SMT component minimum height

[Winfield Hill]

Many SMT components have maximum height specs,
but that can get one into trouble. My little
14-parameter center-of-the-beehive monitor,
has five sensor ICs with tiny holes, to let
air in for measurement. To keep bees from
covering sensor holes with waxy propolis, my
board will have a clear thermo-plastic cover.
The sensor heights are 0.8 to 0.9mm. So, to
keep the cover from covering the sensor holes,
I placed five 1.4mm tall parts on the PCB.
These should keep the plastic 0.4 to 0.6 mm
above the sensor holes.

AVX spec'd their 1uF, 25V 0805 caps at 1.4mm
tall, but, oops, they measure 0.77mm thick.
OK, Arrgh, the spec was for maximum height.
What we need is a minimum height spec!

Poking around in my inventory, I found TDK
C2012X7R1H225K125AC 2.2uF,50V 0805 caps,
that measured 1.34mm thick. TDK's spec is
nominal, 1.25 +/- 0.25 mm.

But if I use them, 3.3V-bus capacitance
will be 11uF, instead of 5uF. The sensor
power has a 50mA current limit, making a
11uF 3.3V / 50mA = 0.7ms turn-on risetime.


--
Thanks,
- Win

 

[Rick C]

On Friday, May 17, 2019 at 5:55:22 PM UTC-4, Winfield Hill wrote:

Many SMT components have maximum height specs,
but that can get one into trouble. My little
14-parameter center-of-the-beehive monitor,
has five sensor ICs with tiny holes, to let
air in for measurement. To keep bees from
covering sensor holes with waxy propolis, my
board will have a clear thermo-plastic cover.
The sensor heights are 0.8 to 0.9mm. So, to
keep the cover from covering the sensor holes,
I placed five 1.4mm tall parts on the PCB.
These should keep the plastic 0.4 to 0.6 mm
above the sensor holes.

AVX spec'd their 1uF, 25V 0805 caps at 1.4mm
tall, but, oops, they measure 0.77mm thick.
OK, Arrgh, the spec was for maximum height.
What we need is a minimum height spec!

Poking around in my inventory, I found TDK
C2012X7R1H225K125AC 2.2uF,50V 0805 caps,
that measured 1.34mm thick. TDK's spec is
nominal, 1.25 +/- 0.25 mm.

But if I use them, 3.3V-bus capacitance
will be 11uF, instead of 5uF. The sensor
power has a 50mA current limit, making a
11uF 3.3V / 50mA = 0.7ms turn-on risetime.

If you need a part for spacing, why bother using it for anything else?

https://www.digikey.com/product-detail/en/vishay-dale/ILSB1206ER100K/541-2607-1-ND/5419048

1.4 mm tall and $0.14 at qty 100. Is that too expensive?

If you really want a cap here is an 0805 at 1.25 Âą0.15 mm, so 1.10 mm min, 1.4 mm max.

https://www.digikey.com/product-detail/en/kemet/C0805C105K3RACTU/399-8004-1-ND/3471727

It looks like to get anything thicker you will need to bump up to the 1206 size. I expect you can get the 1 uF you want if you let the voltage rise. Use the Digikey search feature. It works great if you know how to use it.

--

Rick C.

- Get a 1,000 miles of free Supercharging
- Tesla referral code - https://ts.la/richard11209

 

On Friday, 17 May 2019 22:55:22 UTC+1, Winfield Hill wrote:

Many SMT components have maximum height specs,
but that can get one into trouble. My little
14-parameter center-of-the-beehive monitor,
has five sensor ICs with tiny holes, to let
air in for measurement. To keep bees from
covering sensor holes with waxy propolis, my
board will have a clear thermo-plastic cover.
The sensor heights are 0.8 to 0.9mm. So, to
keep the cover from covering the sensor holes,
I placed five 1.4mm tall parts on the PCB.
These should keep the plastic 0.4 to 0.6 mm
above the sensor holes.

AVX spec'd their 1uF, 25V 0805 caps at 1.4mm
tall, but, oops, they measure 0.77mm thick.
OK, Arrgh, the spec was for maximum height.
What we need is a minimum height spec!

Poking around in my inventory, I found TDK
C2012X7R1H225K125AC 2.2uF,50V 0805 caps,
that measured 1.34mm thick. TDK's spec is
nominal, 1.25 +/- 0.25 mm.

