Anybody using pcb2gcode?

[Nico Coesel]

[email protected] wrote:

(...)


Why is that? Done properly, they can be every bit as good as a
board made by a commercial house. Well, mine are, anyway.

With solder mask, white print and plated holes?

The competition is in dollars spent, not quality.
Even with a nearly manual conversion process, I had a double-sided
board within a week. Automate the most lengthy task
and that time would shrink dramatically.

Within a week? What is taking so long?

Quick turn around PCB manufacturers can get you a PCB within 3 days
(including shipping). I agree making a PCB yourself is always faster
(a matter of hours) but the end result cannot be compared with a PCB
made by a PCB manufacturer.

 

[Winston]

With solder mask, white print and plated holes?

(I like yellow silkscreen better because it contrasts more.)

You are right. A milled board does not have any of those things
unless one applies them later. The purpose of the board is
simply to verify proper operation of the electronics; not
even as a 'first article' product.

Within a week? What is taking so long?

The manual process of creating a toolpath for the cutter in G-code.
That's why I'm asking for the opinion of other folks that have
also made PCBs using the milling process. Not photolithography.

Milling. The application of a mechanical cutter to remove undesirable
copper from a laminated substrate. Not the use of resist and acid
to etch away the copper. I've had boards made using that process and
am comfortable with it and understand its strengths.

I would like information about people's experiences with the
mechanical rather than the electrochemical process though.

I'm really trying to be clear here.

Quick turn around PCB manufacturers can get you a PCB within 3 days
(including shipping).

You misspelled 'two weeks' in the context of a reasonably priced PCB.

I agree making a PCB yourself is always faster
(a matter of hours) but the end result cannot be compared with a PCB
made by a PCB manufacturer.

Sure it can.

* Fewer unnecessary features
* Much higher cost for tools to create the board unless you already
own the hardware.
* Very much cheaper and faster if the toolpath delay can be solved
* Admittedly a much greater time commitment.

There are many other points of comparison as well.

Can we please stop beating the 'photo lithography horse'?
I've used that process before and will use it in the future.
It is outside the context of my question, though.

--Winston

 

[Nico Coesel]

The manual process of creating a toolpath for the cutter in G-code.
That's why I'm asking for the opinion of other folks that have
also made PCBs using the milling process. Not photolithography.

You already came up with the pcb2gcode software. It looks like it
solves your problem. Why don't you just try it?

Most 'millers' I know use the LPKF machine for a while but then switch
back to using PCB manufacturers. This is mostly in commercial
environments though. The software for the LPKF machine seems to have
quite a learning curve.

The machines from Sherline look nice but I wonder about the
tolerances. The DIN connectors on their stepping motor kits are flaky
though and no position feedback.

Can we please stop beating the 'photo lithography horse'?
I've used that process before and will use it in the future.
It is outside the context of my question, though.

The real solution to a problem is almost always outside the context of
a question

 

[John Nagle]

I wonder if the vias can be plated using an 'electroforming' process?

Yes, but the setup for making plated-through holes is complex
enough that it's not worth it for single board runs.

With plated-through holes, the traces are held to the board at
every hole, and are much less likely to come off the board. So
you can use finer lines.

If you really want to make your own boards, photo-etching is
the way to go. (The toner-transfer thing sort of works, but there
are usually flaws in the board.) But really, the quick-turn
board houses are fast, cheap, easy to use, and Internet-connected
to CAD systems.

As someone who actually does CNC milling, it's just not the right
tool for board-making.

John Nagle

 

[Winston]

Nico Coesel wrote:

(...)

You already came up with the pcb2gcode software. It looks like it
solves your problem. Why don't you just try it?

I shall.

I'm just doing 'due diligence' by asking about other's experiences.
If 'pcb2gcode' was known to be far less effective than some other
package, I want to know.

For example if you said, "Maytag washers are on sale!"
I would tell you about my experience with their products
and give you the opportunity to have a happier life.

Reasonable, yes?

"Why didn't you just *ask?*" is the flip side of "Why don't you just try it?".

Most 'millers' I know use the LPKF machine for a while but then switch
back to using PCB manufacturers. This is mostly in commercial
environments though. The software for the LPKF machine seems to have
quite a learning curve.

I see that both of them are selling theirs:

http://cgi.ebay.com/LPKF-ProtoMat-C...585?pt=LH_DefaultDomain_0&hash=item19c4acc5e1
http://cgi.ebay.com/LPKF-ProtoMat-C...356?pt=LH_DefaultDomain_0&hash=item5add7942e4

The machines from Sherline look nice but I wonder about the
tolerances. The DIN connectors on their stepping motor kits are flaky
though and no position feedback.

I've not worked with the DIN connector. The connectors
that came with the Flashcut motors worked well.

