Searching \ for '[EE] Board assembly' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=board+assembly
Search entire site for: 'Board assembly'.

Exact match. Not showing close matches.
PICList Thread
'[EE] Board assembly'
2007\11\12@093912 by Martin Klingensmith

face
flavicon
face
Is it atypical to put SMT components on both sides of a board? In your
experience is it significantly more expensive to have boards assembled
this way?

Thanks,
Martin K

2007\11\12@102615 by alan smith

picon face
Yes and yes.
 
 For BGA decoupling, it would be pretty ineffecient to push the caps that far from the planes and connections.  If you put one part on the back, might as well put as many R and C's as you can as the process and costs are the same, for the most part.  What happens is the board gets flowed twice, so larger and heavier parts shouldn't go on the back as they tend to misalign from the past thats holding it there and they end up being glued or held on...another cost added.  But the small parts usually hold onto the board just fine the first run the the oven and then it gets flipped and run thru again.
 
 So, its two paste stencils, two flows, two times thru the pick n place. So your costs are going to be higher just from the process.  Can it be avoided?  Completly depends on the design and the technology being used, and how dense you are making the board.  I've been doing double sided placements for more than 10 years now so its nothing that new.

Martin Klingensmith <spam_OUTmartinTakeThisOuTspamnnytech.net> wrote:
 Is it atypical to put SMT components on both sides of a board? In your
experience is it significantly more expensive to have boards assembled
this way?

Thanks,
Martin K

2007\11\12@103454 by Timothy J. Weber

face picon face
Martin Klingensmith wrote:
> Is it atypical to put SMT components on both sides of a board?

Depends on cost and the value of smallness.

> In your
> experience is it significantly more expensive to have boards assembled
> this way?

Yes.
--
Timothy J. Weber
http://timothyweber.org

2007\11\12@104021 by Vasile Surducan

face picon face
So, how you're doing for 1700+ balls where there is no room for
capacitors on the side where the BGA part is soldered ?
Usually could be the same distance to the supply planes to the top or
to the bottom...(imagine a 12 layer board with supply planes on layers
4,5 and 8,9.
On actual technology there is no other way than styffing on both sides.

On 11/12/07, alan smith <.....micro_eng2KILLspamspam@spam@yahoo.com> wrote:
{Quote hidden}

> -

2007\11\12@104808 by Martin Klingensmith

face
flavicon
face
Thanks Tim & Alan, that's what I wanted to know. I'll keep them all on
the top.
-
MK

Timothy J. Weber wrote:
{Quote hidden}

2007\11\12@110340 by John Ferrell

face picon face
I don't know about commercial production but hobby/experimental applications
make extensive use of the technique.

Look at the circuit boards in the yahoo group at
http://groups.yahoo.com/group/softrock40/


John Ferrell    W8CCW
"Life is easier if you learn to plow
      around the stumps"
http://DixieNC.US

----- Original Message -----
From: "Martin Klingensmith" <.....martinKILLspamspam.....nnytech.net>
To: "Microcontroller discussion list - Public." <EraseMEpiclistspam_OUTspamTakeThisOuTmit.edu>
Sent: Monday, November 12, 2007 9:39 AM
Subject: [EE] Board assembly


> Is it atypical to put SMT components on both sides of a board? In your
> experience is it significantly more expensive to have boards assembled
> this way?
>
> Thanks,
> Martin K
> --

2007\11\12@110344 by Herbert Graf

flavicon
face
On Mon, 2007-11-12 at 07:39 -0800, Vasile Surducan wrote:
> So, how you're doing for 1700+ balls where there is no room for
> capacitors on the side where the BGA part is soldered ?
> Usually could be the same distance to the supply planes to the top or
> to the bottom...(imagine a 12 layer board with supply planes on layers
> 4,5 and 8,9.
> On actual technology there is no other way than styffing on both sides.

Agreed, a couple years ago we did a board, not sure how many layers (20
+) and the only option was to put decoupling for a very large FPGA on
the "other" side. Xilinx really went a little nuts with decoupling
requirements on the Vertex4 FPGAs, there are whole pages of our
schematics dedicated just to decoupling. The datasheet has formulas to
determine how many caps you need. And they aren't all the same value
either. Those 0201 caps are REALLY small!

Fortunately, the decoupling requirements were reduced for the Vertex5
FPGAs, although since they have more capacity the actual number of caps
didn't go down that much! :)

While I try to reduce the number of components on the "other side" of
boards I design to a relatively small number, it isn't very high on my
priority list, especially for "small" components. The extra cost isn't
an issue for me, the issue is just the annoyance of having to flip a
board to probe something or remove/install a part on the underside.

TTYL

2007\11\12@125924 by Bob Axtell

face picon face
Martin Klingensmith wrote:
> Is it atypical to put SMT components on both sides of a board? In your
> experience is it significantly more expensive to have boards assembled
> this way?
>
> Thanks,
> Martin K
>  
Yes. In fact you will have trouble locating assembly houses that will do
this (unless it is only one or
two caps or resistors)..

The reasons are myriad, but the main ones are (1) reflow soldering
problems and (2) reliability
problems due to so much manual rework. If you look at electronic
cameras, cellphones etc you
will find FLEX circuits in which the devices are mounted on one side,
but are folded back on itself
to obtain the effect of both sides.

Another strategy I have seen is to use two very thin PCBs seperated by
only a spacer; both are populated
on one side only.

