Homebrew PCBs

Hello,

   Since encountering this group a month or so ago, I have refined my 
toner transfer process to the point that I am able to reliably do 15-
mil traces.  Maybe smaller.  Hope so.

   I'm doing a home project and have settled on PIC24F 
microcontrollers.  The latest version needs USB, and all the PICs 
that support USB are surface mount.  So I have been dragged kicking 
and screaming into the 21st century.

   I bought an Aoyue 968 hot-air SMD rework station and some fine 
tweezers and have been  having a good time scrounging parts off some 
defunct computer boards - just to get used to the tool.  I'm going to 
need a microscope - half of these parts just look like bits of dirt!

   One part I scrounged is the same package as the 24FJ256GB106 
processor I'm planning to use - a 64-pin TQFP.  Good LORD those pins 
are small.  Can Toner Transfer make boards this tight? According to 
the spec sheet, the pin pitch is .5mm, or about 20 thousandths of an 
inch.  So the traces need to be 10 thousandths of an inch, and will 
only have 10 thousandths of separation.  Ouch.

                               - Jerry Kaidor

On Fri, 20 Feb 2009 20:27:33 -0000, you wrote:

>Hello,
>
>   Since encountering this group a month or so ago, I have refined my 
>toner transfer process to the point that I am able to reliably do 15-
>mil traces.  Maybe smaller.  Hope so.

I'm roughly at 10 mils, 12 is run of the mil.....


>
>   I'm doing a home project and have settled on PIC24F 
>microcontrollers.  The latest version needs USB, and all the PICs 
>that support USB are surface mount.  So I have been dragged kicking 
>and screaming into the 21st century.

Ain't Technology wonderful?

>
>   I bought an Aoyue 968 hot-air SMD rework station and some fine 
>tweezers and have been  having a good time scrounging parts off some 
>defunct computer boards - just to get used to the tool.  I'm going to 
>need a microscope - half of these parts just look like bits of dirt!

Yeah, now try to measure them.  Seems to be the 50 dollar solution and
the 350 dollar solution, other than the 10 dollar ebay tweezers
(only).

I prefer 1206 or 805 at the smallest, just because of ease of using.
You're probably getting 4xx or so.

>
>   One part I scrounged is the same package as the 24FJ256GB106 
>processor I'm planning to use - a 64-pin TQFP.  Good LORD those pins 
>are small.  Can Toner Transfer make boards this tight? According to 
>the spec sheet, the pin pitch is .5mm, or about 20 thousandths of an 
>inch.  So the traces need to be 10 thousandths of an inch, and will 
>only have 10 thousandths of separation.  Ouch.

In theory, yes.  You might find yourself using 8 mm traces, or 10 mm
traces and somewhat larger pins.  I inadvertently made a board with
0.5 mm spacing pins on one 4 pin connector, and it turned out semi
"OK".  The whole board was a bit over etched, so the pins did not look
as nice as I would want, but otherwise ok, I'd suppose.

You'll either want a small tip (almost a needle) soldering pencil for
the leads, or you'll want to tin-lead coat the pads before soldering.
Very find gauge solder is also a must.  I find that upon occasion, a
larger soldering iron to solder multiple pins at once, and a
desoldering tool (built in air pump kind, not spring release) cleans
up things nicely.  NOT elegant, I know.

If you had the right stuff, then a syringe of solder paste and your
hot air pencil would be the optimum method.  (Right stuff defined as
right tools and paste, etc...)

Yeah, possible to do, messy, and then there's the matter of vias...

An alternate option is to use a carrier, such as the schmartboard
carriers, and go from there.  Chip will be on a carrier that's 2
inches by 2 inches.  Decent for prototypes...

Harvey

>
>                               - Jerry Kaidor
>
>

It will be just fine Jerry.
I've done 0.5mm spacing a number of times.

The trick is to make slightly narrower pads than spaces. Something
like 6.66mil or 8mil works good for me (select a value which the
resolution of your printer can resolve).
You want to make narrower pads than gaps because this way you have
some margin with the toner spreading and stuff, and soldering will
also be easier.

TQFP can be annoying to solder, because it likes to bridge between the
leads. Flux will help and and desolder braid to clear shorts.
QFN seems smaller at first glance, but because of the leadless nature
i find it much easier to solder, there is much less tendency to
bridge.

ST

On Fri, Feb 20, 2009 at 9:27 PM, jerrytr2.com <jerry@...> wrote:
> Hello,
>
>   Since encountering this group a month or so ago, I have refined my
> toner transfer process to the point that I am able to reliably do 15-
> mil traces.  Maybe smaller.  Hope so.
>
>   I'm doing a home project and have settled on PIC24F
> microcontrollers.  The latest version needs USB, and all the PICs
> that support USB are surface mount.  So I have been dragged kicking
> and screaming into the 21st century.
>
>   I bought an Aoyue 968 hot-air SMD rework station and some fine
> tweezers and have been  having a good time scrounging parts off some
> defunct computer boards - just to get used to the tool.  I'm going to
> need a microscope - half of these parts just look like bits of dirt!
>
>   One part I scrounged is the same package as the 24FJ256GB106
> processor I'm planning to use - a 64-pin TQFP.  Good LORD those pins
> are small.  Can Toner Transfer make boards this tight? According to
> the spec sheet, the pin pitch is .5mm, or about 20 thousandths of an
> inch.  So the traces need to be 10 thousandths of an inch, and will
> only have 10 thousandths of separation.  Ouch.
>
>                               - Jerry Kaidor
>
>

----- Original Message ----- 

From: "jerrytr2.com" <jerry@...>
To: <Homebrew_PCBs@yahoogroups.com>
Sent: Friday, February 20, 2009 8:27 PM
Subject: [Homebrew_PCBs] Surface Mount


> Hello,
>
>   Since encountering this group a month or so ago, I have refined my
> toner transfer process to the point that I am able to reliably do 15-
> mil traces.  Maybe smaller.  Hope so.
>
>   I'm doing a home project and have settled on PIC24F
> microcontrollers.  The latest version needs USB, and all the PICs
> that support USB are surface mount.  So I have been dragged kicking
> and screaming into the 21st century.
>
>   I bought an Aoyue 968 hot-air SMD rework station and some fine
> tweezers and have been  having a good time scrounging parts off some
> defunct computer boards - just to get used to the tool.  I'm going to
> need a microscope - half of these parts just look like bits of dirt!
>
>   One part I scrounged is the same package as the 24FJ256GB106
> processor I'm planning to use - a 64-pin TQFP.  Good LORD those pins
> are small.  Can Toner Transfer make boards this tight? According to
> the spec sheet, the pin pitch is .5mm, or about 20 thousandths of an
> inch.  So the traces need to be 10 thousandths of an inch, and will
> only have 10 thousandths of separation.  Ouch.