But if I use them, 3.3V-bus capacitance
will be 11uF, instead of 5uF. The sensor
power has a 50mA current limit, making a
11uF 3.3V / 50mA = 0.7ms turn-on risetime.

So don't connect some


NT

 

[George Herold]

On Friday, May 17, 2019 at 5:55:22 PM UTC-4, Winfield Hill wrote:

Many SMT components have maximum height specs,
but that can get one into trouble. My little
14-parameter center-of-the-beehive monitor,
has five sensor ICs with tiny holes, to let
air in for measurement. To keep bees from
covering sensor holes with waxy propolis, my
board will have a clear thermo-plastic cover.
The sensor heights are 0.8 to 0.9mm. So, to
keep the cover from covering the sensor holes,
I placed five 1.4mm tall parts on the PCB.
These should keep the plastic 0.4 to 0.6 mm
above the sensor holes.

AVX spec'd their 1uF, 25V 0805 caps at 1.4mm
tall, but, oops, they measure 0.77mm thick.
OK, Arrgh, the spec was for maximum height.
What we need is a minimum height spec!

Poking around in my inventory, I found TDK
C2012X7R1H225K125AC 2.2uF,50V 0805 caps,
that measured 1.34mm thick. TDK's spec is
nominal, 1.25 +/- 0.25 mm.

But if I use them, 3.3V-bus capacitance
will be 11uF, instead of 5uF. The sensor
power has a 50mA current limit, making a
11uF 3.3V / 50mA = 0.7ms turn-on risetime.


--
Thanks,
- Win

Hi Win, I have no idea. But bee season is here,
and it's time to let the bee's find all the mistakes in your design.
Bee-ta testing. :^)
George H.

 

[Winfield Hill]

tabbypurr@gmail.com wrote...

So don't connect some.

This is a good suggestion, find and add tall
parts, even if that's their only function.


--
Thanks,
- Win

 

[Winfield Hill]

Rick C wrote...

https://www.digikey.com/product-detail/en/vishay-dale/ILSB1206ER100K/541-26=
07-1-ND/5419048

1.4 mm tall and $0.14 at qty 100. Is that too expensive?

Inductors for height, that's a great idea.

If you really want a cap here is an 0805 at 1.25 =C2=B10.15 mm, so 1.10 mm =
min, 1.4 mm max.

https://www.digikey.com/product-detail/en/kemet/C0805C105K3RACTU/399-8004-1=
-ND/3471727

That;s a good find.


--
Thanks,
- Win

 

[Tom Del Rosso]

Winfield Hill wrote:

tabbypurr@gmail.com wrote...

So don't connect some.

This is a good suggestion, find and add tall
parts, even if that's their only function.

If there's room for dummy parts why not add an even cheaper spacer which
is available in exact sizes? Then performance would be more similar
among all units. Heat will conduct to the shield and radiate to the
sensors - inverse square.

 

[Tim Williams]

"Tom Del Rosso" <fizzbintuesday@that-google-mail-domain.com> wrote in
message news:qbp636$6aa$1@dont-email.me...

If there's room for dummy parts why not add an even cheaper spacer which
is available in exact sizes? Then performance would be more similar among
all units. Heat will conduct to the shield and radiate to the sensors -
inverse square.

Or cut-off chunks of PCB glued or soldered on; they could even be fabbed as
part of the build, on breakaway tabs.

On a related note, it's a powerful design and construction method, building
things by layer -- obviously, 3D printing is usually done this way, but it's
been around for a long time in mechanical builds. Some of the first
magnetrons for example were stacked punchings, furnace brazed. Heat
exchangers are typically stacked plates, alternating, and furnace brazed.
You can't fab just anything with a stack of plates -- each individual layer
preferably should be a connected shape -- but that can be worked around in
the 3rd dimension (if at some expense to layer count / design complexity, of
course).

So, you can do interesting mechanical, even pneumatic/hydraulic, things with
stacked PCBs, and judicious use of something to bond the layers together.

Tim

--
Seven Transistor Labs, LLC
Electrical Engineering Consultation and Design
Website: https://www.seventransistorlabs.com/

 

On Saturday, 18 May 2019 15:50:19 UTC+1, Tom Del Rosso wrote:

Winfield Hill wrote:
tabbypurr wrote...

So don't connect some.

This is a good suggestion, find and add tall
parts, even if that's their only function.

