Author Topic: Noob question: can high temperature during soldering kill/erase an EPROM chip ?  (Read 4431 times)

0 Members and 1 Guest are viewing this topic.

Offline JLLTopic starter

  • Newbie
  • Posts: 5
  • Country: jp
Hello,

I'm currently trying to debug an issue I have and was wondering if high temperature could be the cause of my problem.

Context: I'm building a SFC cart repro for the Gradius 3 SA-1 enhanced mod.
I built two carts: one with a ZIF socket so I could test the EPROM chip quickly (27C322 chip) and a 2nd one so I could solder the chip directly on it.
I have confirmed that the chip works and was written to successfully using the debug cart.

But then after soldering the chip on the 2nd cart, it didn't work at all. Since I had to cut the legs on the chip, I cannot insert it on the debug cart anymore to test it.

While I was soldering the chip on the 2nd cart, I made a bit of a mess and took way longer than I would have liked: dozens of seconds instead of just a few per pin, at 300 degrees (celcius). I noticed the chip very very hot (I could not touch it without getting burned).

Could this have altered the data stored on the chip or killed it in some way ?

Thanks in advance for the feedback.
 

Offline Jwillis

  • Super Contributor
  • ***
  • Posts: 1725
  • Country: ca
I don't think they like a lot of heat. I've only used sockets for this type of sensitive chip. But that was years ago.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Unlikely. Even modern Flash (very tiny cells holding multiple levels of charge) can usually be resoldered with e.g. hot air machine (since we'd be talking SMT parts where the whole body reaches soldering temp).

If that's a ceramic body EPROM, the heat will be effectively sunk by the body, so the die doesn't see extraordinary temps.  It might've still been within rated storage temp, though still very much untouchably hot.

Might not hurt to check its contents or re-burn it though.  Also possible you zapped a pin or something.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: JLL

Offline JLLTopic starter

  • Newbie
  • Posts: 5
  • Country: jp
Thank you for the info. I'll find a way to test/reburn it.
 

Offline Brumby

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
I winced at this
... and took way longer than I would have liked: dozens of seconds instead of just a few per pin, at 300 degrees (celcius). I noticed the chip very very hot (I could not touch it without getting burned).

Certainly, electronic components of all sorts must withstand soldering temperatures - but they always have a time limit next to the temperature limits.  I've seen 10 seconds at 300 (and something) degrees Celsius for a lot of things, but some are more resilient ... and some less.

Between 1 and 3 seconds per pin is my normal range - and I think it's the same for most.  For multi pin packages, I tend to pause every few pins or so, to limit the thermal assault.  The thought of "dozens of seconds" per pin is toothache territory.  This can allow heat to find its way to the active region where it can degrade or destroy the function of the component.

From your description, I am inclined to think it has been "baked".
 

Offline Brumby

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Unlikely. Even modern Flash (very tiny cells holding multiple levels of charge) can usually be resoldered with e.g. hot air machine (since we'd be talking SMT parts where the whole body reaches soldering temp).

If that's a ceramic body EPROM, the heat will be effectively sunk by the body, so the die doesn't see extraordinary temps.  It might've still been within rated storage temp, though still very much untouchably hot.
All true - but this is still subject to the time consideration.

Quote
Might not hurt to check its contents or re-burn it though.  Also possible you zapped a pin or something.
This is also possible.  A handling error rather than an overheating one.
 
The following users thanked this post: T3sl4co1l

Offline mcovington

  • Regular Contributor
  • *
  • Posts: 191
  • Country: us
When you install the next one, use a bit of flux; also make sure your iron is as hot as you think it is.  It is much easier to overheat things with an iron that is too cool than with an iron that is too  hot, because of the time factor.  I set mine to 375 C.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Which to extrapolate a bit further -- one of the most common mistakes beginners make with soldering, is not having a wet tip -- you'll find you are pretty much always applying solder to the joint, even when removing parts, or removing solder (say with a hollow tipped desoldering iron).  If you don't have a wet, molten bead, it's no good!

And the reason that's a problem is the most common step people take is just cranking up the iron temp, which oxidizes the tip and burns away flux, rapidly.  You shouldn't need over 350°C for most work, occasionally going up to 400-430 for large joints, heavy ground plane, etc.