Some people can do that sort thing with TT, but it can be difficult. I can 
do fine pitch devices and 8/8 mil tracks with photo-etch without any 
problems.

Leon

So I have been dragged kicking 
> >and screaming into the 21st century.
> 
> Ain't Technology wonderful?

*** It sure has come a long way since the 6V6GT's and 5U4GB's of my 
youth...

> Yeah, now try to measure them.  Seems to be the 50 dollar solution 
and
> the 350 dollar solution, other than the 10 dollar ebay tweezers
> (only).

*** I bought the $10 ebay tweezers.  They work pretty good, except 
that the tips are pretty dull/rounded, and the  pivot a just a little
wobbly.  Also, my DMM that matches them ( a Fluke 10 ) doesn't do 
capacitance.  I have another credit-card sized DMM that does do C, 
but it has it's own can't-change-em test leads.  Lucky I'm a chinese 
food fan, and long ago mastered the use of chopsticks :).  

 
> You'll either want a small tip (almost a needle) soldering pencil 
for
> the leads, or you'll want to tin-lead coat the pads before 
soldering.

*** Probably tin-coat before.  I just ordered flux and a roll of 
20-mil rosin-core 63/37 solder.

> Very find gauge solder is also a must.  I find that upon occasion, a
> larger soldering iron to solder multiple pins at once, and a
> desoldering tool (built in air pump kind, not spring release) cleans
> up things nicely.  NOT elegant, I know.

*** I ordered some rolls of solder wick, with just that in mind.

> 
> If you had the right stuff, then a syringe of solder paste and your
> hot air pencil would be the optimum method.  (Right stuff defined as
> right tools and paste, etc...)

*** I did order a kit of solder paste from www.celeritous.com.  The
lead-free kind, because paste has to be kept in the fridge, and I 
really don't want to keep a tube of lead in with my sandwiches!
But further research into the characteristics of lead-free solder has
me thinking that I might want to skip that. Maybe get one of those
little coolers with peltier junctions just for the solder paste.

> 
> Yeah, possible to do, messy, and then there's the matter of vias...

*** Close inspection of my scrounged TQFP-64 convinces me that I 
cannot put any through-holes under it, because there is no space for
the height of the wire sticking through.

> 
> An alternate option is to use a carrier, such as the schmartboard
> carriers, and go from there.  Chip will be on a carrier that's 2
> inches by 2 inches. 
*** I looked at that.  Really porks out the part...

                                    - Jerry Kaidor

----- Original Message ----- 

From: "Harvey White" <madyn@...>
To: <Homebrew_PCBs@yahoogroups.com>
Sent: Friday, February 20, 2009 9:01 PM
Subject: Re: [Homebrew_PCBs] Surface Mount


> On Fri, 20 Feb 2009 20:27:33 -0000, you wrote:
> 
> 
>>
>>   One part I scrounged is the same package as the 24FJ256GB106 
>>processor I'm planning to use - a 64-pin TQFP.  Good LORD those pins 
>>are small.  Can Toner Transfer make boards this tight? According to 
>>the spec sheet, the pin pitch is .5mm, or about 20 thousandths of an 
>>inch.  So the traces need to be 10 thousandths of an inch, and will 
>>only have 10 thousandths of separation.  Ouch.
> 
> In theory, yes.  You might find yourself using 8 mm traces, or 10 mm
> traces and somewhat larger pins.  I inadvertently made a board with
> 0.5 mm spacing pins on one 4 pin connector, and it turned out semi
> "OK".  The whole board was a bit over etched, so the pins did not look
> as nice as I would want, but otherwise ok, I'd suppose.
> 
> You'll either want a small tip (almost a needle) soldering pencil for
> the leads, or you'll want to tin-lead coat the pads before soldering.
> Very find gauge solder is also a must.  I find that upon occasion, a
> larger soldering iron to solder multiple pins at once, and a
> desoldering tool (built in air pump kind, not spring release) cleans
> up things nicely.  NOT elegant, I know.

Drag-soldering is much quickerl.

Leon
--
Leon Heller
Amateur radio call-sign  G1HSM
Yaesu FT-817ND transceiver
Suzuki SV1000S motorcycle
leon355@...
http://www.geocities.com/leon_heller

"leon Heller" <leon355@...> writes:
> Drag-soldering is much quickerl.

I've found that drag soldering is MUCH more reliable if I pen-flux the
bottom of the pins on the part, as well as the board.  It just sucks
the solder under each pin, and reduces the risk of bridging.

----- Original Message ----- 

From: "DJ Delorie" <dj@...>
To: <Homebrew_PCBs@yahoogroups.com>
Sent: Friday, February 20, 2009 11:06 PM
Subject: Re: [Homebrew_PCBs] Surface Mount


> 
> "leon Heller" <leon355@...> writes:
>> Drag-soldering is much quickerl.
> 
> I've found that drag soldering is MUCH more reliable if I pen-flux the
> bottom of the pins on the part, as well as the board.  It just sucks
> the solder under each pin, and reduces the risk of bridging.

I use plenty of gel flux and rarely get any bridging.

Leon

"leon Heller" <leon355@...> writes:
> I use plenty of gel flux and rarely get any bridging.

