OmniAtari

[ijor]

Actually, I understand that FB3 is(was) supposed to be based on an FPGA. Are the FB1/2 ASIC based?

 

In all this debate about ASIC vs. FPGA, nobody answered my question.

 

I just saw pics of the FB2 PCB for the first time (yeah, sorry, I'm guilty, don't have one). And I see the two main chips are sealed in epoxy. Now, why do you think they are sealed? Well, more than likely because there is no ASIC at all in the FB2!

[Retro Rogue]

Actually, I understand that FB3 is(was) supposed to be based on an FPGA. Are the FB1/2 ASIC based?

 

In all this debate about ASIC vs. FPGA, nobody answered my question.

 

I just saw pics of the FB2 PCB for the first time (yeah, sorry, I'm guilty, don't have one). And I see the two main chips are sealed in epoxy. Now, why do you think they are sealed? Well, more than likely because there is no ASIC at all in the FB2!

 

 

Boy, talk about drawing conclusions and reaching for straws, the FB2 main chip is ASIC.

 

From this old thread:

"The new FB 2.0 2600 chip (by the way for those who like how the old Atari had codenames for projects/chips the new 2600 FB 2.0 chip is codenamed "Michele") is an ASIC and takes much smaller nibbles in power consumption so it is a perfect device for handheld implementation."

 

 

Also:

Actually, I understand that FB3 is(was) supposed to be based on an FPGA.

 

I think you're missunderstanding from this statement that was posted in the FB3 thread:

 

"The PAL FB2's are sitting here in my office, Atari didn't commission and pay for their development, it was done independently, as everyone knows, the NSTC 2600 chip and PAL chip are have numerous differences, making them completely independant of one another. so Atari doesn't own them and they will unfortunately sit here and gather dust on the development shelf, they are only FPGA's and not product ASIC's inside."

Edited January 6, 2007 by wgungfu

[ijor]

Boy, talk about drawing conclusions and reaching for straws, the FB2 main chip is ASIC.

 

I see, thanks for the clarification.

 

But then, why the epoxy??? The chances that somebody would try to reverse engineer the ASIC are so small. If somebody has the willing/money/time/resources for that, then the epoxy likely is not going to stop him.

 

Or the epoxy is there for reasons not related to reverse engineering? (I'll understand if you can't/don't want to answer).

Edited January 6, 2007 by ijor

[ijor]

Btw, speaking about ASIC reverse engineering. I don't know if you or Curt noted my post in the other thread about the reverse engineered Sally schematics. (I was expecting lot of WOWs, so I guess many missed it).

 

Studying the schematics might be useful for those elusive undocumented op-codes that are still not fully understood.

[supercat]

But then, why the epoxy??? The chances that somebody would try to reverse engineer the ASIC are so small. If somebody has the willing/money/time/resources for that, then the epoxy likely is not going to stop him.

 

Integrated circuit chips need to be encased in airtight packaging or else they will be destroyed pretty quickly by oxidation and other causes. This is usually accomplished in one of three ways:

 

-1- The chip may be installed in a package designed to be soldered to a printed circuit board. When the Atari 2600 was produced, the most common form of package, called a DIP (dual inline package) featured two rows of pins on 0.1" centers. Packages with 20 or fewer pins would have the two rows 0.3" apart; for 22-pin parts, 0.4"; for 24-48 pin parts, 0.6"; for 64-pin parts, 0.8". Most newer chips in 24 or 28-pin DIP packages use 0.3" row spacing. Further, newer packaging styles have been developed which are much smaller than DIPs. A 44-pin PLCC package, for example, is less than half the size of a 48-pin DIP, and some newer styles are less than half the size of that.

 

-2- The chip may be molded into a user-manipulable device like a SmartMedia card.

 

-3- The chip may be soldered directly to a printed circuit board. In this case it is necessary to add an air-tight covering, typically epoxy, to protect the chip from oxidation or other damage.

 

Packaging chips costs money, and thus packaged chips cost more than unpackaged chips (called "bare dice"). On the other hand, chips that are packaged are a lot more durable than bare dice; the latter require specialized equipment for handling, soldering, and encapsuation. Further, bare dice are often sold in containers holding hundreds or thousands of chips. Once a container is opened, the chips inside it must be encapsulated within days if not hours.

 

If a company is going to be making a run of 100,000 circuit boards, the cost of setting up the specialized equipment necessary to handle bare dice may be much less than the cost savings of bare dice over packaged chips even if a hundred chips left over at the end of the batch have to be junked. On the other hand, it would be silly to spend thousands of dollars setting up a die-bonding machine, and then buy a package of 350 bare dice, if one only needed to make 50 boards. Much cheaper in that case to simply buy and use packaged chips.

[ijor]

If a company is going to be making a run of 100,000 circuit boards, the cost of setting up the specialized equipment necessary to handle bare dice may be much less than the cost savings of bare dice over packaged chips even if a hundred chips left over at the end of the batch have to be junked.

 

Thanks supercat, very nice explanation. I didn't know that.

 

Just one question. Then why you never see (at least I never noted) the ASIC chips being sealed in mass produced boards, like PC cards?

 

Or is it because when you produce in huge quantities, then the packaged vs. bar dice equation changes?