Pencil butane torch for Surface Mount desolder?

[N_Cook]

I've only used full size hot air gun and protective masking-off before. I
prefer the short sharp shock of full size rather than the "proper" pulse jet
type things, presumably for rework only really.

Couldn't resist one in a pound shop, unbelievably precisely small spot of
flame and also adjustable (upwards), amazing engineering, minimum about 1mm
diameter.
For SM salvaging rather than just removal, any other tips from actual users?
Seems necessary to hold the pcb vertical in a vice and hold the pencil
with flame upwards about 20 degrees of angle so butane liquid is not at the
outlet.
Mask off surrounding areas (including body of IC) with PTFE or woven glass
matt cut to fit the paricular situation.
Sub mm tungsten wire or something strong enough in those dimensions to
push under the IC or the pins , to loop around, and tug on while heating.
Freezer spray blast just prior to heating pins.

For DIP do one side of pins at a time, while tugging ?
Scalpel blade , perhaps bent, under pins ?
What sort of lighting is best ie to see the flame but also the work?

 

[William Sommerwerck]

I've only used full size hot air gun and protective masking-off.

I prefer the short sharp shock of full size rather than the "proper"
pulse jet type things, presumably for rework only really.

Don't stint on the removal tool -- a cheap and chippy chopper is not a good
choice. You also need a big black block for proper support.

 

[Winston]

I've only used full size hot air gun and protective masking-off before. I
prefer the short sharp shock of full size rather than the "proper" pulse jet
type things, presumably for rework only really.

Couldn't resist one in a pound shop, unbelievably precisely small spot of
flame and also adjustable (upwards), amazing engineering, minimum about 1mm
diameter.
For SM salvaging rather than just removal, any other tips from actual users?

That torch produces plenty of temperature but not nearly enough heat.

I use two hot air guns, one aimed at the back of the board and one
on the front. (The copper ground planes act as heat sinks, so you need
to pour a bunch of thermal energy into the area in question in order
to get sufficient heat rise to melt solder.)

I normally heat up the back of the board before I even start the
'front side' air gun. Two guns are more convenient than one.
Most of the time however, I can use the same gun for both.
Heat the back thoroughly and then concentrate on the front.

I dab a little solder paste in the area. That is my local
'thermometer'. When it melts, I know I'm only seconds away
from 'liftoff'.

Strip some 30 ga. wirewrap wire to route underneath BGAs.
You can form a cradle to gently lift the device off the board.


--Winston

 

[N_Cook]

users?

That torch produces plenty of temperature but not nearly enough heat.

I use two hot air guns, one aimed at the back of the board and one
on the front. (The copper ground planes act as heat sinks, so you need
to pour a bunch of thermal energy into the area in question in order
to get sufficient heat rise to melt solder.)

I normally heat up the back of the board before I even start the
'front side' air gun. Two guns are more convenient than one.
Most of the time however, I can use the same gun for both.
Heat the back thoroughly and then concentrate on the front.

I dab a little solder paste in the area. That is my local
'thermometer'. When it melts, I know I'm only seconds away
from 'liftoff'.

Strip some 30 ga. wirewrap wire to route underneath BGAs.
You can form a cradle to gently lift the device off the board.


--Winston

Some interesting ideas there. I don't often meet ground planes but if I do ,
how do you gauge the rear heat ?
I like the solder paste indicator.

 

[N_Cook]

I've only used full size hot air gun and protective masking-off before. I
prefer the short sharp shock of full size rather than the "proper" pulse jet
type things, presumably for rework only really.

Couldn't resist one in a pound shop, unbelievably precisely small spot of
flame and also adjustable (upwards), amazing engineering, minimum about 1mm
diameter.
For SM salvaging rather than just removal, any other tips from actual users?
Seems necessary to hold the pcb vertical in a vice and hold the pencil
with flame upwards about 20 degrees of angle so butane liquid is not at the
outlet.
Mask off surrounding areas (including body of IC) with PTFE or woven glass
matt cut to fit the paricular situation.
Sub mm tungsten wire or something strong enough in those dimensions to
push under the IC or the pins , to loop around, and tug on while heating.
Freezer spray blast just prior to heating pins.

For DIP do one side of pins at a time, while tugging ?
Scalpel blade , perhaps bent, under pins ?
What sort of lighting is best ie to see the flame but also the work?

Do you have a particular rationale for freeze blasting prior to
rework? I think the resulting thermal shock would increase the
likelihood of damage. Reflow ovens have temperature ramps for a
reason. As long as you aren't charring the numbers off the package
with a bernz-o-matic until it falls off the board, you are safe
heating the entire device to reflow temp. That's how it was put on in
the first place. Just do it uniformly and gradually.

