Programming before reflow soldering - is it possible?

Go To Last Post
17 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi,
I have inquiry from my big client of supply to him preprogrammed ATMegaXXXX in TQFP pack. It is intended for final assembly on PCB with another SMD parts and standard reflow soldering.
Code is mine and I have idea to ship programmed/protected by me chips rather than shipping twice ready boards for programming.
This is of course normally done with DIL package in socket but here is very small board and automated assembly.
In Atmel materials I can't find information is Flash and EEPROM content erase or corrupted in higher temperature when soldering. All I find is typical temperature range.
Is it all possible?
Have anybody such experience? Or helped link?
Best regards, Szymon

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Sure, you can get test sockets for any chip. Probably want to get a gang programmer for this job. You probably have to fab your own socket board in a layout that suits the gang you want. Serial program them and fuse the lock bit when you're done to protect contents.

Regards,
Scott

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'm curious about this as well.. Programming them with a test socket is (to me) obvious, but then what? You have to put them back in the tape they came in? How would you do that for a couple thousand chips?

/* John Butera */

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

have you done search for programming houses, like :
http://www.arrowne.com/news/ar10...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Arrow Global Programming Services:
http://www.arrow.com/aws/pg_tbGe...$$5013$$0$$Component%20Programming

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Programming by experts

MSC High-Tech-Programming Service - Your Complete Service From One Site !

http://www.msc-ge.com/cgi-bin/fr...
http://www.msc-ge.com/download/p...

MSC is a European Atmel ditributor located in Germany which also has a programming service.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi,
thanks a lot, but my problem is what happened with program (flash, eeprom) in AVR when it goes to soldering oven (over 200'C for some seconds, after preheating). Will be lost or no, or after some time?
Szymon

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

So the question: does heat hurt the chip memory contents if you stay in the specs for soldering? My guess is not. The temperature rating of the chip (C or I) is for operating range, with Vcc applied, its ability to execute. Nothing to do with retaining flash and eeprom contents. Maybe there's a statement on flash durability that has a temp range for retention. I know it has a lifespan, and a write-span, maybe a temp-span too?

Regards,
Scott

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

As long as you stay within the soldering profile for the chip, the non-volatile memories will be unaffected by the soldering process. Prologued exposure to extreme heat, can not only damage the non-volatile memories, but can damage the entire chip.

Writing code is like having sex.... make one little mistake, and you're supporting it for life.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Normally it should be OK. But this is based upon a 20-30 second reflow time on lead based soldering. I suggest that you be vary weary of this. Especially if your customer is going to use a lead free solder paste. Prolonged exposure to reflow type heat will reduce the memory retention. I once enquired in a similar way to Atmel when I wanted to look at using a part in an overmoulded housing, and not just for one or two parts. The IC would have been subject to the 230C preheat in the mould tool for around 20 seconds, with a further 30 second plastic flow and then cool down. This was all of course on top of the normal manufacturing profiles. I never did get a direct answer. And with the lead free pastes, the reflow temperature can exceed this figure.

If you are looking for solder profiles, they may be in the quality section. Otherwise, you may be best looking at the website of a company named Amkor. Although this won't answer your original question.

Is there no way you can supply your customer with an in-circuit programmer? This way, you can ensure that they have the correct code just before their products are tested.

Sacha.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

It sounds like your basic requirement is to avoid your client "accidentally" programming more parts than he's paid royalties on. :shock:

There could be other ways. You could design an ISP that programs your client's Mega and contains a "meter" that requires you to "recharge" it at regular intervals. If he's supposed to make 25 widgets this month, set the EEPROM so it will program 25, then he'll ship the ISP back to you (instead of shipping 25 boards), with his payment, and you'll cash his check and reset his EEPROM. :D

If shipping the ISP is an issue, you could have secret codes to enable the device for X number of programming, and when he pays you give him the next code to punch in.

Your ISP device would only need to know how to program one type of Mega, and its own code (as well as the program) would be protected in its FLASH.

Your primary risk becomes, will he try to monitor the ISP signals? Well, if he really wanted to he could hire someone for as little as a few thousand to dissect your chip. You can only be so safe. You can also put tricks in your ISP signals to make it difficult to capture accurately. And, if he tries to unplug it before the fuses are set, disable it (He only gets one shot at making an 'unprotected' chip to read! :evil: )

Hey, then you could even mass produce "the secure ISP" and sell that! :twisted:

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If your doing large production runs, you can get Atmel to pre-program the chips for you. Usually though you need to take into account the lead time to get ready for this.

Hummm. I had a wild thought.
Why not have the pick and place machine pick up the MCU from the tape, then place it in a "Test socket" then program the MCU, then pick it back up and place it on the PCB to be reflow soldered on down the line.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

earlwb wrote:
Hummm. I had a wild thought.
Why not have the pick and place machine pick up the MCU from the tape, then place it in a "Test socket" then program the MCU, then pick it back up and place it on the PCB to be reflow soldered on down the line.

I think it would be chaeper and much easier to order the AVRs preprogrammed from one of the distibutors mentioned in this topic or Atmel.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks to all for interesting answers and concepts.
The another one idea is to manually solder smd chip with hot air solder station, with special holder to place part on board. I think it is safer to memory content.
Szymon

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You should have no worries for the FLASH content of the MCU if it programmed before shipping as long as the correct soldering profile is used. I got the "IPC/JEDEC J-STD-020B" standard from Atmel when I requested the soldering profiles or SMD devices. This one should cover all packages if I am not mistaken, both leaded and lead-free. It also talks about the moisture sensitivity of the packages (grouping for actual packages can be found on Amkors website).

Hope this helps you out.
Regards

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

palka wrote:
I got the "IPC/JEDEC J-STD-020B" standard from Atmel when I requested the soldering profiles or SMD devices.

You can download "IPC/JEDEC J-STD-020B - July 2002" direcly from JEDEC here:
http://www.jedec.org/download/se...

And here's the newer "IPC/JEDEC J-STD-020C - July 2004":
http://www.jedec.org/download/se...

IPC/JEDEC J-STD-020C - July 2004 supersedes:
IPC/JEDEC J-STD-020B - July 2002
IPC/JEDEC J-STD-020A - April 1999
J-STD-020 - October 1996
JEDEC JESD22-A112
IPC-SM-786A - January 1995
IPC-SM-786 - December 1990

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I just came across this FAQ at Atmel's website:
http://www.atmel.com/quality/qua...

Quote:
Question
1. How can I get more information on reflow profiles and moisture sensitivity

Answer
1. The JEDEC Web Site is a good reference. Both the J-STD-020A (Moisture/Reflow Sensitivity Classification for Non-Hermetic Solid State Surface-Mount Devices) and the J-STD-033 (Standard for Handling, Packing, Shipping, and Use of Moisture/Reflow Sensitive Surface-Mount Devices) can be downloaded for free at: www.jedec.org/DOWNLOAD/default.cfm

The FAQ was probably made before J-STD-020B was introduced.