Removing Large Electrolytics

[Cursitor Doom]

I have just spent about 40m trying to unsolder an electro on the board of
the Tek 466 I'm currently working on. It showed a ESR of 45 ohms so I
needed to hook it out of circuit just to be doubly sure it was really
that bad.
The electros in question are all fairly large ones in the (linear) PSU
section. For some reason they've not used separate boards for the various
sub-circuits, so the PSU is just one area of a v.large board which also
has components for all sorts of other functions. Anyway, The cap in
question has 5 connections to the board: 3 equi-spaced ground tabs around
the outside that come straight from the case of the cap and the 2 + and -
wires close to the centre. Given 5 in total through-hole connections,
it's proving very difficult to remove the cap using suction pump and
solder braid cos I cannot wiggle it at all. The best solution IMO would
be to heat all 5 tabs at the same time whilst pulling the cap's case from
the other side until it breaks free. Is there any tool that enables you
to do this? I fear the amount of fucking around I'm having to do
otherwise will lead to delamination of the PCB traces and an ugly mess.
I cannot even slip a junior hacksaw blade between the bottom of the cap
and the PCB and cut it free cos it's obscured by other components.



--
This message may be freely reproduced without limit or charge only via
the Usenet protocol. Reproduction in whole or part through other
protocols, whether for profit or not, is conditional upon a charge of
GBP10.00 per reproduction. Publication in this manner via non-Usenet
protocols constitutes acceptance of this condition.

 

[jfeng@my-deja.com]

Some people use a butane torch to heat the solder joints gently while pulling on the component. Practice this technique before you do it on something you want to save. I have done this for DIP ICs.

Since you are pretty sure it is bad, I would clip the leads, destroying the can beforehand if necessary.

 

On Wednesday, 22 August 2018 01:18:33 UTC+1, jf...@my-deja.com wrote:

Some people use a butane torch to heat the solder joints gently while pulling on the component. Practice this technique before you do it on something you want to save. I have done this for DIP ICs.

Since you are pretty sure it is bad, I would clip the leads, destroying the can beforehand if necessary.

A torch is a recipe for a badly burnt board. I've always been able to move a multi-legged component at least a tiny amount while melting one or more legs, and rock it back & forth, letting the board cool well if paper. The pcb is slightly flexible after all. Paper boards are weak, vulnerable to heat & the copper comes unglued when hot.

Maybe you could cut the 3 outer tabs off flush.


NT

 

[Phil Hobbs]

On 08/21/2018 07:46 PM, Cursitor Doom wrote:

I have just spent about 40m trying to unsolder an electro on the board of
the Tek 466 I'm currently working on. It showed a ESR of 45 ohms so I
needed to hook it out of circuit just to be doubly sure it was really
that bad.
The electros in question are all fairly large ones in the (linear) PSU
section. For some reason they've not used separate boards for the various
sub-circuits, so the PSU is just one area of a v.large board which also
has components for all sorts of other functions. Anyway, The cap in
question has 5 connections to the board: 3 equi-spaced ground tabs around
the outside that come straight from the case of the cap and the 2 + and -
wires close to the centre. Given 5 in total through-hole connections,
it's proving very difficult to remove the cap using suction pump and
solder braid cos I cannot wiggle it at all. The best solution IMO would
be to heat all 5 tabs at the same time whilst pulling the cap's case from
the other side until it breaks free. Is there any tool that enables you
to do this? I fear the amount of fucking around I'm having to do
otherwise will lead to delamination of the PCB traces and an ugly mess.
I cannot even slip a junior hacksaw blade between the bottom of the cap
and the PCB and cut it free cos it's obscured by other components.

+1 for can-opening it to save wear and tear on the board. If you
measure 45 ohms ESR in-circuit, it won't be better by itself.

Cheers

Phil Hobbs

 

[Terry Schwartz]

On Tuesday, August 21, 2018 at 6:46:24 PM UTC-5, Cursitor Doom wrote:

I have just spent about 40m trying to unsolder an electro on the board of
the Tek 466 I'm currently working on. It showed a ESR of 45 ohms so I
needed to hook it out of circuit just to be doubly sure it was really
that bad.
The electros in question are all fairly large ones in the (linear) PSU
section. For some reason they've not used separate boards for the various
sub-circuits, so the PSU is just one area of a v.large board which also
has components for all sorts of other functions. Anyway, The cap in
question has 5 connections to the board: 3 equi-spaced ground tabs around
the outside that come straight from the case of the cap and the 2 + and -
wires close to the centre. Given 5 in total through-hole connections,
it's proving very difficult to remove the cap using suction pump and
solder braid cos I cannot wiggle it at all. The best solution IMO would
be to heat all 5 tabs at the same time whilst pulling the cap's case from
the other side until it breaks free. Is there any tool that enables you
to do this? I fear the amount of fucking around I'm having to do
otherwise will lead to delamination of the PCB traces and an ugly mess.
I cannot even slip a junior hacksaw blade between the bottom of the cap
and the PCB and cut it free cos it's obscured by other components.