IIRC the only bridging I get is on the last two pins of a "swipe"
because there's nothing after the last pin to suck the solder blob
along.  I guess with practice I'll get the blob size correct ;-)

Leon,     

I use the TT method and with care I can do 1 mil lines.
I can do 3mill without problems.

This is how I make the boards.

The best solution I have ever used is one part
muriatic acid(27% the kind used to clean bricks) 
and two parts hydrogen peroxide ( 3 % , the
kind you can get at Walmart or drug stores).

This will  etch in 5 - 6 minutes and does not need heat.
You mix only enough to cover the board in a small plastic tray.

The printer you use is very important(at least the toner is).
A brother printer will not work.
I use an HP P1505 and it works the best I have seen.
I think HP printer is part of why I get good resolution.

I clean the board very good. I use Synthetic Steel 
Wool 2 made by hodes/American to scrub the board 
first .Then I clean it with BAR KEEPERS FRIEND 
that you can get at Walmart. The cleaner the board 
the better.  Do not get finger prints(oil from fingers)on 
the board.  You might want to use latex gloves starting 
at this point.  After you scrub the board good with
BAR KEEPERS FRIEND wash it good with soap 
and cold water.  Now take some 91% alcohol and 
wipe the board clean.  Wipe until there is no residue
left on the wipe cloth.  Now put the board in an oven that 
is preheated to 350 degrees and let it stay for 
3 minutes.

Take the board out of the oven and put the toner paper 
on the board. Be careful at this point and lay the paper 
down easy because it will stick to the hot board.  Now lay 
a clean piece of cloth over the paper and iron the board 
for one minute.  Apply very light pressure as you move 
the iron all over the board.

Now put the board in water and let it soak for about 20 
minutes the paper will float off the board with very little help.  
Now after the paper is off put the board back in the oven 
for about 3 minutes. It is now ready to etch.

I print the circuit on some Laser Jet color paper that
I bought at Sams. It is made by a company named 
Royal Brights,  108 Main Street,  Norwalk,  CT 06851 
phone 1-800-526-4280. It is Glossy Ink  Jet Paper 200 
sheets per box,  this is all I can find on 
the box.  You could call them and find out where to get it.

The reason I use this paper is after you iron it on and soak
it about 20 minuter the paper will float off the board with 
very little help.  I think the paper is one reason I get such 
high resolution. 

  ----- Original Message ----- 
  From: leon Heller 
  To: Homebrew_PCBs@yahoogroups.com 
  Sent: Friday, February 20, 2009 4:28 PM
  Subject: Re: [Homebrew_PCBs] Surface Mount


  ----- Original Message ----- 
  From: "jerrytr2.com" <jerry@...>
  To: <Homebrew_PCBs@yahoogroups.com>
  Sent: Friday, February 20, 2009 8:27 PM
  Subject: [Homebrew_PCBs] Surface Mount

  > Hello,
  >
  > Since encountering this group a month or so ago, I have refined my
  > toner transfer process to the point that I am able to reliably do 15-
  > mil traces. Maybe smaller. Hope so.
  >
  > I'm doing a home project and have settled on PIC24F
  > microcontrollers. The latest version needs USB, and all the PICs
  > that support USB are surface mount. So I have been dragged kicking
  > and screaming into the 21st century.
  >
  > I bought an Aoyue 968 hot-air SMD rework station and some fine
  > tweezers and have been having a good time scrounging parts off some
  > defunct computer boards - just to get used to the tool. I'm going to
  > need a microscope - half of these parts just look like bits of dirt!
  >
  > One part I scrounged is the same package as the 24FJ256GB106
  > processor I'm planning to use - a 64-pin TQFP. Good LORD those pins
  > are small. Can Toner Transfer make boards this tight? According to
  > the spec sheet, the pin pitch is .5mm, or about 20 thousandths of an
  > inch. So the traces need to be 10 thousandths of an inch, and will
  > only have 10 thousandths of separation. Ouch.

  Some people can do that sort thing with TT, but it can be difficult. I can 
  do fine pitch devices and 8/8 mil tracks with photo-etch without any 
  problems.

  Leon 


  

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

----- Original Message ----- 

From: "DJ Delorie" <dj@...>
To: <Homebrew_PCBs@yahoogroups.com>
Sent: Saturday, February 21, 2009 12:45 AM
Subject: Re: [Homebrew_PCBs] Surface Mount


>
> "leon Heller" <leon355@...> writes:
>> I use plenty of gel flux and rarely get any bridging.
>
> IIRC the only bridging I get is on the last two pins of a "swipe"
> because there's nothing after the last pin to suck the solder blob
> along.  I guess with practice I'll get the blob size correct ;-)

I use a mini-hoof cartridge with my Metcal system and don't get that 
problem. It holds just the right amount of solder.

Leon
--
Leon Heller
Amateur radio call-sign  G1HSM
Yaesu FT-817ND transceiver
Suzuki SV1000S motorcycle
leon355@...
http://www.geocities.com/leon_heller

----- Original Message ----- 

From: "Derward Myrick" <wdmyrick@...>
To: <Homebrew_PCBs@yahoogroups.com>
Cc: <wdmyrick@...>
Sent: Saturday, February 21, 2009 2:14 AM
Subject: Re: [Homebrew_PCBs] Surface Mount


> Leon,     
> 
> I use the TT method and with care I can do 1 mil lines.
> I can do 3mill without problems.

Can you really do 3 mil (003") tracks? I don't believe it!

Leon
--
Leon Heller
Amateur radio call-sign  G1HSM
Yaesu FT-817ND transceiver
Suzuki SV1000S motorcycle
leon355@...
http://www.geocities.com/leon_heller

The mini hoof probably pulls the excess solder from the last two pins
into the hoof where a normal iron will not.

One tip for drag-soldering with a normal penciltip iron: Don't drag
the point behind or hold it perpendicular, instead push the tip over
the pins at an angle with the point leading. This will lead to fewer
bridges because the trailing edge of the iron is at a narrow angle
allowing the excess solder to transfer to the iron. If the tip is
trailing the solder is "left behind" with nowhere to go except the
pins.