If there's room for dummy parts why not add an even cheaper spacer which
is available in exact sizes? Then performance would be more similar
among all units. Heat will conduct to the shield and radiate to the
sensors - inverse square.

Presumably because it then requires another fabrication process, $$$. Makes adding another p&p part look cheap.


NT

 

[Tom Del Rosso]

tabbypurr@gmail.com wrote:

On Saturday, 18 May 2019 15:50:19 UTC+1, Tom Del Rosso wrote:
Winfield Hill wrote:
tabbypurr wrote...

So don't connect some.

This is a good suggestion, find and add tall
parts, even if that's their only function.

If there's room for dummy parts why not add an even cheaper spacer
which is available in exact sizes? Then performance would be more
similar among all units. Heat will conduct to the shield and
radiate to the sensors - inverse square.

Presumably because it then requires another fabrication process, $$$.
Makes adding another p&p part look cheap.

Well there has to be a step to attach the cover anyway. If components
are holding it away from the board then something else must hold it in
place.

 

[Rick C]

On Saturday, May 18, 2019 at 1:36:34 PM UTC-4, Tom Del Rosso wrote:

tabbypurr@gmail.com wrote:
On Saturday, 18 May 2019 15:50:19 UTC+1, Tom Del Rosso wrote:
Winfield Hill wrote:
tabbypurr wrote...

So don't connect some.

This is a good suggestion, find and add tall
parts, even if that's their only function.

If there's room for dummy parts why not add an even cheaper spacer
which is available in exact sizes? Then performance would be more
similar among all units. Heat will conduct to the shield and
radiate to the sensors - inverse square.

Presumably because it then requires another fabrication process, $$$.
Makes adding another p&p part look cheap.

Well there has to be a step to attach the cover anyway. If components
are holding it away from the board then something else must hold it in
place.

When he described the cover as thermo-plastic and the requirement for multiple spacers around the device, I assumed he was talking about something similar to shrink wrap that will be applied and heated to fit after the boards are fabbed and working.

--

Rick C.

-- Get a 1,000 miles of free Supercharging
-- Tesla referral code - https://ts.la/richard11209

 

[Rick C]

On Saturday, May 18, 2019 at 12:42:24 PM UTC-4, Tim Williams wrote:

"Tom Del Rosso" <fizzbintuesday@that-google-mail-domain.com> wrote in
message news:qbp636$6aa$1@dont-email.me...
If there's room for dummy parts why not add an even cheaper spacer which
is available in exact sizes? Then performance would be more similar among
all units. Heat will conduct to the shield and radiate to the sensors -
inverse square.


Or cut-off chunks of PCB glued or soldered on; they could even be fabbed as
part of the build, on breakaway tabs.

That's not a bad idea, but it may be a fair amount of handling. The spacer doesn't need to go through assembly, so doesn't need to be PCB material even. A plastic spacer could be used. The trick is to find something *very* economical to make. Perhaps a plastic washer if you have plenty of room on the board. Or a Nylon nut? This would be held in place by the shrink plastic.

--

Rick C.

+ Get 5,000 miles of free Supercharging
+ Tesla referral code - https://ts.la/richard11209

 

[Joe Chisolm]

On Fri, 17 May 2019 14:55:13 -0700, Winfield Hill wrote:

Many SMT components have maximum height specs,
but that can get one into trouble. My little 14-parameter
center-of-the-beehive monitor, has five sensor ICs with tiny holes, to
let air in for measurement. To keep bees from covering sensor holes
with waxy propolis, my board will have a clear thermo-plastic cover.
The sensor heights are 0.8 to 0.9mm. So, to keep the cover from
covering the sensor holes,
I placed five 1.4mm tall parts on the PCB. These should keep the
plastic 0.4 to 0.6 mm above the sensor holes.

AVX spec'd their 1uF, 25V 0805 caps at 1.4mm tall, but, oops, they
measure 0.77mm thick. OK, Arrgh, the spec was for maximum height. What
we need is a minimum height spec!

Poking around in my inventory, I found TDK C2012X7R1H225K125AC
2.2uF,50V 0805 caps,
that measured 1.34mm thick. TDK's spec is nominal, 1.25 +/- 0.25 mm.

But if I use them, 3.3V-bus capacitance will be 11uF, instead of 5uF.
The sensor power has a 50mA current limit, making a 11uF 3.3V / 50mA =
0.7ms turn-on risetime.