--Bob

2007\11\12@130551 by Vasile Surducan

face picon face
On 11/12/07, Herbert Graf <mailinglist3spamspam_OUTfarcite.net> wrote:
> On Mon, 2007-11-12 at 07:39 -0800, Vasile Surducan wrote:
> > So, how you're doing for 1700+ balls where there is no room for
> > capacitors on the side where the BGA part is soldered ?
> > Usually could be the same distance to the supply planes to the top or
> > to the bottom...(imagine a 12 layer board with supply planes on layers
> > 4,5 and 8,9.
> > On actual technology there is no other way than styffing on both sides.
>
> Agreed, a couple years ago we did a board, not sure how many layers (20
> +) and the only option was to put decoupling for a very large FPGA on
> the "other" side. Xilinx really went a little nuts with decoupling
> requirements on the Vertex4 FPGAs, there are whole pages of our
> schematics dedicated just to decoupling. The datasheet has formulas to
> determine how many caps you need. And they aren't all the same value
> either. Those 0201 caps are REALLY small!
>
> Fortunately, the decoupling requirements were reduced for the Vertex5
> FPGAs, although since they have more capacity the actual number of caps
> didn't go down that much! :)

I've just finished a design with the largest Virtex5 package on the
market. The fortune is not the smallest number of capacitors (if you
take a look on the suggested documentation, there are still too many
tantalum C size), but the unused IO pins which make room to other
components on the opposite layer (assuming through hole are used
instead of microvias and buried vias).
On the other hand 0201 capacitors have the worst decoupling parameters
because have quite large inductance.
I've tried to change even the 0402 with 0204 (which is quite difficult
on a Stratix II with 500+ balls and 95procent IO usage) but
succesfully on virtex5 with 50% IO usage.
What I mean is the rule we are talking here differs from design to
design. On a 90 ball BGA is one (simple) solution, ona 1760 ball BGA
is other sollution.

Vasile

2007\11\12@213315 by Vitaliy

flavicon
face
Bob Axtell wrote:
> Yes. In fact you will have trouble locating assembly houses that will do
> this (unless it is only one or
> two caps or resistors)..
>
> The reasons are myriad, but the main ones are (1) reflow soldering
> problems and (2) reliability
> problems due to so much manual rework. If you look at electronic
> cameras, cellphones etc you
> will find FLEX circuits in which the devices are mounted on one side,
> but are folded back on itself
> to obtain the effect of both sides.

I don't know which board houses you guys are talking about. We've been doing
double sided SMT since 2005, and none of the board houses we have contacted,
had any problem with the PCB being double sided. In our experience, the
additional cost is *not* significant (~$1/2 per board).

The old board house simply populated and ran the PCB through the oven twice.
Surface tension keeps components on the bottom from falling off. I'm
assuming the new board house we just switched to, builds them the same way.
Glue can be used to secure heavier components -- never had to do it, but the
rumor is that it adds quite a bit to the cost because of the glue drying
time.

I think Martin's best bet is to find a contract manufacturer, and put this
question to them. It's a good idea to run the design by them anyway, even
before you have the prototype -- ask them if they can see any opportunities
to reduce your cost with a small design change, or improve the board's
manufacturability.

> Another strategy I have seen is to use two very thin PCBs seperated by
> only a spacer; both are populated
> on one side only.

I've never seen such arrangement, but I've seen plenty of PCBs with SMT
components on both sides. They're not by any means "atypical".

Best regards,

Vitaliy

2007\11\12@224942 by Timothy J. Weber

face picon face
Vitaliy wrote:
> I don't know which board houses you guys are talking about. We've been doing
> double sided SMT since 2005, and none of the board houses we have contacted,
> had any problem with the PCB being double sided. In our experience, the
> additional cost is *not* significant (~$1/2 per board).

Interesting.  But that still is *some* money, so you would put them all
on one side unless you had a reason to do otherwise, right?  That seemed
like the OP's question.
--
Timothy J. Weber
http://timothyweber.org

2007\11\13@062944 by peter green

flavicon
face
Timothy J. Weber wrote:
> Vitaliy wrote:
>  
>> I don't know which board houses you guys are talking about. We've been doing
>> double sided SMT since 2005, and none of the board houses we have contacted,
>> had any problem with the PCB being double sided. In our experience, the
>> additional cost is *not* significant (~$1/2 per board).
>>    
>
> Interesting.  But that still is *some* money, so you would put them all
> on one side unless you had a reason to do otherwise, right?

Remember board area costs too so if your board house doesn't charge
extra or only charges a small ammount extra for going to components on
both sides then the space savings should make going to components on
both sides worthwhile.


2007\11\13@064358 by Spehro Pefhany

picon face
At 10:49 PM 11/12/2007, you wrote:
>Vitaliy wrote:
> > I don't know which board houses you guys are talking about. We've
> been doing
> > double sided SMT since 2005, and none of the board houses we have
> contacted,
> > had any problem with the PCB being double sided. In our experience, the
> > additional cost is *not* significant (~$1/2 per board).
>
>Interesting.  But that still is *some* money, so you would put them all
>on one side unless you had a reason to do otherwise, right?  That seemed
>like the OP's question.

The PCB can certainly be smaller and possibly can have a better layout if
you put parts on both sides. This is particularly true when there are more
than two layers. Unless you are fixed to a certain size of board (eg. by
corner mounting holes, grooves or clips), the cost savings from reducing
the square inches of a multilayer board will likely overshadow
the additional cost of placing parts on both sides (eg. using the adhesive
method before reflow).

With 2-layer boards you may wish to do a quasi-single-sided
layout in order to maintain a partial ground plane under the parts.

Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
@spam@speffKILLspamspaminterlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com



2007\11\13@071435 by Bryan Bishop

picon face
On Monday 12 November 2007 20:32, Vitaliy wrote:
> I don't know which board houses you guys are talking about.

What are board houses?