Between hot-air rework and hotplate reflow, i rarely drag solder parts.

ST

On Sat, Feb 21, 2009 at 12:45 PM, leon Heller <leon355@...> wrote:

> I use a mini-hoof cartridge with my Metcal system and don't get that
> problem. It holds just the right amount of solder.
>
> Leon
> --

----- Original Message ----- 

From: "Stefan Trethan" <stefan_trethan@...>
To: <Homebrew_PCBs@yahoogroups.com>
Sent: Saturday, February 21, 2009 3:06 PM
Subject: Re: [Homebrew_PCBs] Surface Mount


> The mini hoof probably pulls the excess solder from the last two pins
> into the hoof where a normal iron will not.

I've just done a 44QFP part on a home-made PCB. It had 0.8 mm lead spacing 
so I used the larger hoof cartridge instead of the mini-hoof. I draw the tip 
towards me on the last lead in a row, to make sure that bridging doesn't 
ocurr. Metcal recommends using a clean hoof tip to remove bridges, the 
depression sucks up the excess solder.

Leon

Leon,

Besides that when I learned this method they called it Zip Soldering
:)  If you stop adding solder 2 pins before the end, there is usually
plenty of excess on the soldering iron to solder the last few pins.

On Fri, Feb 20, 2009 at 4:45 PM, DJ Delorie <dj@...> wrote:
>
> "leon Heller" <leon355@...> writes:
>> I use plenty of gel flux and rarely get any bridging.
>
> IIRC the only bridging I get is on the last two pins of a "swipe"
> because there's nothing after the last pin to suck the solder blob
> along.  I guess with practice I'll get the blob size correct ;-)
>
>

-- 
Erik L. Knise
Pacific Shipping Company
Seattle, WA

On Fri, 20 Feb 2009, jerrytr2.com wrote:

>    One part I scrounged is the same package as the 24FJ256GB106
> processor I'm planning to use - a 64-pin TQFP.  Good LORD those pins
> are small.  Can Toner Transfer make boards this tight? According to
> the spec sheet, the pin pitch is .5mm, or about 20 thousandths of an
> inch.  So the traces need to be 10 thousandths of an inch, and will
> only have 10 thousandths of separation.  Ouch.

10/10 is not too difficult, one of my first toner transfer PCBs were
10/10. I've since got down to 8/8 (for 0.4mm pitch LQFP).

I find hot air and solder paste does the easiest/neatest job when
soldering SMD.

Dylan Smith wrote:
> On Fri, 20 Feb 2009, jerrytr2.com wrote:
> 
>>    One part I scrounged is the same package as the 24FJ256GB106
>> processor I'm planning to use - a 64-pin TQFP.  Good LORD those pins
>> are small.  Can Toner Transfer make boards this tight? According to
>> the spec sheet, the pin pitch is .5mm, or about 20 thousandths of an
>> inch.  So the traces need to be 10 thousandths of an inch, and will
>> only have 10 thousandths of separation.  Ouch.
> 
> 10/10 is not too difficult, one of my first toner transfer PCBs were
> 10/10. I've since got down to 8/8 (for 0.4mm pitch LQFP).
> 
> I find hot air and solder paste does the easiest/neatest job when
> soldering SMD.
> 
> 

Do you apply a continuous trail of solder paste along each row of pads 
of a quad plat pack IC ? If so you you use a syringe to apply the paste ?.
I found it easier to just use solder wire and dragging the iron across. 
There are many videos on the net showing this method.

For all small SMD parts (resistors, caps, transistors,...) I found 
trying to apply paste to individual pads by syringe was waste of time. 
It was much easier to:

1) tin all pads using a soldering iron and 0.5mm solder wire.

2) use tweezers to lay all components on the board in approximate location.

3) apply liquid rosin flux over the whole PCB.

4) use two soldering irons to simultaneously maneuver the component and 
melt solder of each pad.


For desoldering large ICs I use a cheap electric hot plate from a 
kitchen shop. I place layer of glass cloth followed by the PCB. I get 
the PCB to 100 ~ 150C and then hot air gun on top to desolder the IC in 
question. Lifts off in a breeze, without any collateral damage.
Double sided heating like this is much less heat stress.

Well, obviously experiences and opinions will vary as always.
Your method involving the two irons seems awfully complicated compared
to paste....

ST

On Sun, Feb 22, 2009 at 11:39 PM, Adam Seychell <a_seychell@...> wrote:
>
> Do you apply a continuous trail of solder paste along each row of pads
> of a quad plat pack IC ? If so you you use a syringe to apply the paste ?.
> I found it easier to just use solder wire and dragging the iron across.
> There are many videos on the net showing this method.
>
> For all small SMD parts (resistors, caps, transistors,...) I found
> trying to apply paste to individual pads by syringe was waste of time.
> It was much easier to:
>
> 1) tin all pads using a soldering iron and 0.5mm solder wire.
>
> 2) use tweezers to lay all components on the board in approximate location.
>
> 3) apply liquid rosin flux over the whole PCB.
>
> 4) use two soldering irons to simultaneously maneuver the component and
> melt solder of each pad.
>
>
> For desoldering large ICs I use a cheap electric hot plate from a
> kitchen shop. I place layer of glass cloth followed by the PCB. I get
> the PCB to 100 ~ 150C and then hot air gun on top to desolder the IC in
> question. Lifts off in a breeze, without any collateral damage.
> Double sided heating like this is much less heat stress.
>

On Mon, 23 Feb 2009, Adam Seychell wrote:

> Do you apply a continuous trail of solder paste along each row of pads
> of a quad plat pack IC ? If so you you use a syringe to apply the paste ?.
> I found it easier to just use solder wire and dragging the iron across.
> There are many videos on the net showing this method.

Yes and yes. I used to do drag soldering but the problem I found was after
carefully lining up the IC, it was very easy to knock it out of alignment
with the pads while trying to tack down the first couple of pins. Using
paste and hot air, that problem didn't happen. (I also do all the discrete
components at the same time - I use 0603 generally, and the paste/hot air
does a very neat job).