--
This message may be freely reproduced without limit or charge only via
the Usenet protocol. Reproduction in whole or part through other
protocols, whether for profit or not, is conditional upon a charge of
GBP10.00 per reproduction. Publication in this manner via non-Usenet
protocols constitutes acceptance of this condition.

Depending on what else is nearby, a heat gun can be effective.

 

I use a Metcal which is a 40 Watt iron very well controlled but when removing caps from computer motherboards where there are no thermal reliefs to the ground plane, NO soldering iron can unsolder it alone. The pin can't conduct the heat fast enough. I also have a cheap Chinese hot air tool. Set the temp on the air tool slightly below the melt temp so it won't 'blow' any SMD parts (none on the 466) off the board. This becomes a 2 handed operation.. While heating with the Metcal, I 'auxiliary' heat the board with the hot air tool for about 20 seconds. This WILL get hot enough to use your Edsyn Soldapullt to clear the hole and not lift any pads or damage the board.

Cutting up the cap and then removing the separate pins is OK IF you can destroy it without damaging the board. I've had excellent results with the aux heating technique which is faster than dismembering the part.

Torches are WAY too brutal especially in a Tek.

G²

 

[Cursitor Doom]

On Tue, 21 Aug 2018 21:28:10 -0400, Phil Hobbs wrote:

+1 for can-opening it to save wear and tear on the board. If you
measure 45 ohms ESR in-circuit, it won't be better by itself.

Well, it's hardly surprising for an electro of this age. Clearly the
original component from a scope with a s/n indicating it was manufactured
44 years ago! What was more surprising was that only 2 out of 6 electros
in the PSU section showed abnormal readings (the remaining faulty one has
gone leaky).




--
This message may be freely reproduced without limit or charge only via
the Usenet protocol. Reproduction in whole or part through other
protocols, whether for profit or not, is conditional upon a charge of
GBP10.00 per reproduction. Publication in this manner via non-Usenet
protocols constitutes acceptance of this condition.

 

On Wednesday, 22 August 2018 11:12:10 UTC+1, Cursitor Doom wrote:

On Tue, 21 Aug 2018 21:28:10 -0400, Phil Hobbs wrote:

+1 for can-opening it to save wear and tear on the board. If you
measure 45 ohms ESR in-circuit, it won't be better by itself.

Well, it's hardly surprising for an electro of this age. Clearly the
original component from a scope with a s/n indicating it was manufactured
44 years ago! What was more surprising was that only 2 out of 6 electros
in the PSU section showed abnormal readings (the remaining faulty one has
gone leaky).

Sounds about par for the course. Go back further to paper caps and it'd be a surprise to find any still working properly. And yes, micamold were paper caps.


NT

 

On Tuesday, August 21, 2018 at 9:28:19 PM UTC-4, Phil Hobbs wrote:

On 08/21/2018 07:46 PM, Cursitor Doom wrote:
I have just spent about 40m trying to unsolder an electro on the board of
the Tek 466 I'm currently working on. It showed a ESR of 45 ohms so I
needed to hook it out of circuit just to be doubly sure it was really
that bad.
The electros in question are all fairly large ones in the (linear) PSU
section. For some reason they've not used separate boards for the various
sub-circuits, so the PSU is just one area of a v.large board which also
has components for all sorts of other functions. Anyway, The cap in
question has 5 connections to the board: 3 equi-spaced ground tabs around
the outside that come straight from the case of the cap and the 2 + and -
wires close to the centre. Given 5 in total through-hole connections,
it's proving very difficult to remove the cap using suction pump and
solder braid cos I cannot wiggle it at all. The best solution IMO would
be to heat all 5 tabs at the same time whilst pulling the cap's case from
the other side until it breaks free. Is there any tool that enables you
to do this? I fear the amount of fucking around I'm having to do
otherwise will lead to delamination of the PCB traces and an ugly mess.
I cannot even slip a junior hacksaw blade between the bottom of the cap
and the PCB and cut it free cos it's obscured by other components.