- Bryan

2007\11\13@072736 by Steve Howes

flavicon
face
>> I don't know which board houses you guys are talking about.
>
> What are board houses?

Where the board people live ;)

2007\11\13@073428 by Jake Anderson

flavicon
face
Bryan Bishop wrote:
> On Monday 12 November 2007 20:32, Vitaliy wrote:
>  
>> I don't know which board houses you guys are talking about.
>>    
>
> What are board houses?
>
> - Bryan
>  
Company that etches boards.

2007\11\13@160339 by Vitaliy

flavicon
face
Jake Anderson wrote:
>>> I don't know which board houses you guys are talking about.
>>>
>>
>> What are board houses?
>>
>> - Bryan
>>
> Company that etches boards.

Yeah, keep making fun of my primitive English... ;)

I meant "board assembly houses", in other words contract manufacturers that
actually populate the PCB.

2007\11\13@194953 by Vitaliy

flavicon
face
Spehro Pefhany wrote:
>>Interesting.  But that still is *some* money, so you would put them all
>>on one side unless you had a reason to do otherwise, right?  That seemed
>>like the OP's question.
>
> The PCB can certainly be smaller and possibly can have a better layout if
> you put parts on both sides. This is particularly true when there are more
> than two layers. Unless you are fixed to a certain size of board (eg. by
> corner mounting holes, grooves or clips), the cost savings from reducing
> the square inches of a multilayer board will likely overshadow
> the additional cost of placing parts on both sides (eg. using the adhesive
> method before reflow).
>
> With 2-layer boards you may wish to do a quasi-single-sided
> layout in order to maintain a partial ground plane under the parts.

Spehro is right. The only reason we have a board that's populated on both
sides, is we are constrained by the enclosure we're using, and the sheer
number of parts, which would not fit on one side.

I was arguing that contrary to what others have said, boards populated on
both sides are common, and they're not "significantly more expensive" (IMO,
of course it depends on the definition of "significantly"). One thing that
is not debatable, is that most (if not all), board assembly houses would not
have a problem with parts on both sides.

Vitaliy

2007\11\13@202901 by Timothy J. Weber

face picon face
Vitaliy wrote:
> Spehro is right. The only reason we have a board that's populated on both
> sides, is we are constrained by the enclosure we're using, and the sheer
> number of parts, which would not fit on one side.
>
> I was arguing that contrary to what others have said, boards populated on
> both sides are common, and they're not "significantly more expensive" (IMO,
> of course it depends on the definition of "significantly").

I think it's a very useful discussion of one of the usual many-variable
optimization problems that make up engineering!

In the typical case in whatever niche it is I'm currently in, my board
is constrained to a *minimum* size by the buttons and displays riding on
it, and there ends up being plenty of space in and around them to put a
PIC and glue logic.  Maybe this niche could be called something like
"trivial vertical-market automation" - no BGAs, PIC16F is a good match,
low density everything, never going to be pocket-sized anyway.

> One thing that
> is not debatable, is that most (if not all), board assembly houses would not
> have a problem with parts on both sides.

Agreed.
--
Timothy J. Weber
http://timothyweber.org

2007\11\14@070641 by Jake Anderson

flavicon
face
Vitaliy wrote:
> Jake Anderson wrote:
>  
>>>> I don't know which board houses you guys are talking about.
>>>>
>>>>        
>>> What are board houses?
>>>
>>> - Bryan
>>>
>>>      
>> Company that etches boards.
>>    
>
> Yeah, keep making fun of my primitive English... ;)
>
> I meant "board assembly houses", in other words contract manufacturers that
> actually populate the PCB.
>
>  
Board house is the "correct" term for it, its just jargon meaning if you
know what it means it makes sense and is generally a useful contraction
but if you don't know what it means it makes no sense. Kind of like a
second language. A native English speaker with a decent vocabulary would
probably be able to understand it due to similar usage with other terms
and the context it is in.

2007\11\14@072704 by Xiaofan Chen

face picon face
On Nov 13, 2007 7:41 PM, Spehro Pefhany <KILLspamspeffKILLspamspaminterlog.com> wrote:
>
> With 2-layer boards you may wish to do a quasi-single-sided
> layout in order to maintain a partial ground plane under the parts.

Yes that will be some cost saving in this case. Last time we will
always try that if possible. We would also try to avoid 4-layer board
as the cost goes up by quite a bit.

I've never used a single layer board. But I hear that quite some consumer
electronics product (eg: TV) are still trying to use single layer
board to reduce
cost. Sometimes I hear FR2 board is used instead of FR4.

Now we are trying to use 4-layer or 6-layer board if possible. Customer
wants to have higher density to reduce cost per I/O channel. So you
have to struggle with real estate on the boards and power consumption.


Xiaofan

2007\11\14@073717 by Xiaofan Chen

face picon face
On Nov 13, 2007 2:05 AM, Vasile Surducan <RemoveMEpiclist9TakeThisOuTspamgmail.com> wrote:
> > either. Those 0201 caps are REALLY small!

0201, even my eyesight is not good enough, you must have to use a
microscope.

> On the other hand 0201 capacitors have the worst decoupling parameters
> because have quite large inductance.
> I've tried to change even the 0402 with 0204

0204 capacitors? Never used them. They are not so common as 0603 or
0805. But Google tells me that 0204 capacitors have lower inductance.

We've just introduced many 0402 capacitors and resistors in the new
job. In the previous job, 0201/0402 ceramic (small size, for sensor, even
01005 is going to be used) is used quite often and 0204 metal frame (big,
for intrinsic safety fail safe resistor) is used also quite often.