Hi Jerry,
Once you get used to SMD, you'll never want to do through hole again!
I try to minimize all holes just to save drilling (which for me is the
biggest pain). Once you get used to them, the SMD parts are actually
pretty nice to work with. 0805's are the nicest and still have the
values marked on the package, but the 0603 size is good too if you
need to save space.

As for traces, beating 15/15 is no problem with TT.  10/10 is easily
obtained and I do it regularly (usually 12/12 rules with 10/10 where I
need it like the TQFP's). In a pinch, I can do 8/8 although consistent
traces and spaces at this level can be difficult as I still apply my
toner with a hand iron ... to little pressure and the trace gets
sparse, to much and ... where's the space? It's very doable though and
many boards have a few 8/8's on them.

My printer is an HP (both my 1200 and the 8000 series at work do an
excellent job ... the 1200 with reman cartridges is actually slightly
better).

Hey, cool microscope for this ... the Dino Lite PC scopes. Just got
one at work and it is excellent! A little pricey at $320 but really a
nice little scope.  There are many off brands for under $100 at Amazon
(some around 50 to 60, not as nice as the Dino Lite though). I'm going
to try one of those for home and hope it's not too cheesy. 

Phil (KA0HBG)


--- In Homebrew_PCBs@yahoogroups.com, "jerrytr2.com" <jerry@...> wrote:

>
> Hello,
> 
>    Since encountering this group a month or so ago, I have refined my 
> toner transfer process to the point that I am able to reliably do 15-
> mil traces.  Maybe smaller.  Hope so.
> 
>    I'm doing a home project and have settled on PIC24F 
> microcontrollers.  The latest version needs USB, and all the PICs 
> that support USB are surface mount.  So I have been dragged kicking 
> and screaming into the 21st century.
> 
>    I bought an Aoyue 968 hot-air SMD rework station and some fine 
> tweezers and have been  having a good time scrounging parts off some 
> defunct computer boards - just to get used to the tool.  I'm going to 
> need a microscope - half of these parts just look like bits of dirt!
> 
>    One part I scrounged is the same package as the 24FJ256GB106 
> processor I'm planning to use - a 64-pin TQFP.  Good LORD those pins 
> are small.  Can Toner Transfer make boards this tight? According to 
> the spec sheet, the pin pitch is .5mm, or about 20 thousandths of an 
> inch.  So the traces need to be 10 thousandths of an inch, and will 
> only have 10 thousandths of separation.  Ouch.
> 
>                                - Jerry Kaidor
>

Hi Phil,
I agree with you on all points. There is one "gotcha"  with surface mount --
it is more difficult to layout the pcb since placing tracks between the pins
is near impossible. You end up with many more vias or jumpers.

Myc

On Mon, Feb 23, 2009 at 10:29 AM, pgdion1 <pgdion1@...> wrote:

>   Hi Jerry,
> Once you get used to SMD, you'll never want to do through hole again!
> I try to minimize all holes just to save drilling (which for me is the
> biggest pain). Once you get used to them, the SMD parts are actually
> pretty nice to work with. 0805's are the nicest and still have the
> values marked on the package, but the 0603 size is good too if you
> need to save space.
>
> As for traces, beating 15/15 is no problem with TT. 10/10 is easily
> obtained and I do it regularly (usually 12/12 rules with 10/10 where I
> need it like the TQFP's). In a pinch, I can do 8/8 although consistent
> traces and spaces at this level can be difficult as I still apply my
> toner with a hand iron ... to little pressure and the trace gets
> sparse, to much and ... where's the space? It's very doable though and
> many boards have a few 8/8's on them.
>
> My printer is an HP (both my 1200 and the 8000 series at work do an
> excellent job ... the 1200 with reman cartridges is actually slightly
> better).
>
> Hey, cool microscope for this ... the Dino Lite PC scopes. Just got
> one at work and it is excellent! A little pricey at $320 but really a
> nice little scope. There are many off brands for under $100 at Amazon
> (some around 50 to 60, not as nice as the Dino Lite though). I'm going
> to try one of those for home and hope it's not too cheesy.
>
> Phil (KA0HBG)
>
> --- In Homebrew_PCBs@yahoogroups.com <Homebrew_PCBs%40yahoogroups.com>, "
> jerrytr2.com" <jerry@...> wrote:
> >
> > Hello,
> >
> > Since encountering this group a month or so ago, I have refined my
> > toner transfer process to the point that I am able to reliably do 15-
> > mil traces. Maybe smaller. Hope so.
> >
> > I'm doing a home project and have settled on PIC24F
> > microcontrollers. The latest version needs USB, and all the PICs
> > that support USB are surface mount. So I have been dragged kicking
> > and screaming into the 21st century.
> >
> > I bought an Aoyue 968 hot-air SMD rework station and some fine
> > tweezers and have been having a good time scrounging parts off some
> > defunct computer boards - just to get used to the tool. I'm going to
> > need a microscope - half of these parts just look like bits of dirt!
> >
> > One part I scrounged is the same package as the 24FJ256GB106
> > processor I'm planning to use - a 64-pin TQFP. Good LORD those pins
> > are small. Can Toner Transfer make boards this tight? According to
> > the spec sheet, the pin pitch is .5mm, or about 20 thousandths of an
> > inch. So the traces need to be 10 thousandths of an inch, and will
> > only have 10 thousandths of separation. Ouch.
> >
> > - Jerry Kaidor
> >
>
>  
>


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

PHIL,

I make my regular run up to the smd and then I go with a 1 to 3 mil
line through the smd and then go back to the regular trace.
This will eliminate nost jumpers.

Derward myerick  KD5WWI 

  ----- Original Message ----- 
  From: Myc Holmes 
  To: Homebrew_PCBs@yahoogroups.com 
  Sent: Monday, February 23, 2009 9:45 AM
  Subject: Re: [Homebrew_PCBs] Re: Surface Mount


  Hi Phil,
  I agree with you on all points. There is one "gotcha" with surface mount --
  it is more difficult to layout the pcb since placing tracks between the pins
  is near impossible. You end up with many more vias or jumpers.