If you are serious about salvaging QFP/TSOP/TSSOP devices, you might
want to look into a regular size heat gun with a nozzle designed for
these packages. An OKi is about 500 bucks but you can get a cheap
chinese knockoff for under 200. If you really want to do it quickly
and safely I'd recommend getting a solder tip for each package. But
this could run you anywhere from 200-1500 depending on what you've got
to start with.

But if you're like me and you are salvaging chips because you don't
have the cash to buy new ones, then I suggest you look into chip-quik.
It's low-temperature solder you "blob" over the leads with an ordinary
iron. It stays liquid for 10-30 seconds so you can do a few laps
around the device with your pencil iron, then grab the device with
forceps and pluck it right off the board. You can set your iron a lot
lower if you have an adjustment.

I used to use that tungsten wire trick (steel or ordinary enameled
magnet wire works great too) but chip quik is much faster and less
traumatic to the device.

I routinely use chip quik these days to upgrade 80-pin
microcontrollers. I am not salvaging them but the lower heat causes
less damage to the board, and that saves me trouble later.

OOOOOOOOO

I'm probably to old to change my ways now unless a proven advantage to do
so. I like to go in and out as quick as possible with hot air source having
heated up for 30 seconds before applying. No known failures in reuse from
pre-freezing. I like the way they come off in their 0.1mm tugsten wire
strops and they can be placed in the palm of the hand, IC body face down,
and be just warm , not hot.

 

[mike]

I've only used full size hot air gun and protective masking-off before. I
prefer the short sharp shock of full size rather than the "proper" pulse jet
type things, presumably for rework only really.

Couldn't resist one in a pound shop, unbelievably precisely small spot of
flame and also adjustable (upwards), amazing engineering, minimum about 1mm
diameter.
For SM salvaging rather than just removal, any other tips from actual users?
Seems necessary to hold the pcb vertical in a vice and hold the pencil
with flame upwards about 20 degrees of angle so butane liquid is not at the
outlet.
Mask off surrounding areas (including body of IC) with PTFE or woven glass
matt cut to fit the paricular situation.
Sub mm tungsten wire or something strong enough in those dimensions to
push under the IC or the pins , to loop around, and tug on while heating.
Freezer spray blast just prior to heating pins.

For DIP do one side of pins at a time, while tugging ?
Scalpel blade , perhaps bent, under pins ?
What sort of lighting is best ie to see the flame but also the work?

You're on the right track.
Get a pencil torch with a catalytic "hot air" tip.
I use a Weller PortaSol.
MUCH better ratio of temperature to heat delivered.
Works way better than I'd expected. Preheting the
board closer to solder temperature with a big hot air
gun or temperature controlled hot plate makes it even easier.

 

[Winston]

On 5/27/2010 7:42 AM, N_Cook wrote:

(...)

I don't often meet ground planes but if I do ,
how do you gauge the rear heat ?

I have not been measuring the temperature at all.
I've been relying on experience to know that I've heated up the back
side of the board 'sufficiently' but not so much as to loosen
components.

I've been eyeing these Tempilstiks to get a better idea:
http://www.use-enco.com/CGI/INPDFF?PMPAGE=1130&PARTPG=INLMK32

One could mark the board with three crayons of increasing
melting temperature and get a better idea when to start
pulling on the BGA.

I like the solder paste indicator.

It works 'better than nothing'.
Of course the virgin solder melts at a much lower temperature
than does the used solder.

--Winston

 

[N_Cook]

On 5/27/2010 7:42 AM, N_Cook wrote:

(...)


I have not been measuring the temperature at all.
I've been relying on experience to know that I've heated up the back
side of the board 'sufficiently' but not so much as to loosen
components.

I've been eyeing these Tempilstiks to get a better idea:
http://www.use-enco.com/CGI/INPDFF?PMPAGE=1130&PARTPG=INLMK32

One could mark the board with three crayons of increasing
melting temperature and get a better idea when to start
pulling on the BGA.



It works 'better than nothing'.
Of course the virgin solder melts at a much lower temperature
than does the used solder.

--Winston

Another reason to investigate these , but 1/2 deg C resolution fine for that
purpose. For general fault hotspot monitoring 0.1 deg C resolution ones
would be more useful but too expensive. Another gizmo I must get around to
trying. Using the pyrometer of a PIR unit , setting in an insulated tube
shroud for pointing and then analogue fashion monitor the output in a
relative way, absolute temp not required for that purpose.