Xiaofan

2007\11\14@084403 by Vasile Surducan

face picon face
On 11/13/07, Steve Howes <spamBeGonestevespamBeGonespamgeekinter.net> wrote:
> >> I don't know which board houses you guys are talking about.
> >
> > What are board houses?
>
> Where the board people live ;)

And where the "board people" are killing other people named "component
engineers",
"PCB design engineers" and finally the "product managers" step by step...

2007\11\14@095612 by Xiaofan Chen

face picon face
On Nov 14, 2007 9:44 PM, Vasile Surducan <TakeThisOuTpiclist9EraseMEspamspam_OUTgmail.com> wrote:
> On 11/13/07, Steve Howes <RemoveMEstevespamTakeThisOuTgeekinter.net> wrote:
> > >> I don't know which board houses you guys are talking about.
> > >
> > > What are board houses?
> >
> > Where the board people live ;)
>
> And where the "board people" are killing other people named
> "component engineers",  "PCB design engineers" and finally the
> "product managers" step by step...
>

Is this really that bad? We have component engineers, PCB design
engineers, design engineers and product managers all in-house.
And we do use partners for some product families. We always
buy PCB boards from outside and mostly do in-house assemblies.

In fact, quite some of the manufacturing people here come from
those EMS companies. And our PCB designer also comes from
a big EMS company.


Xiaofan

2007\11\14@125254 by Dave Tweed

face
flavicon
face
Xiaofan Chen <xiaofancEraseMEspam.....gmail.com> wrote:
> 0204 capacitors? Never used them. They are not so common as 0603 or
> 0805. But Google tells me that 0204 capacitors have lower inductance.

They're also very handy when working with 1 mm pitch BGAs -- they can be
fitted direcly within the grid of vias on the opposite side of the PCB
from the chip.

-- Dave Tweed

2007\11\14@134341 by M. Adam Davis

face picon face
To join the chorus:

On Nov 12, 2007 9:39 AM, Martin Klingensmith <EraseMEmartinspamnnytech.net> wrote:
> Is it atypical to put SMT components on both sides of a board?

No.  Any place where size, or signal path distance, or several other
of a variety of factors are important double sided placement is used.
However, you'll find that many (if not most) mass manufactured items
are still single sided assembly, single sided boards, with through
hole components hand assembled in factories in Asia.  Open up your
average home consumer device for an example.   Even modern DVD players
and VCRs are often single sided fab, single sided boards, with a
mixture of SMT (on the bottom when a suitable through hole is not
available) and through hole on the top.

That is cheap, cheap, cheap, and when it's suitable it's fine.  But
double sided assembly is far from atypical.

> In your experience is it significantly more expensive to have boards assembled
> this way?

For some values of 'significant' yes.  Careful design and engineering
can not only eliminate the extra cost, but possibly make it cheaper to
go double sided placement.  This is application specific, though.

The basic cost of assembly is time.  Let's look at the process:

Single side assembly:
0. Load pick and place machine, setup oven profile, load stencil, set
up other equipment
1. Clean board (if needed)
2. Apply solder paste (stencil)
3. Place parts
4. Bake
5. Wash boards (flux removal)
6. Visual inspection
7. Test, program, etc

Double sided assembly:
0. Load pick and place machine, setup oven profile, load stencil, set
up other equipment
1. Clean board (if needed)
2. Apply solder paste (stencil)
3. Place parts
4. Bake
5. Flip board
6. Load pick and place machine, load stencil, reset other equipment
7. Apply solder paste (stencil)
8. Place parts
9. Bake
10. Wash boards (flux removal)
11. Visual inspection
12. Test, program, etc

Practically speaking the assembly of a double sided board is twice the
time, and therefore twice the cost of a single sided assembly.  As
quantity goes up, however, certain cost savings come into effect - a
larger board house will have invested in a larger oven which, while
not faster, has a greater throughput (boards per hour) so each board
will cost less.  They may have chip shooters that load caps,
resistors, etc at thousands of parts per minute.  They invest in
automated test and inspection equipment so that inspection takes very
little time.

So larger quantities often make the cost insignificant.  For very
small quantities, machine setup and test/inspection dominate cost, so
the cost is very significant.

Then you need to consider the savings.  As others have mentioned,
placing parts on both sides reduces board size, which can be a
significant cost savings the more layers the board has.  On the flip
side (heh heh heh) a smaller double sided layout often requires more
layers.  But if you need a 6 layer board anyway for other reasons,
then there's little reason not to look at double sided assembly - the
board cost savings is significant.

Please note that pick and place time can be optimized to match the
oven speed (so the line goes as fast as it's slowest machine - the
oven).  Carefully decide how many and what type of parts you place on
each side.  You can reduce the setup time if you restrict all the 10k
resistors, for instance, to one side of the board - they won't have to
be loaded on the machine for the other side.   Panelize your boards
appropriately for the line setup - if the pick and place is taking too
long put fewer boards on the panel next time. Place heavy or large
parts on one side of the board and assemble that side last - surface
tension will keep small parts on the other side when upside down
during reflow.  Avoid through hole parts if possible - they often
require a separate assembly process, and will generally add more cost
to the board than double sided will (except in the case of very high
quantities with hand assembly in Asia).

I need to set up a design wiki.  I can tell I'm forgetting things that
I've learned in the past because I haven't used them recently...

-Adam

--
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - - - - - - - - - - - - - - - - - - - -
Moving in southeast Michigan? Buy my house: http://ubasics.com/house/

Interested in electronics? Check out the projects at http://ubasics.com

Building your own house? Check out http://ubasics.com/home/

2007\11\14@205634 by Vitaliy

flavicon
face
Jake Anderson wrote:
>> I meant "board assembly houses", in other words contract manufacturers
>> that
>> actually populate the PCB.
>>
>>
> Board house is the "correct" term for it, its just jargon meaning if you
> know what it means it makes sense and is generally a useful contraction
> but if you don't know what it means it makes no sense. Kind of like a
> second language. A native English speaker with a decent vocabulary would
> probably be able to understand it due to similar usage with other terms
> and the context it is in.