+1 for can-opening it to save wear and tear on the board. If you
measure 45 ohms ESR in-circuit, it won't be better by itself.

Cheers

Phil Hobbs

Right I'd hack up the component and pull pin by pin.
I do use a hot air rework station on some rotary switches...
it's faster than all that hacking.

George H.

 

[Cursitor Doom]

On Tue, 21 Aug 2018 21:54:12 -0700, stratus46 wrote:

I use a Metcal which is a 40 Watt iron very well controlled but when
removing caps from computer motherboards where there are no thermal
reliefs to the ground plane, NO soldering iron can unsolder it alone.

You're not kidding. My largest iron is 40W/80W switchable. 40W won't
touch it and I'm not risking 80W! The biggest problem seems to be the cap
case - a long 1" dia aluminium tube with 3 tags at the bottom going
straight to PCB grounds. It's acting like a very effective heat sink.
Going to have to trash the caps from above.





--
This message may be freely reproduced without limit or charge only via
the Usenet protocol. Reproduction in whole or part through other
protocols, whether for profit or not, is conditional upon a charge of
GBP10.00 per reproduction. Publication in this manner via non-Usenet
protocols constitutes acceptance of this condition.

 

[John-Del]

On Tuesday, August 21, 2018 at 7:46:24 PM UTC-4, Cursitor Doom wrote:

I have just spent about 40m trying to unsolder an electro on the board of
the Tek 466 I'm currently working on. It showed a ESR of 45 ohms so I
needed to hook it out of circuit just to be doubly sure it was really
that bad.
The electros in question are all fairly large ones in the (linear) PSU
section. For some reason they've not used separate boards for the various
sub-circuits, so the PSU is just one area of a v.large board which also
has components for all sorts of other functions. Anyway, The cap in
question has 5 connections to the board: 3 equi-spaced ground tabs around
the outside that come straight from the case of the cap and the 2 + and -
wires close to the centre. Given 5 in total through-hole connections,
it's proving very difficult to remove the cap using suction pump and
solder braid cos I cannot wiggle it at all. The best solution IMO would
be to heat all 5 tabs at the same time whilst pulling the cap's case from
the other side until it breaks free. Is there any tool that enables you
to do this? I fear the amount of fucking around I'm having to do
otherwise will lead to delamination of the PCB traces and an ugly mess.
I cannot even slip a junior hacksaw blade between the bottom of the cap
and the PCB and cut it free cos it's obscured by other components.

You need to preheat the board. BTW, rocking a component while desoldering is a recipe for disaster if the board in question is a multi-layer. An internal foil can be separated from the plated-through hole if physical assertion is used.

If you can't remove the board to put it in a pre-heater, aim a hair dryer at it for about 10-15 minutes (a real heat gun can blister it). When the board is good and hot, it will have far less propensity to sink off heat from your soldering iron tip.

 

[Ron D.]

Once you remove the solder "ADD MORE". Yep, add more. Suck that out.

It has the tendency to heat other areas.

You can also add a solder removal alloy form Chip-Quik.

You can break off the tabs sometimes. You can cut the can sometimes. I'd use these methods at last resort.

Solder with a low temperature allow after removing most of the solder and suck out is my best suggestion.

 

[Phil Hobbs]

On 08/22/2018 06:12 AM, Cursitor Doom wrote:

On Tue, 21 Aug 2018 21:28:10 -0400, Phil Hobbs wrote:

+1 for can-opening it to save wear and tear on the board. If you
measure 45 ohms ESR in-circuit, it won't be better by itself.

Well, it's hardly surprising for an electro of this age. Clearly the
original component from a scope with a s/n indicating it was manufactured
44 years ago! What was more surprising was that only 2 out of 6 electros
in the PSU section showed abnormal readings (the remaining faulty one has
gone leaky).

I have a Keithley 410 Micro-Microammeter from about 1960 that has all
its original electros and still works fine. Of course it uses an
electrometer tube, so it takes a good couple of hours' warm-up to settle
down on the 100-fA FS range, but it eventually does.

It has way more soul than my 610C(*), but really isn't nearly as good a
meter. (Don't anybody tell it, though--it has an honoured place on my
bench shelf.)

