Once you replace the power transistors you might want to screw the heatsink tabs to a 1/8" thick aluminum heatsink plate as big as space allows just like your old CW-80 uses. Use a silver-based heatsink paste on the back of each transistor tab, again like the old CW-80. This may not protect your new transistors against big transients like what blew them in the first place but you'll get better continuous load reliability out of them.
@Bill Swatos posted:Once you replace the power transistors you might want to screw the heatsink tabs to a 1/8" thick aluminum heatsink plate as big as space allows just like your old CW-80 uses. Use a silver-based heatsink paste on the back of each transistor tab, again like the old CW-80. This may not protect your new transistors against big transients like what blew them in the first place but you'll get better continuous load reliability out of them.
Unless you also insulate them electrically you would just short out the board. Let's not go modifying this just yet. It was designed to be fanless and does not require a heatsink. Load testing also was validated and these are not getting hot under load. Heck, not even warm.
I honestly considered this myself (adding a heatsink) when I cracked mine open to evaluate it. That said, between the work required to measure, cut, and drill precision spacing of the mounting holes to not stress the transistors, as well as the supplies (thermal insulator, bushing for tab hole, screws, nuts) needed to properly insulate the tabs let alone getting it to fit in the case that is now sealed without a fan.
Again, this modification is work and money with very little payoff as it simply is not required on this model.
Yes, the old design needed a heatsink and a fan to remain in safe range. this new one, completely opposite end of the spectrum.
The fact that they used FET's for voltage control greatly reduces the power dissipation in the control circuit, a whole different ballgame than the old CW-80 with the triacs. You send a few amps through a triac and you need a significant heatsink and ventilation to keep it in it's operating temperature range.
Update on CW-80 repair - I finally got the transistors and tried to replace the two that were shorted out. Vernon was dead bang on target when he said my challenge would be removing the old transistors. I tried using the solder sucker but couldn't make it work very well. I had better luck with the wick and was able to get most of the solder out. But the transistors are still firmly implanted in the board - it looks like they were machine pressed into it. I am afraid to use force to pry them out so I am wondering if there is any technique or tool for removing them. Any ideas would be greatly welcomed. Thanks.
What I do for similar parts is to cut both sides off flush with the PCB. Then I turn up my soldering station to 650-700F and use the tip of the iron to push the remaining lead out from the bottom. The might be "tight" in the holes, but they're not truly machine pressed. In my experience, it's much harder to try to remove the parts intact, and since you're tossing them anyway, there's no incentive not to simply chop them off and toss them.
I agree 100% with the above methods. Do not yank out the remaining leads and rip out the through hole plating of the PCB. This is why I said doing it right would be difficult but is possible and you don't have to have specialized equipment, but it sure makes the repair a heck of a lot easier.
Again, you ideally need heat to melt the solder, but a temp controlled iron is important to not overheat and damage the PCB.
So yes, as John described, cut the leads off at the transistor so you can work on them one by one rather than 3 at the same time. You may even want to add solder to aid in heat conduction to ensure both sides of the PCB (dual layer board) are melting so the pin just floats out of molten solder. Then worry about using your solder sucker or solder wick to clear the holes.
ARRRRGH!@#%$!#@! After spending 3+ hours trying to get the transistor pins out of the board I had almost no luck. I read the posts, watched the video links and went in for another try (it seemed so easy). I cut off the shorted transistor and then tried to push out the remaining piece with the conical tip of my iron but it was too big to get into the hole. One of the pins was long enough to grab with pliers and I got it out but the other two were too short to grab. I increase the temp of the iron from 650 to 750 and tried adding a little fresh solder but still no luck - obviously a lot of operator error going on. At this point I started handling the board too roughly out of frustration so I stopped. Figuring I had ruined the board with too much heat and rough handling I reassembled it to see if the Accessory terminals still worked and to my amazement they did! Its probably time to give up as I was able to replace the fan in the old CW-80 and it works for the accessories I need to power. But its so hard to admit defeat. The only things I can think of are: 1) to drill out the old pins (but I am sure that is a bad idea), or to clip down the pins of the new transistors and try to solder to the top of the old pins (which also seems like a bad plan). If anyone has any ideas I am all ears and willing to give it one last shot. At least I am gaining some experience though, and appreciate the gracious efforts on your part to help me.
Drilling them out will kill the board, so that's certainly a bad plan! If you are really at your wits end, the idea of cutting down the new transistors and soldering them to the stub should work.
Ran into this sometimes and cut it flush then used a beading needle to push through the hole to clear it while hot. Worked for me most of the time.
John
Yikes...and here I thought I was doing well by rescuing a 75-year old Type S yesterday...😳... this is way above my pay grade. More power to you! 🙂
Mark in Oregon