Hm, then what do you call places like Advanced Circuits or PCBexpress?

2007\11\14@220040 by James Newton

face picon face
On Nov 12, 2007 9:39 AM, Martin Klingensmith <RemoveMEmartinEraseMEspamEraseMEnnytech.net> wrote:
> Is it atypical to put SMT components on both sides of a board?

You stick the little beggars where ever you can find a place.

I've seen SMT components soldered INSIDE a hole drilled (and not plated)
between board layers. They couldn't find a place for the decoupling caps or
pull up resistors anywhere else on the board.

I've even seen one case where they stacked a batch of 0806 resistors on end
on the data pads and then air wired along the tops and down to the +Vcc pad
'cause the original design didn't need pull-ups and there was no room in the
physical space for them to be added. It was actually pretty solid.

That sort of thing requires hand assembly, of course, and it should not be
assumed that I'm recommending that sort of crap.

But nothing surprises me any more. The atypical is typical.

--
James.



2007\11\15@011229 by Xiaofan Chen

face picon face
On 11/15/07, James Newton <RemoveMEjamesnewtonspam_OUTspamKILLspammassmind.org> wrote:
> On Nov 12, 2007 9:39 AM, Martin Klingensmith <RemoveMEmartinTakeThisOuTspamspamnnytech.net> wrote:
> > Is it atypical to put SMT components on both sides of a board?
>
> You stick the little beggars where ever you can find a place.
>
> I've seen SMT components soldered INSIDE a hole drilled (and not plated)
> between board layers. They couldn't find a place for the decoupling caps or
> pull up resistors anywhere else on the board.
>
> I've even seen one case where they stacked a batch of 0806 resistors on end
> on the data pads and then air wired along the tops and down to the +Vcc pad
> 'cause the original design didn't need pull-ups and there was no room in the
> physical space for them to be added. It was actually pretty solid.
>
> That sort of thing requires hand assembly, of course, and it should not be
> assumed that I'm recommending that sort of crap.
>
> But nothing surprises me any more. The atypical is typical.
>

Are you sure they can indeed ship these kind of things in the final product?
At least not in the company I worked before.

Maybe you are right after all since I hear flying wires are considered
acceptible even in Thinkpad anyway...

Xiaofan

2007\11\15@041302 by Alan B. Pearce

face picon face
>> I've even seen one case where they stacked a batch of 0806 resistors on
>> end
>> on the data pads and then air wired along the tops and down to the +Vcc
>> pad
>> 'cause the original design didn't need pull-ups and there was no room in
>> the
>> physical space for them to be added. It was actually pretty solid.
>
>Are you sure they can indeed ship these kind of things in the final
>product?
>At least not in the company I worked before.
>
>Maybe you are right after all since I hear flying wires are considered
>acceptible even in Thinkpad anyway...

Umm, yes, we are close to shipping an instrument that has a resistor string
similar to that mentioned, for a mod on a board that is due for launch into
moon orbit next April ...

2007\11\15@050609 by Morgan Olsson
flavicon
face
This is a very good discussion!

I ant to add that keeping a board small, often do not save cost of board only:

o  Easier to fit in application; manhour and material cost

o  Smaller, cheaper encapsulation


--
Morgan Olsson

2007\11\15@051013 by Morgan Olsson

flavicon
face
Den 2007-11-15 04:01:13 skrev James Newton <EraseMEjamesnewtonspamspamspamBeGonemassmind.org>:

> On Nov 12, 2007 9:39 AM, Martin Klingensmith <RemoveMEmartinKILLspamspamnnytech.net> wrote:
>> Is it atypical to put SMT components on both sides of a board?
>
> You stick the little beggars where ever you can find a place.
>
> I've seen SMT components soldered INSIDE a hole drilled (and not plated)
> between board layers. They couldn't find a place for the decoupling caps or
> pull up resistors anywhere else on the board.
>
> I've even seen one case where they stacked a batch of 0806 resistors on end
> on the data pads and then air wired along the tops and down to the +Vcc pad
> 'cause the original design didn't need pull-ups and there was no room in the
> physical space for them to be added. It was actually pretty solid.
>
> That sort of thing requires hand assembly, of course, and it should not be
> assumed that I'm recommending that sort of crap.

It is also high performance; 3D: little interference, short track: low parasitic L, C, and R

Not to mention you can use real conductors for the power stage, not thooe wimpy copper foils...

I have only done this on prototypes... so far.
(except one low number design that has *NO* board, just wires and leaded and leadless cpmponents soldered on a few wires in a tight box of connectors...)


--
Morgan Olsson

2007\11\15@080215 by Martin Klingensmith

face
flavicon
face
Morgan Olsson wrote:
> It is also high performance; 3D: little interference, short track: low parasitic L, C, and R
>
> Not to mention you can use real conductors for the power stage, not thooe wimpy copper foils...
>
> I have only done this on prototypes... so far.
> (except one low number design that has *NO* board, just wires and leaded and leadless cpmponents soldered on a few wires in a tight box of connectors...)
>
>
>  

Maybe I should eliminate the circuit board and use point to point wiring
with solder terminal strips attached to a sheet metal chassis. Hell, it
wouldn't be that hard to just put some 12AX7s in there instead of opamps
too.
=)
-

2007\11\15@085359 by Morgan Olsson

flavicon
face
Den 2007-11-15 14:02:06 skrev Martin Klingensmith <martinSTOPspamspamspam_OUTnnytech.net>:

> Maybe I should eliminate the circuit board and use point to point wiring
> with solder terminal strips attached to a sheet metal chassis. Hell, it
> wouldn't be that hard to just put some 12AX7s in there instead of opamps
> too.