Cheers

Phil Hobbs

(*) "Now Nortons and Indians and Greeveses won't do...They don't have a
soul like a Vincent '52." -- Richard Thompson

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[Baron]

Cursitor Doom prodded the keyboard with:

On Tue, 21 Aug 2018 21:54:12 -0700, stratus46 wrote:

I use a Metcal which is a 40 Watt iron very well controlled but
when removing caps from computer motherboards where there are no
thermal reliefs to the ground plane, NO soldering iron can unsolder
it alone.

You're not kidding. My largest iron is 40W/80W switchable. 40W won't
touch it and I'm not risking 80W! The biggest problem seems to be
the cap case - a long 1" dia aluminium tube with 3 tags at the
bottom going straight to PCB grounds. It's acting like a very
effective heat sink. Going to have to trash the caps from above.

Have you tried making a soldering bit with a piece of split copper
tube and a cooks torch to heat it. Apply to all the pins up at once.

--
Best Regards:
Baron.

 

[jfeng@my-deja.com]

Some styles of electrolytics may be particularly difficult to remove gracefully. The ground tabs of the FP style were often twisted for mechanical stability before soldering. And sometimes, the pins were folded over before soldering.

 

On Friday, 24 August 2018 02:24:43 UTC+1, jf...@my-deja.com wrote:

> Some styles of electrolytics may be particularly difficult to remove gracefully. The ground tabs of the FP style were often twisted for mechanical stability before soldering. And sometimes, the pins were folded over before soldering.

If enough solder is removed, twisted tabs can then be twisted back. A bit of remaining solder can be ignored, it lacks enough strength to get in the way. Folded over legs can be pushed up with a soldering iron before attempting removal.


NT

 

[Phil Hobbs]

On 08/24/2018 04:27 AM, tabbypurr@gmail.com wrote:

On Friday, 24 August 2018 02:24:43 UTC+1, jf...@my-deja.com wrote:

Some styles of electrolytics may be particularly difficult to remove gracefully. The ground tabs of the FP style were often twisted for mechanical stability before soldering. And sometimes, the pins were folded over before soldering.

If enough solder is removed, twisted tabs can then be twisted back. A bit of remaining solder can be ignored, it lacks enough strength to get in the way. Folded over legs can be pushed up with a soldering iron before attempting removal.


NT

This might be a good use for Chip Quik--it's a bismuth-based solder that
drops the liquidus point of a joint low enough that gentle use of a heat
gun will melt it.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[Cursitor Doom]

On Fri, 24 Aug 2018 13:18:02 -0400, Phil Hobbs wrote:

This might be a good use for Chip Quik--it's a bismuth-based solder that
drops the liquidus point of a joint low enough that gentle use of a heat
gun will melt it.

I had to use one of these so-called multi-tools to slice through the
pins. There was about 2mm of clearance between the bottom of the caps and
the PCB. This is one of those cutters that vibrates from side to side at
high speed and the blade had *just* enough reach to cut through the most
inaccessible pins. Double checking after removal confirmed they had all
definitely failed (3 out of 6 altogether).





--
This message may be freely reproduced without limit or charge only via
the Usenet protocol. Reproduction in whole or part through other
protocols, whether for profit or not, is conditional upon a charge of
GBP10.00 per reproduction. Publication in this manner via non-Usenet
protocols constitutes acceptance of this condition.

 

[Cursitor Doom]

I had 3 faulty electrolytics; all 30V/5500uF. In my spares bin I found 6
NOS (new old stock) electros of values 35V/3300uF which when stacked 2
high and wired in parallel occupy - fortunately - the same footprint as
the ones that failed.
These spares, though unused, are probably 40 years old. Will they need re-
forming gradually over 24hrs before installing?




--
This message may be freely reproduced without limit or charge only via
the Usenet protocol. Reproduction in whole or part through other
protocols, whether for profit or not, is conditional upon a charge of
GBP10.00 per reproduction. Publication in this manner via non-Usenet
protocols constitutes acceptance of this condition.

 

On Saturday, 25 August 2018 00:42:23 UTC+1, Cursitor Doom wrote:

I had 3 faulty electrolytics; all 30V/5500uF. In my spares bin I found 6
NOS (new old stock) electros of values 35V/3300uF which when stacked 2
high and wired in parallel occupy - fortunately - the same footprint as
the ones that failed.
These spares, though unused, are probably 40 years old. Will they need re-
forming gradually over 24hrs before installing?

I'd get newer ones if possible. At that age they'll be dead or near to it. In apps where high ESR is acceptable they last better, but still 40 is 80 human years for lytics.


NT