  Myc

  On Mon, Feb 23, 2009 at 10:29 AM, pgdion1 <pgdion1@...> wrote:

  > Hi Jerry,
  > Once you get used to SMD, you'll never want to do through hole again!
  > I try to minimize all holes just to save drilling (which for me is the
  > biggest pain). Once you get used to them, the SMD parts are actually
  > pretty nice to work with. 0805's are the nicest and still have the
  > values marked on the package, but the 0603 size is good too if you
  > need to save space.
  >
  > As for traces, beating 15/15 is no problem with TT. 10/10 is easily
  > obtained and I do it regularly (usually 12/12 rules with 10/10 where I
  > need it like the TQFP's). In a pinch, I can do 8/8 although consistent
  > traces and spaces at this level can be difficult as I still apply my
  > toner with a hand iron ... to little pressure and the trace gets
  > sparse, to much and ... where's the space? It's very doable though and
  > many boards have a few 8/8's on them.
  >
  > My printer is an HP (both my 1200 and the 8000 series at work do an
  > excellent job ... the 1200 with reman cartridges is actually slightly
  > better).
  >
  > Hey, cool microscope for this ... the Dino Lite PC scopes. Just got
  > one at work and it is excellent! A little pricey at $320 but really a
  > nice little scope. There are many off brands for under $100 at Amazon
  > (some around 50 to 60, not as nice as the Dino Lite though). I'm going
  > to try one of those for home and hope it's not too cheesy.
  >
  > Phil (KA0HBG)
  >
  > --- In Homebrew_PCBs@yahoogroups.com <Homebrew_PCBs%40yahoogroups.com>, "
  > jerrytr2.com" <jerry@...> wrote:
  > >
  > > Hello,
  > >
  > > Since encountering this group a month or so ago, I have refined my
  > > toner transfer process to the point that I am able to reliably do 15-
  > > mil traces. Maybe smaller. Hope so.
  > >
  > > I'm doing a home project and have settled on PIC24F
  > > microcontrollers. The latest version needs USB, and all the PICs
  > > that support USB are surface mount. So I have been dragged kicking
  > > and screaming into the 21st century.
  > >
  > > I bought an Aoyue 968 hot-air SMD rework station and some fine
  > > tweezers and have been having a good time scrounging parts off some
  > > defunct computer boards - just to get used to the tool. I'm going to
  > > need a microscope - half of these parts just look like bits of dirt!
  > >
  > > One part I scrounged is the same package as the 24FJ256GB106
  > > processor I'm planning to use - a 64-pin TQFP. Good LORD those pins
  > > are small. Can Toner Transfer make boards this tight? According to
  > > the spec sheet, the pin pitch is .5mm, or about 20 thousandths of an
  > > inch. So the traces need to be 10 thousandths of an inch, and will
  > > only have 10 thousandths of separation. Ouch.
  > >
  > > - Jerry Kaidor
  > >
  >
  > 
  >

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


  

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

Well you seem to be the only one able to make 1 to 3 mil lines with TT.

Please describe what printer you use, what etchant, and what
precedure. I'm sure most people here are interested to hear about your
method.

Personally 6.66mil is the limit of what i am comfortable with. Much
thinner and i don't get continuity because of under-etching.

ST

On Mon, Feb 23, 2009 at 5:29 PM, Derward Myrick <wdmyrick@...> wrote:
> PHIL,
>
> I make my regular run up to the smd and then I go with a 1 to 3 mil
> line through the smd and then go back to the regular trace.
> This will eliminate nost jumpers.
>
> Derward myerick  KD5WWI
>
>

Stefan,

This is how I make the boards.

The best solution I have ever used is one part
muriatic acid(27% the kind used to clean bricks) 
and two parts hydrogen peroxide ( 3 % , the
kind you can get at Walmart or drug stores).
I only use it once I want it to etch fast.
This will  etch in 5 - 6 minutes and does not need heat.
You mix only enough to cover the board in a small plastic tray.
Rock it back and to untill etched.

The printer you use is very important(at least the toner is).
A brother printer will not work.
I use an HP P1505 and it works the best I have seen.
I think HP printer is part of why I get good resolution.

I clean the board very good with BAR KEEPERS FRIEND 
that you can get at Walmart. The cleaner the board 
the better.  Do not get finger prints(oil from fingers)on 
the board.  You might want to use latex gloves starting 
at this point.  After you scrub the board good with
BAR KEEPERS FRIEND wash it good with soap 
and cold water.  Now take some 91% alcohol and 
wipe the board clean.  Wipe until there is no residue
left on the wipe cloth.  Now put the board in an oven that 
is preheated to 350 degrees and let it stay for 
3 minutes.

Take the board out of the oven and put the toner paper 
on the board. Be careful at this point and lay the paper 
down easy because it will stick to the hot board.  
Apply very light pressure (almost no pressure) as you 
move the iron all over the board.  Iron for one minute.

Now put the board in water and let it soak for about 20 
minutes the paper will float off the board with very little help.
(I print the circuit on some Laser Jet color paper that
I bought at Sams. It is made by a company named 
Royal Brights,  108 Main Street,  Norwalk,  CT 06851 
phone 1-800-526-4280. It is Glossy Ink  Jet Paper 200 
sheets per box,  this is all I can find on the box.  You could 
call them and find out where to get it)  THIS  IS THE ONLY
PAPER I HAVE SEEN THAT FLOATS OFF THE BOARD 
WHEN SOAKED.
Now after the paper is off put the board back in the hot oven 
for about 3 minutes. It is now ready to etch.

This works very good for going under smd but is harder to do
for long runs.


Derward Myrick   KD5WWI

  ----- Original Message ----- 
  From: Stefan Trethan 
  To: Homebrew_PCBs@yahoogroups.com 
  Sent: Monday, February 23, 2009 11:54 AM
  Subject: Re: [Homebrew_PCBs] Re: Surface Mount


  Well you seem to be the only one able to make 1 to 3 mil lines with TT.