I have actually made a couple small prototypes by:
printing out the layout,
put double sided adhesive on the print
put thin copper vires on that as tracks, following the print
usig needle, put wires through on other side
put down the components
soldered

:)

But now i have improved my etch system and have a better printer so i can do single sided etched board with 0.5mm spacing.

--
Morgan Olsson

2007\11\17@132646 by Vitaliy

flavicon
face
Morgan Olsson wrote:
> This is a very good discussion!
>
> I ant to add that keeping a board small, often do not save cost of board
> only:
>
> o  Easier to fit in application; manhour and material cost
>
> o  Smaller, cheaper encapsulation

Good points.

2007\11\17@135637 by Vitaliy

flavicon
face
Morgan Olsson wrote:
> I have actually made a couple small prototypes by:
> printing out the layout,
> put double sided adhesive on the print
> put thin copper vires on that as tracks, following the print
> usig needle, put wires through on other side
> put down the components
> soldered

The method you describe sounds similar to what we used in the "electronics
club" (I was 12 at the time). Basically, we would take a piece of thin
cardboard (preferably white), draw the schematic, then punch holes and pull
wires where the connections should be. Then we would solder the parts to the
wires.

Here's something I dug out of my dusty old circuit box. You can tell it's
ancient (from my late 'teens) by the fact that I did not use US symbols for
resistors. :)

http://maksimov.org/circuit_top.jpg
http://maksimov.org/circuit_bottom.jpg

Some parts fell off, and in general stuff is "squished", but I think you get
the idea.

IMO, the advantage of such "schematic circuits" is that they are easier to
follow and troubleshot than, say, something assembled on veroboard. You
don't need to refer to a separate schematic -- the circuit *is* the
schematic. Changes are easy to make. And the thing does not fall apart as
easily as something put together on a breadboard.

Vitaliy

2007\11\19@154125 by James Newton

face picon face
That is pretty cool...

You left the component leads long to allow for possible future re-use?

Did any one use wood and nails?

--
James.

{Original Message removed}

2007\11\19@155312 by Bob Blick

face picon face

--- James Newton <spamBeGonejamesnewtonSTOPspamspamEraseMEmassmind.org> wrote:

> That is pretty cool...
>
> You left the component leads long to allow for
> possible future re-use?
>
> Did any one use wood and nails?

Yes, I did this with vacuum tube projects when I was a
kid!

Cheerful regards,

Bob

2007\11\19@160705 by Marcel Duchamp

picon face
Bob Blick wrote:

>> Did any one use wood and nails?
>
> Yes, I did this with vacuum tube projects when I was a
> kid!
>
> Cheerful regards,
>
> Bob
>

Coool! Although, I would have thought the glass would shatter when you
nailed down the tubes...

2007\11\19@162718 by David VanHorn

picon face
> You left the component leads long to allow for possible future re-use?
>
> Did any one use wood and nails?

Wood and brass nails, yes indeed!

2007\11\19@162755 by David VanHorn

picon face
>
> Coool! Although, I would have thought the glass would shatter when you
> nailed down the tubes...

Milspec metal case tubes!  :)

2007\11\19@162832 by M. Adam Davis

face picon face
On 11/19/07, Marcel Duchamp <KILLspammarcel.duchampspamBeGonespamsbcglobal.net> wrote:
> >> Did any one use wood and nails?
> >
> > Yes, I did this with vacuum tube projects when I was a
> > kid!
>
> Coool! Although, I would have thought the glass would shatter when you
> nailed down the tubes...

You have to use the more expensive glass nails...

-Adam

--
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - - - - - - - - - - - - - - - - - - - -
Moving in southeast Michigan? Buy my house: http://ubasics.com/house/

Interested in electronics? Check out the projects at http://ubasics.com

Building your own house? Check out http://ubasics.com/home/

2007\11\19@230408 by Peter van Hoof

face picon face
As a 8 year old I used to copy the schematic on paper life size, nail it
with thumtacks to a piece of board (the nails on all the connecting points
and then tin all thumtacks and solder the compinents and connecting wires
above the ones in the schematic

This way making errors with it was very difficult and as long as you knew
for sure what wire of a component was what it was hard to screw it up

One of the most impressive things I built with this was a 3 transistor FM
transmitter. with carbon microphone. It's amazing what cool gong sound
you create tapping your finger on a 5 turn 2.5mm copper air coil.

Peter van Hoof

----- Original Message ----
From: David VanHorn <EraseMEmicrobrixspamEraseMEgmail.com>
To: Microcontroller discussion list - Public. <@spam@piclist@spam@spamspam_OUTmit.edu>
Sent: Monday, November 19, 2007 4:27:18 PM
Subject: Re: [EE] Board assembly

> You left the component leads long to allow for possible future re-use?
>
> Did any one use wood and nails?

Wood and brass nails, yes indeed!

2007\11\20@071204 by Morgan Olsson

flavicon
face
Den 2007-11-19 21:41:58 skrev James Newton <spamBeGonejamesnewtonspamKILLspammassmind.org>:

> Did any one use wood and nails?

Of course.
Very good for learning.
I have used it very successively in educaiton too.
Very apparent schematic-like buildup.
Cheap = pupils can carry it home and are very happy.
Happy pupils learn much :)


--
Morgan Olsson

2007\11\20@123526 by Martin

face
flavicon
face
It just kills the gain.

