[Coco] Sort of OT: Recapping

[Rick Ulland]

Glad to hear you got it fixed!

Speaking of longer leads, anyone who tries this trick should note that 
physically attaching the 'remote' part to the board with glue or ??? is 
important. If you don't do that the fan vibes will eventually break your 
tack weld.


On 11/29/20 5:23 PM, Salvador Garcia via Coco wrote:
>   Thanks. Believe me, I considered doing just that, piggybacking the new cap on top of the old one's leads. The only thing I didn't like about this is that the cap's leads would be much longer and I didn't want them to act as antennas, especially with components working at giga Hertz speeds nearby.
> Salvador
>
>
>      On Saturday, November 28, 2020, 05:33:47 PM CST, Rick Ulland <rickulland1 at gmail.com> wrote:
>   
>   I've got to admit, without tools I would probably just stick a new cap
> to the board with hotsnot and tack it to the remains of the old leads....
>
>
>
> On 11/28/20 2:38 PM, Richard Lorbieski via Coco wrote:
>> The problem is not with the copper thickness or amount of layers. Most
>> IBM retro motherboards (1990 era) are 6 or more layers.
>>
>> The three most common problems that occur with DIP multi-layer rework:
>>
>> 1) Failure to get the lead(s) out of the thru hole ground/power plane.
>>
>> 2) Old solder clogged inside the thru-hole of via.
>>
>> 3) Damage to the thru-hole plating (most catastrophic).
>>
>> I discourage the use of sharp objects or using brute force can lead to
>> damage the plating and thus cut the connection between internal and
>> surface pads of the PCB.
>>
>>
>> On 11/28/2020 6:55 AM, Dave Philipsen wrote:
>>>> On Nov 27, 2020, at 11:09 PM, Richard Lorbieski via Coco
>>>> <coco at maltedmedia.com> wrote:
>>>>
>>>> Based on the photos, this is a multi layer IBM clone PCB with
>>>> possible internal ground and power planes.
>>>>
>>>> The problem is that the heat is being dispersed onto the ground (or
>>>> power) plane and not enough heat is being applied to the solder.
>>>>
>>>> This kind of rework requires more than a 70W soldering iron.
>>>>
>>>> You will need a rework station with a heat gun and possibly a
>>>> variable convection type oven.  A conventional oven could damage
>>>> components on the PCB or melt some of the plastic parts. You will
>>>> also need a soldering iron that is capable of reaching 900F and use
>>>> a proper solder tip. Most consumer soldering irons usually reach
>>>> 600F-650F.
>>> You could even use a larger Weller “soldering gun” (you know, the one
>>> that actually looks like a gun!) to heat it up enough (for the
>>> weekend hobbyist who doesn’t do enough of this sort of work to
>>> justify the cost of a hot-air rework system or an oven). Keep a
>>> slight pull on the cap lead while heating the joint. As soon as the
>>> solder melts you pull the lead out and remove the heat of the gun. On
>>> radial lead parts like those in the photo you may have to alternate
>>> back and forth between the two leads a couple of times to “rock” it out.
>>>
>>> Practice on a junk board first until you feel comfortable with it.
>>>
>>>> I also recommend NOT to sharp knives or force the leads out of the
>>>> PCB. This could damage the internal connections on the PCB.
>>> Although I would only recommend cutting the connection to the ground
>>> plane as a last resort, I would have to disagree with your comment on
>>> not using a sharp knife to cut the copper foil. The copper is
>>> relatively soft at a thickness of maybe 2 thousands of an inch. An
>>> Exacto or similar knife can cut through it fairly easily. However,
>>> the prepreg layers used to separate an internal copper layer of a
>>> 4-layer board are going to be probably more than 7 thousands of an
>>> inch thick and *much* tougher to cut through than the copper. When
>>> cutting through copper traces/planes it is very easy to see when
>>> you’ve gotten through the copper and reached the substrate. If you
>>> are fearful of cutting too deep, use less pressure on the knife and
>>> several successive cuts to get through the copper. I have never cut a
>>> trace deep enough on a multi-layer board to damage an internal layer.
>>>
>>>
>