 

[N_Cook]

Another reason to investigate these , but 1/2 deg C resolution fine for that
purpose. For general fault hotspot monitoring 0.1 deg C resolution ones
would be more useful but too expensive. Another gizmo I must get around to
trying. Using the pyrometer of a PIR unit , setting in an insulated tube
shroud for pointing and then analogue fashion monitor the output in a
relative way, absolute temp not required for that purpose.

In this situation would thore remote IR thermometers pick up the temp of the
board or the much higher temp of the hot air above it , but in line of
"sight" of the sensor?

 

[Winston]

(...)

I dunno. Plus / minus 10 C is all that is necessary.
Your real indication is when you can lift the BGA off the board
without undue drama.

I understand that PIRs are A.C. devices, so one would have to be adept
at 'blinking' the shroud at the target while capturing the resulting
pulse from the PIR. As you say below, there is a much better solution.

In this situation would thore remote IR thermometers pick up the temp of the
board or the much higher temp of the hot air above it , but in line of
"sight" of the sensor?

That is a good experiment to do.

I *guess* that the infrared emissivity of nitrogen *at that temperature*
is low enough that the infrared from the PCB would swamp it out.
I do not know for sure.

Product Idea! An infrared sensor mounted to your headband magnifier.

Lensed LCD chart presented as a heads-up display in your magnifier
continuously scrolls to show temperature at the center of your crosshairs.

--Winston

 

[N_Cook]

I dunno. Plus / minus 10 C is all that is necessary.
Your real indication is when you can lift the BGA off the board
without undue drama.


I understand that PIRs are A.C. devices, so one would have to be adept
at 'blinking' the shroud at the target while capturing the resulting
pulse from the PIR. As you say below, there is a much better solution.


That is a good experiment to do.

I *guess* that the infrared emissivity of nitrogen *at that temperature*
is low enough that the infrared from the PCB would swamp it out.
I do not know for sure.

Product Idea! An infrared sensor mounted to your headband magnifier.

Lensed LCD chart presented as a heads-up display in your magnifier
continuously scrolls to show temperature at the center of your crosshairs.

--Winston

The "ac" action is from the effect of the facetted mirrors in the PIR,
producing signal change when a moving warm object moves across the
boundaries of facet projected areas , a bit like fly's multi-lens eye
structured as a movement sensor

 

[Winston]

On 5/28/2010 9:03 AM, N_Cook wrote:

(...)

The "ac" action is from the effect of the facetted mirrors in the PIR,
producing signal change when a moving warm object moves across the
boundaries of facet projected areas , a bit like fly's multi-lens eye
structured as a movement sensor

The way I misunderstand it is that the raw output of the sensor
itself is A.C. With or without the faceted lens, a hot object
presented to the sensor itself will result in a pulse, not
a D.C. offset as one expects of a Si photocell, for example.

I'm here to learn however, so I am open to new information.



--Winston

 

[Grant]

I dunno. Plus / minus 10 C is all that is necessary.
Your real indication is when you can lift the BGA off the board
without undue drama.

I tap the board to encourage components to 'unstick' from the molten
solder.

I understand that PIRs are A.C. devices, so one would have to be adept
at 'blinking' the shroud at the target while capturing the resulting
pulse from the PIR. As you say below, there is a much better solution.

Yeah, PIR detector is hopeless at absolute temperature. What it does
instead is rely on the warm body moving in and out of the lensed zones
to create an AC IR signal. That's why they have that fresnel lens thing
on them to define the IR zones.

That is a good experiment to do.

I *guess* that the infrared emissivity of nitrogen *at that temperature*
is low enough that the infrared from the PCB would swamp it out.
I do not know for sure.

Product Idea! An infrared sensor mounted to your headband magnifier.

Lensed LCD chart presented as a heads-up display in your magnifier
continuously scrolls to show temperature at the center of your crosshairs.

Hmm, seen the price of a point-at-target IR thermometer? Add a couple
more times that for the headsup display? Not cheap

Grant.

 

[Winston]

On 5/28/2010 7:48 PM, Grant wrote:

(....)

Hmm, seen the price of a point-at-target IR thermometer? Add a couple
more times that for the headsup display? Not cheap

http://www.dealextreme.com/details.dx/sku.29079~r.24492410

Thirty smackers? Looks hackable too.

--Winston

 

[Grant]

On 5/28/2010 7:48 PM, Grant wrote:

(....)


http://www.dealextreme.com/details.dx/sku.29079~r.24492410

Thirty smackers? Looks hackable too.

Nice. I was thinking more of the 'brand name' pricing, 10 or more times
that

Grant.

 

[Winston]

Nice. I was thinking more of the 'brand name' pricing, 10 or more times
that

The uSoft version *is* pricey.

http://iml.jou.ufl.edu/projects/students/purdy/Gates_of_Borg.jpg

--Winston