Desoldering through-hole transistor mounted to heat sink

[matthuszagh]

I'm attempting (unsuccessfully) to desolder a transistor which I believe to be faulty in an HP 11729C. I've attached two images (part is circled). They show the lead and top sides of the transistor (it's the 2N3054A). I'm using a JBC desoldering iron with a tip whose max supported lead diameter is 0.8mm (opening diameter is 1mm), which is the actual lead diameter. The procedure I'm using is to:

1. ensure there is enough existing solder on the lead to get good thermal transfer (the picture shows most of the solder removed, but I don't start this way)
2. set the desoldering iron to 400C (I previously tried 350).
3. place the desoldering iron over the pin and wait for the solder to melt
4. turn on the vacuum

This has not been effective at removing all of the solder from the through-hole. Any strategies for how  to do this? If I had access to the lead on the top side of the PCB I could use a soldering iron on this side with the desoldering iron on the backside. Unfortunately, I don't have access to the lead on the top side as the heat sink blocks this. Moreover, the heatsink cannot be removed until all of the transistors are desoldered, so there is a bit of a chicken and egg problem.

[xrunner]

I might have an idea. But first - are the two leads of the transistor bent over or are they straight coming out of the hole? I am not sure from the pic. The idea might not work well if they are bent coming out of the hole.

[matthuszagh]

I might have an idea. But first - are the two leads of the transistor bent over or are they straight coming out of the hole? I am not sure from the pic. The idea might not work well if they are bent coming out of the hole.

They're straight.

[xrunner]

OK. If their aren't any better ideas coming (I'm sure there will be) ...

You might try getting a thick piece of solid copper wire. You are going to make a heat pipe with it. It will extend from one pin to the other laying next to them. Heat the wire and use solder to transfer the heat back and forth so that the solder in both pins will melt. When you see they are melted - pull the transistor out from the other side. The after that suck out the remaining solder.

It's just one idea so wait for more suggestions and pick the best. 

[andy3055]

Looks like the pins are slightly bent. If so, while heating each one, try to straighten them more. Put more fresh solder on them and make sure there is enough flux in it. Then heat the joints well. Alternate between the pins so that they do not cool down and solidify the lead. Silicon transistors can take a lot of heat, so no need to worry. also, if you are removing it I assume you plan replacing it anyway. If you have a de-soldering gun, it is much easier.

You may be able to get some very thin screw driver blade between the H/sink and the transistor from the other side and lever it bit by bit. Do not force it too much either as you can damage the traces.

[BrokenYugo]

Heat a few seconds longer before vacuuming? Obviously you want to put in minimum heat, but with a plated through hole you have to get it good and molten all the way to the other side.

[David Hess]

For parts like that, I usually *add* a little bit of solder for good heat transfer and then alternately heat each pin and "walk" the part out.  If you can bridge the pins like xrunner suggests, then it should just slide out.

[coromonadalix]

do i see metal tube rivets in the pcb vias

[SeanB]

do i see metal tube in the pcb vias

Yes there are rivets crimped to the PCB to make the connection more robust, but the leads will come loose if you add more solder, then allow the joint to heat all the way through the board first before trying to suck the solder out. Best though is to add extra flux, and then use desolder braid to pull the solder away, using a wide tip on your soldering iron to get good heat transfer. TO66 packages are pretty good at removing heat from the pins to the frame, so you will need a lot of contact, lots of braid and time to get the hole molted to the base.

[xrunner]

I'm attempting (unsuccessfully) to desolder a transistor which I believe to be faulty in an HP 11729C.

Did you notice the joint at the top in your pic? It looks like the solder hasn't flowed onto the pin of the other transistor (I think that's what it goes to). Looks like an original joint. Are you sure that isn't the problem with the circuit?

[PKTKS]

Forget your "soft"  iron and use a high profile 150W
hand GUN...  usually just  some brands available...

Combine that by flooding your spot with a
LOW MELT ALLOW ALLOY like

https://www.aliexpress.com/item/2Pcs-lot-MECHANIC-Lead-Free-Solder-Soldering-Wire-0-3-0-4-0-5-0-6/33016004971.html

Good enough for normal budget and in time results.

Paul

[RayRay]

Just create a solder bridge with your soldering iron (bridge both pins with solder, adding flux can help it flow too), when done, heat it up fron one side, and pull it out from the other (using tweezers or pliers) and when done, clean up the solder with a wick (or pump/wick combo). Should be pretty straightforward. Temperature wise, I'd advise you to never go over 370C at worst.

[matthuszagh]

Yes there are rivets crimped to the PCB to make the connection more robust, but the leads will come loose if you add more solder, then allow the joint to heat all the way through the board first before trying to suck the solder out.

Interesting, I was wondering what these were.

Best though is to add extra flux, and then use desolder braid to pull the solder away, using a wide tip on your soldering iron to get good heat transfer. TO66 packages are pretty good at removing heat from the pins to the frame, so you will need a lot of contact, lots of braid and time to get the hole molted to the base.

To clarify, are you saying that the wide tip + desolder braid will be more effective than the vacuum desolder gun?

[matthuszagh]

Did you notice the joint at the top in your pic? It looks like the solder hasn't flowed onto the pin of the other transistor (I think that's what it goes to). Looks like an original joint. Are you sure that isn't the problem with the circuit?