Or if you don't have a temp controlled iron, the average 30-40W passive iron can get pretty toasty just sitting there.  Keep this in mind when conditioning the tip -- clean it with a sponge and fresh solder before use, etc.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: Mp3

Offline Mp3

  • Frequent Contributor
  • **
  • Posts: 361
  • Country: us
It sounds like the IC package may have actually got too hot. Hopefully the ceramic absorbed the heat and kept the IC safe, but if you can't verify the ROM contents against the image then it's a moot point.

I have gotten some IC's so hot while soldering i was sure i toasted them, but they kept working after i finished the job and let it cool down. Then again, i woudn't say any of them were "burning" feeling.

As a general rule. If the soldering action doesn't happen within 2 seconds, you should remove the iron.

OP - Did you cut the legs of the old EPROM and solder the new EPROM to the legs? I have done this before on my own projects, but it's really risky soldering the EPROM directly.  It's best to desolder from the other side of the board and install a socket in its place. But, i have done the procedure of cutting from the top and soldering legs to legs directly myself. If you do that, secure the corners first so it has some good stability. check every so often to be sure you haven't dripped solder, and be careful to test the continuity of every pin to its neighboring pin after soldering. Lastly, be sure you haven't dripped solder anywhere.

I'd also install a battery holder while you're in there. Nasty stuff happens with soldered-in batteries that get way too low. I have a few saved from repairs I did that are particularly impressive in terms of grodiness.
High school graduate
 

Offline NivagSwerdna

  • Super Contributor
  • ***
  • Posts: 2507
  • Country: gb
I'm currently trying to debug an issue I have and was wondering if high temperature could be the cause of my problem.
Certain components e.g. LEDs don't respond well to prolonged heat but in my experience they are hard to kill.  If you have an interesting failure mode you should probably investigate it (add flying leads to cut pins) by some diagnostics... it might be worth the effort to understand what went wrong.

Interestingly the old datasheets eg. http://hxc2001.free.fr/Squale/datasheet/27C32.PDF, https://static1.squarespace.com/static/51f517f0e4b01da70d01ca2a/t/5575515de4b0888fee6170ec/1433751901367/27C322.pdf etc. don't talk about reflow... I guess the assumption is they are socketed.
« Last Edit: February 18, 2020, 11:20:08 am by NivagSwerdna »
 

Offline bluevd

  • Contributor
  • Posts: 13
While this is most certainly not the case, it's worth mentioning: ESD safe irons can do this if they are not properly handled. I've managed to fry a PIC once, despite the fact the iron was grounded, my circuit was laying on a cutting mat which I presume was statically charged.
And ESD mat *and* ESD iron is a must when dealing with IC's.
 

Offline JLLTopic starter

  • Newbie
  • Posts: 5
  • Country: jp
Thanks to everyone for the feedback and tips. I'm still learning how to properly solder so mistakes are definitely bound to happen...

I actually ended up programming a new chip and used that one for my final cartridge.
This time, I made sure to not go too long on each pin and leave some time for it to cool off before moving to the next ones.
I also triple checked each solder point on my PCB with a multimeter and made sure everything was soldered properly.


OP - Did you cut the legs of the old EPROM and solder the new EPROM to the legs?

I was actually following this video from Voultar



and he cuts the pins at 11:20 because of the adapter that is used (the red PCB in the video)
When removing the EPROM, I did it properly and desoldered everything with braid and a pump (that took ages...)

Quote
and be careful to test the continuity of every pin to its neighboring pin after soldering

Yes, lesson learnt now, this is also what got me at some point: I had a faulty adapter (it's failing a simple continuity check, just the adapter itself) and it took me quite some time to diagnose. I still have to contact the seller and see what can be done about this.

As for the old chip, I wanted to try soldering some wire to each pin and reconnect it to my programmer (TL866II with an adapter for the 27c322) to test if it's working or not.
Or is that too much trouble ? Should I not bother and just trash it ? Is there any better way to test it ?

Finally here's a picture of the two finished carts, both are working.

930432-0
« Last Edit: February 18, 2020, 01:45:11 pm by JLL »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf