlling
> the amount of toner helps somewhat with the coated papers - I use a
> copier so I can tweak the toner level to what ever I want.

The bad thing is if i reduce the toner density i can't make good ground  
planes any more.
The HPIIID has a nice analog pot to set print density, i hope the lexmark  
printer has at least some software setting, it certainly doesn't have any  
analog pots.

You are right with the paper, glossy is bad, matt is good, the toner seems  
to fill the gaps rather than spreading. Of course fusing pressure and no.  
of passes is a big influence too.

ST