After I determined the maximum step rate for each axis (with
a reasonable amount of back force), I selected 60% of that
number and programmed that in. 'Missing steps', if they
did occur, did not impair the functionality or appearance
of the resulting PCBs.

I found the tolerances of my Sherline mill to be quite good
enough for the task.

The real solution to a problem is almost always outside the context of
a question

OK, but few of us live next door to a PCB house that
charges almost no money to fabricate our boards
very quickly. Well, I don't anyway.

--Winston

 

[Winston]

Just found out the hard way that Maytag fabric softener dispensers are
auto-clogging

Saves on softener! It's a Feature!

Two days of soaking in denatured alcohol plus using a high pressure
washer to get it to a point that it would disassemble :-(

The instructions make no mention, but various web sites warn of the
problem... take apart monthly and clean with white vinegar.

Ouch. MTBF for Maytag washers and fridges is about 4 years in my
experience. Just Nasty.

--Winston

Economic recovery, one hysterically angry customer at a time.

 

[Winston]

John Nagle wrote:

(...)

But really, the quick-turn
board houses are fast, cheap, easy to use, and Internet-connected
to CAD systems.

So, a photo lithographic process then, hmmm? Wow!

Thanks John!


--Winston

 

[[email protected]]

With solder mask, white print and plated holes?


Within a week? What is taking so long?

I just got one in three days, from China. ...at least I think it was from
China. It was one of our regular suppliers. Other than a stupid mistake
(swapped TCK and TMS on the CPLD programming port), easily fixed on the test
fixture on the same board, it was flawless. We were worrying about the .5mm
pitch BGA. Another manufacturing roadblock bites the dust. ;-)

Quick turn around PCB manufacturers can get you a PCB within 3 days
(including shipping). I agree making a PCB yourself is always faster
(a matter of hours) but the end result cannot be compared with a PCB
made by a PCB manufacturer.

In man-hours it's rarely faster to roll your own. We had one done outside
without solder mask to save $20. What a mistake that was but management has
to learn somehow. ;-)

 

[[email protected]]

Saves on softener! It's a Feature!


Ouch. MTBF for Maytag washers and fridges is about 4 years in my
experience. Just Nasty.

Whirlpool isn't any different, evidently. We're into the third year on our
Cabrio washer and the bearings are getting *really* noisy. It's a $750 job to
replace them, from what I understand. Not a good deal on a $1000 washer.
That'll be the last Whirpool product I buy.

The 'fridge is an Electrolux. Hopefully it'll last a *lot* longer.

 

[John KD5YI]

Whirlpool isn't any different, evidently. We're into the third year on our
Cabrio washer and the bearings are getting *really* noisy. It's a $750 job to
replace them, from what I understand. Not a good deal on a $1000 washer.
That'll be the last Whirpool product I buy.

I have a Whirlpool dishwasher that I bought less than 4 years ago. It
has started leaking around the bottom of the door. I've looked for any-
and every-thing that might be causing it. No joy. No more Whirlpool for me.

John

 

[Winston]

[email protected] wrote:

(...)

Whirlpool isn't any different, evidently. We're into the third year on our
Cabrio washer and the bearings are getting *really* noisy. It's a $750 job to
replace them, from what I understand. Not a good deal on a $1000 washer.
That'll be the last Whirpool product I buy.

OMG.

When Whirlpool bought Maytag, I hoped Maytag
would start to care about what they sold.

Apparently the Maytag Monster infected Whirlpool
instead. That is sad. I replaced my very defective
Maytag washer with a Whirlpool a decade ago and it has been
like a dream. No issues. Well, OK a solenoid valve, but
I found a cheap NIB workalike and it runs like a top again.

I really hope the Whirlpool era is not ending.

The 'fridge is an Electrolux. Hopefully it'll last a *lot* longer.

Got my fingers crossed for ya.

If you have absolutely *nothing* to do, suggest looking at the make
of the compressor. If it doesn't say Tecumseh on the side, you
have a fighting chance.

--Winston

 

[Winston]

I just got one in three days, from China. ...at least I think it was from
China. It was one of our regular suppliers. Other than a stupid mistake
(swapped TCK and TMS on the CPLD programming port), easily fixed on the test
fixture on the same board, it was flawless. We were worrying about the .5mm
pitch BGA. Another manufacturing roadblock bites the dust. ;-)

Wow! That is astonishing. They must have turned the board in about
20 minutes and then put it in a time machine to get it through
Customs in time. Man!

Can you share the vendor name?

In man-hours it's rarely faster to roll your own. We had one done outside
without solder mask to save $20. What a mistake that was but management has
to learn somehow. ;-)

Yup.