  Please describe what printer you use, what etchant, and what
  precedure. I'm sure most people here are interested to hear about your
  method.

  Personally 6.66mil is the limit of what i am comfortable with. Much
  thinner and i don't get continuity because of under-etching.

  ST

  On Mon, Feb 23, 2009 at 5:29 PM, Derward Myrick <wdmyrick@...> wrote:
  > PHIL,
  >
  > I make my regular run up to the smd and then I go with a 1 to 3 mil
  > line through the smd and then go back to the regular trace.
  > This will eliminate nost jumpers.
  >
  > Derward myerick KD5WWI
  >
  >

  

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

From your description our methods differ very little.

What copper foil thickness do you have on your boards?

Do you have any photos or scans of boards with such thin runs?

You can see a picture of a test PCB i did some time ago here:
<http://i40.tinypic.com/fk7dir.jpg>
As you can see the 3.33mil line already starts to break up, this is on
18um copper.

Thanks,

ST

On Mon, Feb 23, 2009 at 8:10 PM, Derward Myrick <wdmyrick@...> wrote:
> Stefan,
>
> This is how I make the boards.
>
> The best solution I have ever used is one part
> muriatic acid(27% the kind used to clean bricks)
> and two parts hydrogen peroxide ( 3 % , the
> kind you can get at Walmart or drug stores).
> I only use it once I want it to etch fast.
> This will  etch in 5 - 6 minutes and does not need heat.
> You mix only enough to cover the board in a small plastic tray.
> Rock it back and to untill etched.
>
> The printer you use is very important(at least the toner is).
> A brother printer will not work.
> I use an HP P1505 and it works the best I have seen.
> I think HP printer is part of why I get good resolution.
>
> I clean the board very good with BAR KEEPERS FRIEND
> that you can get at Walmart. The cleaner the board
> the better.  Do not get finger prints(oil from fingers)on
> the board.  You might want to use latex gloves starting
> at this point.  After you scrub the board good with
> BAR KEEPERS FRIEND wash it good with soap
> and cold water.  Now take some 91% alcohol and
> wipe the board clean.  Wipe until there is no residue
> left on the wipe cloth.  Now put the board in an oven that
> is preheated to 350 degrees and let it stay for
> 3 minutes.
>
> Take the board out of the oven and put the toner paper
> on the board. Be careful at this point and lay the paper
> down easy because it will stick to the hot board.
> Apply very light pressure (almost no pressure) as you
> move the iron all over the board.  Iron for one minute.
>
> Now put the board in water and let it soak for about 20
> minutes the paper will float off the board with very little help.
> (I print the circuit on some Laser Jet color paper that
> I bought at Sams. It is made by a company named
> Royal Brights,  108 Main Street,  Norwalk,  CT 06851
> phone 1-800-526-4280. It is Glossy Ink  Jet Paper 200
> sheets per box,  this is all I can find on the box.  You could
> call them and find out where to get it)  THIS  IS THE ONLY
> PAPER I HAVE SEEN THAT FLOATS OFF THE BOARD
> WHEN SOAKED.
> Now after the paper is off put the board back in the hot oven
> for about 3 minutes. It is now ready to etch.
>
> This works very good for going under smd but is harder to do
> for long runs.
>
>
> Derward Myrick   KD5WWI
>
>
>
>
>  ----- Original Message -----
>  From: Stefan Trethan
>  To: Homebrew_PCBs@yahoogroups.com
>  Sent: Monday, February 23, 2009 11:54 AM
>  Subject: Re: [Homebrew_PCBs] Re: Surface Mount
>
>
>  Well you seem to be the only one able to make 1 to 3 mil lines with TT.
>
>  Please describe what printer you use, what etchant, and what
>  precedure. I'm sure most people here are interested to hear about your
>  method.
>
>  Personally 6.66mil is the limit of what i am comfortable with. Much
>  thinner and i don't get continuity because of under-etching.
>
>  ST
>
>  On Mon, Feb 23, 2009 at 5:29 PM, Derward Myrick <wdmyrick@...> wrote:
>  > PHIL,
>  >
>  > I make my regular run up to the smd and then I go with a 1 to 3 mil
>  > line through the smd and then go back to the regular trace.
>  > This will eliminate nost jumpers.
>  >
>  > Derward myerick KD5WWI
>  >
>  >
>
>
>
> [Non-text portions of this message have been removed]
>
>
>
> ------------------------------------
>
> Be sure to visit the group home and check for new Links, Files, and Photos:
> http://groups.yahoo.com/group/Homebrew_PCBsYahoo! Groups Links
>
>
>
>

----- Original Message ----- 

  From: Stefan Trethan 
  To: Homebrew_PCBs@yahoogroups.com 
  Sent: Monday, February 23, 2009 1:21 PM
  Subject: Re: [Homebrew_PCBs] Re: Surface Mount


  I am using one mil copper.
  Will etch  a test board in a couple days and
  put it up for you to see.

  Derward


  __SII,_._,___

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

But one mil copper would be like 25um thick, as far as i know only
18um and 35um is available easily?
(It is called 1oz and 1/2oz of copper weight per some specific area in the US.)

I look forward to your photos, thanks.

ST

On Mon, Feb 23, 2009 at 8:43 PM, Derward Myrick <wdmyrick@...> wrote:

>  I am using one mil copper.
>  Will etch  a test board in a couple days and
>  put it up for you to see.
>
>  Derward
>
>

You are correct I did not mean 1 mil but 1 oz.

1 oz is about all I use for faster etch.

Derward

  ----- Original Message ----- 
  From: Stefan Trethan 
  To: Homebrew_PCBs@yahoogroups.com 
  Sent: Monday, February 23, 2009 1:51 PM
  Subject: Re: [Homebrew_PCBs] Re: Surface Mount


  But one mil copper would be like 25um thick, as far as i know only
  18um and 35um is available easily?
  (It is called 1oz and 1/2oz of copper weight per some specific area in the US.)

  I look forward to your photos, thanks.