Might try these Mill-Max circular test points. The -3 is .055in.
Not cheap but if you only need a few per board might not too bad.

https://www.mill-max.com/products/pin/1578/1578-0-57-15-00-00-03-0

Digi-Key, Mouser and Heilind carry them.


--
Chisolm
Republic of Texas

 

[whit3rd]

On Friday, May 17, 2019 at 2:55:22 PM UTC-7, Winfield Hill wrote:

Many SMT components have maximum height specs,
but that can get one into trouble. My little
14-parameter center-of-the-beehive monitor,
has five sensor ICs with tiny holes, to let
air in for measurement. ...So, to
keep the cover from covering the sensor holes,
I placed five 1.4mm tall parts on the PCB.
These should keep the plastic 0.4 to 0.6 mm
above the sensor holes.

Usually, you'd see the plastic cover with punched holes, and matching holes in the
PCB, with little plastic standoff/snapins. For standoffs, there's adhesive bump items, too.

Oddly, surface mount inductors seem cheaper (unless I just don't know where to shop)
than plastic snaps.

 

you can do interesting mechanical,
even pneumatic/hydraulic, things with
stacked PCBs, and judicious use of
something to bond the layers together.

Hydraulics, with PCB? Never heard about that kind of usage before. Can you elaborate?

Regards

Klaus

 

[Jasen Betts]

On 2019-05-18, Winfield Hill <hill@rowland.harvard.edu> wrote:

tabbypurr@gmail.com wrote...

So don't connect some.

This is a good suggestion, find and add tall
parts, even if that's their only function.

Use resistors, but only put solder at one end

the trick would be getting the right amount of solder to promote a 90
degree tombstone perhaps a lifting pad under the belly
of the resistor would help.


--
When I tried casting out nines I made a hash of it.

 

On Tuesday, 21 May 2019 07:31:02 UTC+1, Jasen Betts wrote:

On 2019-05-18, Winfield Hill <hill@rowland.harvard.edu> wrote:
tabbypurr wrote...

So don't connect some.

This is a good suggestion, find and add tall
parts, even if that's their only function.

Use resistors, but only put solder at one end

the trick would be getting the right amount of solder to promote a 90
degree tombstone perhaps a lifting pad under the belly
of the resistor would help.

cute idea but too unreliable.

 

Jasen Betts <jasen@xnet.co.nz> wrote in news:qc05cp$8ls$2
@gonzo.revmaps.no-ip.org:

On 2019-05-18, Winfield Hill <hill@rowland.harvard.edu> wrote:
tabbypurr@gmail.com wrote...

So don't connect some.

This is a good suggestion, find and add tall
parts, even if that's their only function.

Use resistors, but only put solder at one end

A stupid suggestion.

the trick would be getting the right amount of solder to promote a
90
degree tombstone perhaps a lifting pad under the belly
of the resistor would help.

Yeah, I am sure the pick and place machine would love you.

Very weak termination on a resistor, especially mounted that way.
A very poor idea.

Hell even a stack of three resistors on top of each other but
connected by both terminations would be better than that

The capacitors are the best choice (for this manner). Two legs
for attachment and as broad and tall as one desires without labor
intensive hand tombstoning of resistors and the part price is also a
non issue.

Slots and thin PCB 'rails' can also be put onto the assembly main
PCB and act as stand offs.

 

[whit3rd]

On Friday, May 17, 2019 at 2:55:22 PM UTC-7, Winfield Hill wrote:

To keep bees from
covering sensor holes with waxy propolis, my
board will have a clear thermo-plastic cover.
The sensor heights are 0.8 to 0.9mm. So, to
keep the cover from covering the sensor holes,
I placed five 1.4mm tall parts on the PCB.

A variant solution would be to mount the plastic cover with
double-sticky foam tape; 3M sells it, with permanent adhesive,
in a variety of thicknesses. 1.4 mm is circa 1/16 inch.

<https://www.parts-express.com/3m-foam-mount-tape-1-2-x-75-roll--082-180>

 

[Winfield Hill]

Winfield Hill wrote...

Rick C wrote...

If you really want a cap here is an 0805 at 1.25mm,
so 1.10 mm min, 1.4 mm max.

https://www.digikey.com/product-detail/en/kemet/C0805C105K3RACTU/399-8004-1=
-ND/3471727

That's a good find.

Kemet 1uF 25V X7R arrived, measure 1.24mm,
very nice! Off to the assembly house.

Thanks, Rick!


--
Thanks,
- Win