Since some of the Commodore chips are in short supply, I'm wondering if this might be viable in the near future? There seems to be steady growth in this field.
Any thoughts?
3D Printed ICs
3D Printed ICs
Learning all the time...
- pixel
- Vic 20 Scientist
- Posts: 1356
- Joined: Fri Feb 28, 2014 3:56 am
- Website: http://hugbox.org/
- Location: Berlin, Germany
- Occupation: Pan–galactic shaman
Re: 3D Printed ICs
I was wondering, since that technology is pretty old, if it wasn't possible to just manufacture them the traditional way. I'm sure you can find some company in China that could do it.
A man without talent or ambition is most easily pleased. Others set his path and he is content.
https://github.com/SvenMichaelKlose
https://github.com/SvenMichaelKlose
Re: 3D Printed ICs
the problem are mostly making the masks and the setup costs. you'll also need a test/selection rig and have someone bond them into packages.... given the fact that you can perhaps sell a few thousends, thats going to be helluva expensive per IC
I'm just a Software Guy who has no Idea how the Hardware works. Don't listen to me.
-
- Vic 20 Newbie
- Posts: 15
- Joined: Sun May 29, 2016 3:18 pm
- Location: U.S.A. WYOMING
Re: 3D Printed ICs
HI,
Here in the U.S. I consulted with a CBM Engineer he told me the following below: (name withheld).
Nowadays the cost would probably be double of what it was in the 1980's way too expensive, you have to take into account the following
costs:
The building, tools, chemical tanks, Insurance to cover the chemicals in case of tank leakage etc, so this is all very expensive these days,
the only other way is FPGA "Field Programmable Gate Array's". FPGA uses either Verilog or VHDL language, but it can be very complex, takes awhile to learn the language.
Commodore always made about 1 million chips in a run so the cost was 1 million dollars (U.S.)
The only other problem is that the MOS chip layouts aren't available, but with FPGA one can try to approximate the chip or the Gates in the chip,
so not all overseas chip manufacturers may not have the actual MOS masks, that is another cost (chip masks).
I had always wanted to do something like you mention but the traditional way is just way too expensive to do unless you have a rich
uncle somewhere, would take Billions of dollars.
You work up the Gate code (of a MOS IC) to kind of test it is called a Benchmark, if that passes then theres a possible chance the
MOS approximation could be tested in any CBM 8-bit circuit. But starting simple is best.
Wish I had better news but these days this is the cheaper option.
Cheers,
traymond
Here in the U.S. I consulted with a CBM Engineer he told me the following below: (name withheld).
Nowadays the cost would probably be double of what it was in the 1980's way too expensive, you have to take into account the following
costs:
The building, tools, chemical tanks, Insurance to cover the chemicals in case of tank leakage etc, so this is all very expensive these days,
the only other way is FPGA "Field Programmable Gate Array's". FPGA uses either Verilog or VHDL language, but it can be very complex, takes awhile to learn the language.
Commodore always made about 1 million chips in a run so the cost was 1 million dollars (U.S.)
The only other problem is that the MOS chip layouts aren't available, but with FPGA one can try to approximate the chip or the Gates in the chip,
so not all overseas chip manufacturers may not have the actual MOS masks, that is another cost (chip masks).
I had always wanted to do something like you mention but the traditional way is just way too expensive to do unless you have a rich
uncle somewhere, would take Billions of dollars.
You work up the Gate code (of a MOS IC) to kind of test it is called a Benchmark, if that passes then theres a possible chance the
MOS approximation could be tested in any CBM 8-bit circuit. But starting simple is best.
Wish I had better news but these days this is the cheaper option.
Cheers,
traymond
-
- Vic 20 Newbie
- Posts: 15
- Joined: Sun May 29, 2016 3:18 pm
- Location: U.S.A. WYOMING
Re: 3D Printed ICs
ah sorry I didn't see your post but yes correct very expensive.groepaz wrote:the problem are mostly making the masks and the setup costs. you'll also need a test/selection rig and have someone bond them into packages.... given the fact that you can perhaps sell a few thousends, thats going to be helluva expensive per IC
- eslapion
- ultimate expander
- Posts: 5458
- Joined: Fri Jun 23, 2006 7:50 pm
- Location: Canada
- Occupation: 8bit addict
Re: 3D Printed ICs
What about this company ?
https://www.mosis.com/
These are the process they offer:
https://www.mosis.com/products
Any process in there compatible with good old NMOS or HMOS ?
About 2.5 million VIC-20s were made and sold between 1981 and 1985 and these were PAL or NTSC. Yet if you look at the wide variety of available 6560 and 6561 chips, you can find a chip with a datecode for almost every single week in that time span.
If I follow your logic, Commodore made only 2 batches of 6560/6561 - one PAL and one NTSC then sold the whole thing over a period of 4 years. I don't think so.
Also, in the early 80s, Commodore needed a relatively wide variety of custom ICs, such as VIC-I (PAL and NTSC), VIC-II (PAL and NTSC), SID, CIAs, VIAs, ROMs (for the VIC-20, C64, 1541 and all sorts of game cartridges), 6502, 6510 and the custom chips for Plus/4 and C16/116 of which they needed a continuous supply.
Making chips in huge batches of 1 million at a time would have choked production and considerably hindered the ability of the company to adapt to market changes. Batches of a specific type of chip around 10k units are much more likely.
https://www.mosis.com/
These are the process they offer:
https://www.mosis.com/products
Any process in there compatible with good old NMOS or HMOS ?
I find that difficult to believe.traymond20 wrote:Commodore always made about 1 million chips in a run so the cost was 1 million dollars (U.S.)
About 2.5 million VIC-20s were made and sold between 1981 and 1985 and these were PAL or NTSC. Yet if you look at the wide variety of available 6560 and 6561 chips, you can find a chip with a datecode for almost every single week in that time span.
If I follow your logic, Commodore made only 2 batches of 6560/6561 - one PAL and one NTSC then sold the whole thing over a period of 4 years. I don't think so.
Also, in the early 80s, Commodore needed a relatively wide variety of custom ICs, such as VIC-I (PAL and NTSC), VIC-II (PAL and NTSC), SID, CIAs, VIAs, ROMs (for the VIC-20, C64, 1541 and all sorts of game cartridges), 6502, 6510 and the custom chips for Plus/4 and C16/116 of which they needed a continuous supply.
Making chips in huge batches of 1 million at a time would have choked production and considerably hindered the ability of the company to adapt to market changes. Batches of a specific type of chip around 10k units are much more likely.
Be normal.