Marcel Duchamp wrote:
>
>
> Coool! Although, I would have thought the glass would shatter when you
> nailed down the tubes...
>  

2007\11\20@131451 by Michael Rigby-Jones

picon face


>-----Original Message-----
>From: .....piclist-bouncesspam_OUTspammit.edu [TakeThisOuTpiclist-bounces.....spamTakeThisOuTmit.edu]
>On Behalf Of James Newton
>Sent: 19 November 2007 20:42
>To: 'Microcontroller discussion list - Public.'
>Subject: RE: [EE] Board assembly
>
>
>That is pretty cool...
>
>You left the component leads long to allow for possible future re-use?
>
>Did any one use wood and nails?

My father gave me one of the first books that got me interested in electronics, "Making a Transistor Radio" published by "Ladybird" who were more famous for young childrens books.  This showed how to build a radio in various stages, starting from a crystal set, adding aplification and finaly moving up to a full TRF set with speaker output.  This was built on a slab of wood using woodscrews and brass screw cups to trap the component wires.  Building this lead to my learning to solder as the screwcups proved rather unreliable and quite difficult when there were many wires to trap.

some pictures from the book:
"http://www.arar93.dsl.pipex.com/mds975/Content/trfradios02.html

Coincidently my wife was clearing out some of her old bits and pieces a couple of weeks back and found the same book that her father (electrical engineer) had given her.

Regards

Mike


=======================================================================
This e-mail is intended for the person it is addressed to only. The
information contained in it may be confidential and/or protected by
law. If you are not the intended recipient of this message, you must
not make any use of this information, or copy or show it to any
person. Please contact us immediately to tell us that you have
received this e-mail, and return the original to us. Any use,
forwarding, printing or copying of this message is strictly prohibited.
No part of this message can be considered a request for goods or
services.
=======================================================================

2007\11\21@042226 by Alan B. Pearce

face picon face
>>That is pretty cool...
>>
>>You left the component leads long to allow for possible future re-use?
>>
>>Did any one use wood and nails?
>
>My father gave me one of the first books that got me interested in
>electronics, "Making a Transistor Radio" published by "Ladybird"
...
>
>some pictures from the book:
> "http://www.arar93.dsl.pipex.com/mds975/Content/trfradios02.html

Boy those were the days ...

I remember getting a book out of the library when I was late primary school
age, on how to build a crystal set. The thing was built up in a manner
similar to the 'cardboard PCB' units that were linked here, but gave
dimensions for mounting lengths of 16 SWG wire onto a piece of Perspex (or
it might have been paxolin or bakelite, but my dad had some Perspex) and
then the components were soldered to that. Made quite a nice little unit,
and really kindled my interest in radio. The book went on to add a one and
two transistor amplifier to the same baseboard, but I never went that far
with it.

I do remember being impressed taking it to my grandparents place, as the
local AM 500kW transmitter was only about a mile or so as the crow flies
from the house, and I could listen to it without an aerial or ground.

2007\11\21@090420 by Martin Klingensmith

face
flavicon
face
Hell, you could probably have listened to it with a speaker connected
between ground and aerial.

Alan B. Pearce wrote:
> I do remember being impressed taking it to my grandparents place, as the
> local AM 500kW transmitter was only about a mile or so as the crow flies
> from the house, and I could listen to it without an aerial or ground.
>
>  

2007\11\21@151511 by Richard Prosser

picon face
If it's the place I'm thinking of (Titahi Bay)  I used to live there.
In dry conditions you could go up to the farmland under the mast, pull an
arc off a fence line & listen to the radio!

Crystal sets were OK, as long as you wanted to listen to either 2YA or 2ZB
(the 2 main stations being transmitted). Otherwise the selectivity was just
not good enough. I eventually figured out that if my (long wire) aerial was
pointed directly at the transmitter I got a null and could hear other
channels also. I tried a variety of "radio powered" circuits but didn't have
the skills, equipment or persistance in those days. Surfing was more fun.

RP


On 22/11/2007, Martin Klingensmith <TakeThisOuTmartinKILLspamspamspamnnytech.net> wrote:
{Quote hidden}

> -

2007\11\21@153514 by Spehro Pefhany

flavicon
face
Quoting Martin Klingensmith <.....martinspamRemoveMEnnytech.net>:

> Hell, you could probably have listened to it with a speaker connected
> between ground and aerial.

You'd also need a diode or some other kind of nonlinear device in there
somewhere. ;-)   A tarnished connection might suffice (copper oxide  
rectifier).

--

Best regards,
Spehro Pefhany, Gedex Inc.


2007\11\22@044025 by Alan B. Pearce

face picon face
>If it's the place I'm thinking of (Titahi Bay)  I used to live there.

Yes, that's right. My grandparents lived at Plimmerton.

>In dry conditions you could go up to the farmland under the mast,
>pull an arc off a fence line & listen to the radio!

I went for a visit to the transmitter as part of the course I did during my
apprenticeship. The stories that got told of various things that happened.

The taller of the two masts had both 2YA (570kHz IIRC) and 2YC (620kHz?) fed
to it through a combining & tuning network, with the mast length being tuned
between the two. There had been a lightening strike a couple of months
before our visit, and the station techs had stories about molten copper
pouring out of the combining network running around the floor. I hate to
think what the 500kW transmitter (2YC was only 60kW or 100kW IIRC) thought
of this sudden lack of load. Anyway apparently the techs lashed up a direct
connection to get 2YA back on the air as quick as possible.