  ST

  On Mon, Feb 23, 2009 at 8:43 PM, Derward Myrick <wdmyrick@...> wrote:

  > I am using one mil copper.
  > Will etch a test board in a couple days and
  > put it up for you to see.
  >
  > Derward
  >
  >

  

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

Derward Myrick wrote:
> PHIL,
> 
> I make my regular run up to the smd and then I go with a 1 to 3 mil
> line through the smd and then go back to the regular trace.
> This will eliminate nost jumpers.
> 

Be sure to include photos of your 1 mil (25um) width/spaced copper 
tracks, with the TT method. I'm curious.
I'm sure others here would be interested in how to create tracks 
narrower than the thickness of copper foil they are etching.

Derward

Could you confirm that by "1 to 3 mil" you mean 0.001 inch to 0.003
inch (1 to 3 milli-inches) and not 1 millimeter to 3 millimeters. This
confusion with the term mil has happened here before.

Regards
Warren

>   >
>   > I make my regular run up to the smd and then I go with a 1 to 3 mil
>   > line through the smd and then go back to the regular trace.
>   > This will eliminate nost jumpers.
>   >
>   > Derward myerick KD5WWI
>   >
>   >
> 
>   
> 
> [Non-text portions of this message have been removed]
>

I've been looking for a good microscope for a while but they are pricey.
My eyes aren't what they used to be I want to get more into smd
Maybe you can take some pics with your Dino Lite and share?
What model # is it?
 
Thanks
 

--- On Mon, 2/23/09, pgdion1 <pgdion1@...> wrote:

From: pgdion1 <pgdion1@...>
Subject: [Homebrew_PCBs] Re: Surface Mount
To: Homebrew_PCBs@yahoogroups.com
Date: Monday, February 23, 2009, 11:29 AM






Hi Jerry,
Once you get used to SMD, you'll never want to do through hole again!
I try to minimize all holes just to save drilling (which for me is the
biggest pain). Once you get used to them, the SMD parts are actually
pretty nice to work with. 0805's are the nicest and still have the
values marked on the package, but the 0603 size is good too if you
need to save space.

As for traces, beating 15/15 is no problem with TT. 10/10 is easily
obtained and I do it regularly (usually 12/12 rules with 10/10 where I
need it like the TQFP's). In a pinch, I can do 8/8 although consistent
traces and spaces at this level can be difficult as I still apply my
toner with a hand iron ... to little pressure and the trace gets
sparse, to much and ... where's the space? It's very doable though and
many boards have a few 8/8's on them.

My printer is an HP (both my 1200 and the 8000 series at work do an
excellent job ... the 1200 with reman cartridges is actually slightly
better).

Hey, cool microscope for this ... the Dino Lite PC scopes. Just got
one at work and it is excellent! A little pricey at $320 but really a
nice little scope. There are many off brands for under $100 at Amazon
(some around 50 to 60, not as nice as the Dino Lite though). I'm going
to try one of those for home and hope it's not too cheesy. 

Phil (KA0HBG)

--- In Homebrew_PCBs@ yahoogroups. com, "jerrytr2.com" <jerry@...> wrote:
>
> Hello,
> 
> Since encountering this group a month or so ago, I have refined my 
> toner transfer process to the point that I am able to reliably do 15-
> mil traces. Maybe smaller. Hope so.
> 
> I'm doing a home project and have settled on PIC24F 
> microcontrollers. The latest version needs USB, and all the PICs 
> that support USB are surface mount. So I have been dragged kicking 
> and screaming into the 21st century.
> 
> I bought an Aoyue 968 hot-air SMD rework station and some fine 
> tweezers and have been having a good time scrounging parts off some 
> defunct computer boards - just to get used to the tool. I'm going to 
> need a microscope - half of these parts just look like bits of dirt!
> 
> One part I scrounged is the same package as the 24FJ256GB106 
> processor I'm planning to use - a 64-pin TQFP. Good LORD those pins 
> are small. Can Toner Transfer make boards this tight? According to 
> the spec sheet, the pin pitch is .5mm, or about 20 thousandths of an 
> inch. So the traces need to be 10 thousandths of an inch, and will 
> only have 10 thousandths of separation. Ouch.
> 
> - Jerry Kaidor
>

















      

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

--- In Homebrew_PCBs@yahoogroups.com, Stefan Trethan
<stefan_trethan@...> wrote:
>
> But one mil copper would be like 25um thick, as far as i know only
> 18um and 35um is available easily?
> (It is called 1oz and 1/2oz of copper weight per some specific area
in the US.)
>


*Laughing* - you know Stefan, we tried to go metric back in the 70's,
it just wouldn't take here. Dare to be different I guess.  Funny how
we won't move from a systems that you fellows invented and abandoned
isn't it? Oh well, personally, I admit I do like the English system of
measurement better.

Anyways, the copper thickness is in oz per square foot. One in^3 of
copper has a mass of 5.179oz, so 1oz of copper has a volume of .1931
in^3. Divide this by 144 (12x12) to spread it out over an area of 1
ft² and you get a thickness of 1.3 mils  (0.0013") ... half that for
1/2oz copper.  Easy no?  (No!)

Phil

--- In Homebrew_PCBs@yahoogroups.com, "Derward Myrick" <wdmyrick@...>
wrote:
>
> PHIL,
> 
> I make my regular run up to the smd and then I go with a 1 to 3 mil
> line through the smd and then go back to the regular trace.


Wow Derward, that is impressive.

I, like Stefan, am amazed. I can do 8/8 well with TT (I prefer 10/10).
I think I could do 6mil if I needed to (especially for short traces)
but not sure I could push past that. the 8's get a little thin at
places as it is. I always figure I'd need to move to some form of
laminating to move past 8. But I see you use an iron too.

Perhaps cleansering the board helps (the Bar Keepers Friend), I always
stop at Dawn dish soap and a green dish scrubby. I agree with the
printer, my best results have come on the HP's.

I'm anxious to see your test board too.
It would be nice to fit between pads on an soic or fit more traces
under the 0805's.

Phil (KA0HBG)