0
\$\begingroup\$

I soldered my first TQFP 100 package, an Altera MAX II EPM570 CPLD, on a PCB.

But I only found the "correct" way to solder (put flux on the pads and on the pins, melt some tin on a thin iron tip, then gently move the tip across the pads without touching the pins) on the fourth side, after several tries with the other three.

Blobs were forming on the pins; I tried using solder wick to remove them, but for some reason it didn't absorb absolutely anything. Even with the iron set to 420°C and pressing the tip for 10-20 seconds a few times. Then I read the advice to apply solder flux to the wick, and finally it started doing its job. So I managed to remove all the blobs.

This is how it looks now. East side went flawlessly, the other sides are repaired:

enter image description here

I'm concerned I've burnt the chip due to excessive heat. Here are my questions:

  • Should I just desolder this chip and try again with another one?
  • If I decide to keep it, can I verify that all the I/Os are still working before mounting all the other components? I can solder what's necessary to power it up and connect via JTAG. Reason I ask, is that I have a CPLD with a burnt I/O with everything else still functioning otherwise.
\$\endgroup\$
4
  • \$\begingroup\$ Try cleaning the pins with plenty of 91% isopropyl alcohol and an old toothbrush. Use a magnifying glass or microscope to inspect each pin. 420°C is really hot, but it's probably ok. Chances are much higher of microscopic solder bridges. \$\endgroup\$
    – rdtsc
    Sep 23, 2020 at 17:19
  • \$\begingroup\$ It's actually quite unlikely that you damaged the silicon, the question would be if you actually got everything correctly connected. Hopefully this is an experimenter project, not a safety critical system. At this point you're as likely to do permanent damage to the board trying to change out the chip, as you are to have a problem with the current chip. At least see if it works before you give up. \$\endgroup\$ Sep 23, 2020 at 17:34
  • \$\begingroup\$ If you want a test routine, try something that's either a "chase" pattern around the I/Os or a counter with a different bit on each. Then probe with a scope. Typically if you have an adjacent pin short, you'll get at least some period of illicit intermediate voltage as the shorted pair of I/Os fight each other. There are a few spots below the writing that I'd want to inspect under magnification regardless. \$\endgroup\$ Sep 23, 2020 at 17:35
  • \$\begingroup\$ if you decide to scrap the chip, then cut all the pins next to the chip body ... that way you can use minimal heat to remove the pins from the pads \$\endgroup\$
    – jsotola
    Sep 24, 2020 at 1:46

2 Answers 2

2
\$\begingroup\$

Pin 100 doesn't look soldered and there are a few spots where there could be shorts (or maybe just reflections). As rdtsc says, clean it well and look at it under a good microscope or magnifier, tilt it to see if any solder has gotten in behind the leads.

I would poke at any leads that look like they might not be soldered first (a dental pick works well), fasten them down to make sure they don't get bent, and then scrub the leads with a brush (one of those free ones from the dentist works well) and 99% isopropanol.

Make sure you have fresh solder wick of appropriate size from a reputable maker. Doesn't hurt to throw some liquid flux on there if you have to touch things up (a flux pen is a handy way to do that).

It's unlikely you've damaged the chip with a soldering iron. Usually the board gets damaged first.

\$\endgroup\$
1
\$\begingroup\$

I see a short between two pins just above C11, under "3W3N...". And another one under "...9Z0..." on the same side. Do these again.

To make sure there is no solder bridge between pins, move a needle between each pin. If the needle stumble, it means there is a solder bridge between them. The needle won't be able to cut the solder bridge and don't try to do it with it. Use your solder iron and the solder wick. The needle helps you to feel that there is nothing and can clean solder micro balls. Don't apply too much strength on the needle to avoid scratching the PCB. The goal is to feel it. Not to scratch out dirt.

Then, you must check each pin with a multimeter. For each pin, you must do two tests:

  1. A connectivity test with the meter on 200 ohms (or the lowest denominator). One probe on the pin, the other on another component supposed to be connected to this pin.
  2. An isolation test: Turn the meter to 200K or 1M or something of this order. In some case you will have to lower the impedance to get a "1" indicating that there is no direct connection. Internal connections can show variable values if the multimeter impedance is set too high. Then check each pin with the one next to it. One probe on one pin, the other one the next one.

This test is not a guarantee that it will work. But one fail is a guarantee that it won't. Do both tests with each pin. Repeat if you are not sure. Don't be lazy.

Using a solder wick, you always have to add flux, as you have found out. But rising the temperature to 420°C is useless and may damage the ic. 270°C is enough and safe. As you applied 420°C for 10 to 210 sec, I would say there is a risk of the chip being damaged by the heat. But since it's a long work to redo everything, I would try like that.

Another way I'm using is to apply solder paste on pads (flux containing tin - this is expensive and more difficult to find). Then applying the solder iron to melt this solder paste and solder the pins. This is more precise than using tin wires. Before I used a hot air gun to imitates reflow soldering, but the air flow is moving the chip to often and I had to redo everything with the solder iron anyway. I noticed I save time using the iron immediately.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.