Another trick was to climb the mast. There was a platform that you stood on
and then jumped across to a similar platform on the base of the mast so you
were not attempting to ground the transmission power through your body ...
The mast cross section was large enough that you climbed inside the lattice
structure, so were not subject to the voltage fields. It would take around
half an hour (if you were good and fit) to climb to the top. Apparently
someone who had done this had peed out through the lattice, and the
resulting liquid stream followed the a curve that represented the field
intensity around the mast.

2007\11\22@051524 by Richard Prosser

picon face
Alan,
I never got up the mast, the one time I was offered a trip up I chickened
out & have regretted it ever since.
We knew one of the riggers and he was not very complementary regarding the
state of the structure - & this was in the late 1960's. The mast is, of
course, still standing but has probably been replaced part by part several
times since then.

IIRC one of the hazards at the top (apart from the obvious) was that the
field strength was about 1kV/m so you were advised to keep below the top
rail as otherwise you would add to the effective mast height (700ft?).

Although the transmitter was rated at 500kW, I was told it wasn't run at
this level, 100kW ? being nearer the mark. Trying to extend the output tube
life. Big water cooled buggers.

2ZB & another station I can't remember  (a part time station) ran on a
slightly smaller mast nearby and the  Radio NZ SW transmitter was also on
site. Right next to the golf club where Michael Campbell learnt his trade.

RP


On 22/11/2007, Alan B. Pearce <RemoveMEA.B.PearcespamspamBeGonerl.ac.uk> wrote:
{Quote hidden}

> -

2007\11\22@054021 by Alan B. Pearce

face picon face
>I never got up the mast, the one time I was offered a trip up
>I chickened out & have regretted it ever since.

I didn't get an offer, but don't think I would have tried.

>We knew one of the riggers and he was not very complementary
>regarding the state of the structure - & this was in the late 1960's.

OK, it would have been '71 or '72 that I went there.

>2ZB & another station I can't remember  (a part time station)

2ZM IIRC, but may well be wrong.

>ran on a slightly smaller mast nearby

Yeah, it was also tuned between the two frequencies, and had a
matching/combiner network.

>and the  Radio NZ SW transmitter was also on site.

IIRC it was a wires on telegraph poles type arrangement. I think they had
several aerials they could switch between, depending on the Pacific Island
programme, the Antarctic program, or whatever else they ran, and where the
target audience was.

>Right next to the golf club where Michael Campbell learnt his trade.

;)

2007\11\22@061643 by Vasile Surducan

face picon face
On 11/14/07, Xiaofan Chen <spamBeGonexiaofanc@spam@spamspam_OUTgmail.com> wrote:
> On Nov 14, 2007 9:44 PM, Vasile Surducan <TakeThisOuTpiclist9spamspamgmail.com> wrote:
> > On 11/13/07, Steve Howes <steveEraseMEspamgeekinter.net> wrote:
> > > >> I don't know which board houses you guys are talking about.
> > > >
> > > > What are board houses?
> > >
> > > Where the board people live ;)
> >
> > And where the "board people" are killing other people named
> > "component engineers",  "PCB design engineers" and finally the
> > "product managers" step by step...
> >
>
> Is this really that bad?

It's worse than in comunism (ok, it's the same not worse, not better)
but you don't realise because you haven't any comparison term.
Fortunately I have it...

2007\11\23@151143 by Martin Klingensmith

face
flavicon
face
I don't claim to know much about AM stations, but FM stations are rated
for some amount of equivalent power. When you hear "500kW" or something
it's always much lower power with a higher antenna with more bays.
--
Martin K

Richard Prosser wrote:
{Quote hidden}

>> --

2007\11\24@233658 by Vitaliy

flavicon
face
James Newton wrote:
> That is pretty cool...
>
> You left the component leads long to allow for possible future re-use?

Yes, how did you know? ;)


2007\11\25@014318 by Vitaliy

flavicon
face
Peter van Hoof wrote:
> One of the most impressive things I built with this was a 3 transistor FM
> transmitter. with carbon microphone.

My FM transmitter had only one transistor, and the carbon mic was placed in
series with the battery. Worked surprisingly well. :)

> It's amazing what cool gong sound
> you create tapping your finger on a 5 turn 2.5mm copper air coil.

:)

I found it worked even better on a shortwave transmitter based on a vacuum
tube "HF generator" design from an ancient electronics book. Its LC had
something like a 10 turn 50 mm coil -- so I would take a pencil and go
"drrrrrrring.... briinggg". Frequency Modulation from an AM circuit, dude!
;-D

The coolest thing about this transmitter, was the recommended testing
procedure. "Connect a 6V light bulb to a wire loop (about 15 cm in
diameter). Place the loop about 20 cm from the generator's coil, and if the
circuit is working properly, the bulb should light up."

This project was my hands-on intro to harmonics -- at first, I couldn't
understand why I was able to pick up the signal on multiple bands (with
decreasing power).

Vitaliy

2007\11\25@015104 by Vitaliy

flavicon
face
Richard Prosser wrote:
> In dry conditions you could go up to the farmland under the mast, pull an
> arc off a fence line & listen to the radio!

On first reading, I imagined a "singing" electric arc. I would pay to see
that. :-D

OT: I once saw the schematic for a "modulated" spark transmitter that was
used on old Russian Navy ships. Essentially, it consisted of an AC generator
(1Khz?) whose output was fed to the spark chamber. Weird stuff...

2007\11\25@094602 by Carl Denk

flavicon
face
As I surf through the channels,  I occasionally hear from a TV
Evangelist: (not a follower) "Because the Bi.... said so" :)  [it must
be still early in the morning here, but, I'll click "Send" anyhow. :)  ]

Vitaliy wrote:
{Quote hidden}

More... (looser matching)
- Last day of these posts
- In 2007 , 2008 only
- Today
- New search...