Nicely spotted. That's not actually the only one as well. All the leads for regulator ICs attached to the same heatsink as this transistor look like that. They all seem to be providing the correct voltage under load, so I didn't think they were an issue. Still, I was planning to reflow them at some point. I was curious how this could have happened. It looks like some sort of mechanical shock dislodged the pins from the joint. But that's just a guess. Any other ideas as to what may have happened here?

[matthuszagh]

Forget your "soft"  iron and use a high profile 150W
hand GUN...  usually just  some brands available...

I'm currently using a desoldering gun that can provide up to 160W.

Combine that by flooding your spot with a
LOW MELT ALLOW ALLOY like

https://www.aliexpress.com/item/2Pcs-lot-MECHANIC-Lead-Free-Solder-Soldering-Wire-0-3-0-4-0-5-0-6/33016004971.html

Good enough for normal budget and in time results.

Admittedly, I am using lead-free solder as I'm really trying to transition away from leaded solder. I've had pretty good results with lead-free, though, if it's going to make a big difference I could use leaded solder for this (all the original joints will probably be leaded anyway).

[SilverSolder]

[...]
Admittedly, I am using lead-free solder[...]

OMG stop that right away, it is much harder to work with and not necessarily compatible with older boards.

Lead free solder is like breast free women - a bad idea.

[PKTKS]

Forget your "soft"  iron and use a high profile 150W
hand GUN...  usually just  some brands available...

I'm currently using a desoldering gun that can provide up to 160W.

Combine that by flooding your spot with a
LOW MELT ALLOW ALLOY like

https://www.aliexpress.com/item/2Pcs-lot-MECHANIC-Lead-Free-Solder-Soldering-Wire-0-3-0-4-0-5-0-6/33016004971.html

Good enough for normal budget and in time results.

Admittedly, I am using lead-free solder as I'm really trying to transition away from leaded solder. I've had pretty good results with lead-free, though, if it's going to make a big difference I could use leaded solder for this (all the original joints will probably be leaded anyway).

Desoldering (if I got you right like vacuum ? ) just will not
cope with high thermal mass..

I am suggesting that hand GUN w/embed transformers..
they can cope w/high thermal mass fine.. 150W minimum

I am suggesting LOW MELT ALLOY BISMUTH based.

NOT LEAD FREE

They will form a final mass with lower melt point.

ALAS my experience w/lead free is just a crap.

They brittle and crack in particular just after being
excessive heated like this..

LEAD FREE is just garbage for serious repair.

I have checked the suggested BISMUTH and my opinion
is that you need to remove that later...

I always replace with leaded..

simple fast example..
https://www.aliexpress.com/item/1005002670551859.html

I have an old school one very robust hand gun..
150W is piece of cake on the wire.

Paul

[TheMG]

Get a larger tip for your vacuum desoldering tool!

If you're trying to desolder a 1mm diameter lead with a tip that has a 1mm opening, that just isn't going to work, the lead blocks most of the suction and it can't pull the solder through effectively.

It's a good idea to have 2 or 3 different size tips for a vacuum desoldering tool for these kinds of scenarios. The tip you use should have an opening that is larger that the lead diameter, but the outside diameter of the tip should not exceed the size of the PCB pad.

[matthuszagh]

Get a larger tip for your vacuum desoldering tool!

If you're trying to desolder a 1mm diameter lead with a tip that has a 1mm opening, that just isn't going to work, the lead blocks most of the suction and it can't pull the solder through effectively.

It's a good idea to have 2 or 3 different size tips for a vacuum desoldering tool for these kinds of scenarios. The tip you use should have an opening that is larger that the lead diameter, but the outside diameter of the tip should not exceed the size of the PCB pad.

The lead is 0.8mm and the tip opening is 1mm. Moreover, the tip is stated to support a maximum 0.8mm diameter. Do you still think I need a larger tip opening?

[BrokenYugo]

Lead free is all well and good now but I wouldn't use it on old stuff built with leaded. Lead free makes you put more heat into the board, increasing chances of damage, and develops cracked joints more easily.

[SilverSolder]

Get a larger tip for your vacuum desoldering tool!

If you're trying to desolder a 1mm diameter lead with a tip that has a 1mm opening, that just isn't going to work, the lead blocks most of the suction and it can't pull the solder through effectively.

It's a good idea to have 2 or 3 different size tips for a vacuum desoldering tool for these kinds of scenarios. The tip you use should have an opening that is larger that the lead diameter, but the outside diameter of the tip should not exceed the size of the PCB pad.

The lead is 0.8mm and the tip opening is 1mm. Moreover, the tip is stated to support a maximum 0.8mm diameter. Do you still think I need a larger tip opening?

1mm will probably work.  The problem in this specific situation is to melt all the solder in the hole, despite the lead for the transistor carrying the heat away all the time. 

A good basic approach here is: 

1) Wick out the lead free solder, by applying fresh leaded solder and removing with solder braid a couple of times.  This stuff is just going to make things harder.

2) Leave a generous amount of leaded solder on the joint to work with, to act as a heat spreader/reservoir.  Heat it up with your regular soldering iron until completely melted through, then

3) very quickly apply the solder sucker (set to a nice hot temperature) into the pool of molten solder, and slurp everything right out.  Quick, as in hold one tool in each hand and swap in a second.