--Winston

 

[[email protected]]

Are you tight on enclosure space or the part just doesn't come in anything
bigger? We've been tending to follow Joerg's advice lately, which is to avoid
BGAs when possible and use the biggest parts that'll actually fit...
especially for devices that get tossed around a lot...

I like BGAs but this is the first time we've gone under .8mm pitch with one
(.5mm isn't a problem on QFPs and such). QFNs caused us a lot of grief, until
we figured out that the problems were self-imposed (cross your feet, aim,
shoot). We've had extremely good luck with BGAs. I don't recall a failure
since we got our oven profile right and that was a RoHS issue. We just bought
a 12-stage oven, so that'll be a big improvement over the 4-stage, too.

Yeah, it's real tight. It's on a board sitting in the back-shell of a DE-9
connector. ;-) I was going to use a Xilinx Coolrunner in a QFN-32 but they
changed the spec and 32 cells wasn't going to cut it. I'm using a 240 cell
Altera Max-IIZ (and half of it is used ;-).

I'll have to post some pictures somewhere. We got the board back yesterday and
got the first sample working today. It'll be on both the Auburn and Alabama
sidelines on Saturday. ;-)

Wow. Definitely not worth $20 to omit a soldermask!

TSOPs and all. Was our tech pissed at him! ;-)

 

[[email protected]]

Wow! That is astonishing. They must have turned the board in about
20 minutes and then put it in a time machine to get it through
Customs in time. Man!

I've received stuff from DealExtream from China in three days. No problem,
though the US mails, even.

Can you share the vendor name?

I'll have to get it.

Yup.

It takes a long time to train a new manager, though sometimes it's hopeless.

 

[[email protected]]

Nice. Real Soon Now we're supposed to be building a new building at which
point the idea is that the production guys will get some fancier toys than
what we have now (...which would include a small 4-stage oven that's used for
prototype/low-volume production runs).

We had a lot of problems getting the 4-stage set up well enough to work and
even then there were real problem parts. We're hoping the 12-stage solves a
lot of these.

With all of what John Larkin's been saying about Xilinx's software lately I
keep meaning to look at Altera again; the only time I've used them in the past
was for some small CPLD where it just so happened they had the fastest part
available for the features I needed (...and this was a decade ago now); we
ended up with a board that had all of Xilinx, Altera, and Vantis (now Lattice)
parts on it!

I like Altera's support better. The Arrow FAE is pretty sharp and always
ready to help (and willing to come down for lunch or dinner ;-).

That'd be great!

I'll have to find a place (never done it). I just discovered that I hadn't
loaded my camera's software after the last Windows install. I'll have to get
it from Olympus, too.

 

[Jasen Betts]

I have a Whirlpool dishwasher that I bought less than 4 years ago. It
has started leaking around the bottom of the door. I've looked for any-
and every-thing that might be causing it.

could it be over full, a clogged or leaky level sensor line?

 

[John KD5YI]

could it be over full, a clogged or leaky level sensor line?

Repair guy said it *might* have been the level sensor which he adjusted.
That did not fix it.

I looks like a seal problem, but damn if I can find it.

 

[Winston]

Repair guy said it *might* have been the level sensor which he adjusted.
That did not fix it.

I looks like a seal problem, but damn if I can find it.

Hey John, Have you had a visit from an appliance repair guy
recently? Say he repaired the stove or something else?

After warranty service on my Jenn-Air (Gack!) cooktop/oven, my
nice new dishwasher started making very *different* noises during the
drain cycle. I checked it out and found a few 1" *rocks* trapped
*behind* the filter screen obstructing the drain.

I made the mistake of not watching the repair guy while he was
busy in the kitchen. After I removed the rocks, the dishwasher
returned to it's normal noises and has been trouble-free since.

Now when I have a service guy in the house, I watch him like a
hawk.

--Winston

 

[John KD5YI]

Hey John, Have you had a visit from an appliance repair guy
recently? Say he repaired the stove or something else?

Yes, I called him to fix the leak. It is no worse than before. Maybe it
even leaks a little less.

After warranty service on my Jenn-Air (Gack!) cooktop/oven, my
nice new dishwasher started making very *different* noises during the
drain cycle. I checked it out and found a few 1" *rocks* trapped
*behind* the filter screen obstructing the drain.

I made the mistake of not watching the repair guy while he was
busy in the kitchen. After I removed the rocks, the dishwasher
returned to it's normal noises and has been trouble-free since.

Now when I have a service guy in the house, I watch him like a
hawk.

--Winston

I didn't watch him, but the problem got no worse and no additional
problems appeared.

John

 

[Winston]

John KD5YI wrote:

(...)

I didn't watch him, but the problem got no worse and no additional
problems appeared.

I'd replace the door seal 'just because':
http://www.ehow.com/how_5675606_fix-seal-dishwasher.html

--Winston