CA3080 being made again?

[StephenGiles]

http://www.rocelec.com/search/build/CA3080/ISL/1/contains/

[armdnrdy]

It seems so.

I posted the same link about a year ago.

If Rochester is producing the 3080s again....why hasn't the cost gone down? They are still priced like a rare NOS IC.

This might be a question for Steve Daniels. $4.95 @ Smallbear?

[merlinb]

Just get over it and buy LM13700s like normal people. 

[R.G.]

If Rochester is producing the 3080s again....why hasn't the cost gone down? They are still priced like a rare NOS IC.

When you don't understand what is happening, it's a reliable indicator that you don't know the rules of the game.

I suspect that Rochester, in common with other replacement/boutique semiconductor places is making 3080s *because* they can price them like a rare NOS IC.  There is a semiconductor sub-industry that buys the old equipment from the mainstream semiconductor guys and uses it to make high-priced replacements for outfits that have to have that one device for something. They make low quantities, and price them high. I first ran into this when I was researching germanium transistors back in the 80s and 90s. Germanium Power Devices was one outfit I talked to. They could make you a device that met datasheet specs for nearly any germanium transistor that had ever existed if you wanted to pay them to do it.

So there exists a boutique semiconductor industry too.

The only reason that normal semiconductors are cheap is that the industry has bootstrapped itself up to do incredibly complex, high-tech stuff in huge quantities. When they were first introduced, individual transistors were individually more expensive than tubes. At the time, tubes were a high-volume industry. The situation is reversed now, and transistors are incredibly cheap, but tubes are expensive. Back in the 1960s, individual transistors cost several dollars each. In 1960 dollars. That's the rough equivalent of several *tens* of dollars each in what passes for a dollar today.

[armdnrdy]

A bit more research has revealed that Rochester purchased Intersils stock of 3080 wafers and dies.
This means that when Rochester's "materials" for the 3080 run out....that's it....no more produced.

So in essence...this is a limited supply run...which would justify the higher (than I would like to pay) price.

[R.G.]

That's even easier than buying and running older semiconductor fab stuff! 

[alanp]

What exactly are "dies" in the semiconductor manufactory context? Are these the tooling necessary to create wafers, given the correct machinery to put them in?

[armdnrdy]

simple explanation here:

http://en.wikipedia.org/wiki/Die_(integrated_circuit)

[amz-fx]

A bit more research has revealed that Rochester purchased Intersils stock of 3080 wafers and dies.

That is certainly true, as Rochester has verified, but I know someone that buys 3080s from them and they were told that Rochester would be making new CA3080s when the wafer stock was exhausted. This seem to be confirmed by their web page that also says they will be manufacturing new LM308 chips:

http://www.rocelec.com/blog/?p=1164

and http://www.rocelec.com/obsolete-semiconductor-manufacturing/

FWIW, the majority of the LM308 chips that you see on Ebay are probably fakes...  I know the ones that I recently bought from a Shanghai supplier were not originals as marked.

Best regards, Jack

[Seljer]

simple explanation here:

http://en.wikipedia.org/wiki/Die_(integrated_circuit)

so basically they've got the finished silicon and just need to add the legs and epoxy then stamp the name on top of it?

[armdnrdy]

simple explanation here:

http://en.wikipedia.org/wiki/Die_(integrated_circuit)

so basically they've got the finished silicon and just need to add the legs and epoxy then stamp the name on top of it?

Yep.
Package it up and send it out the door.

[armdnrdy]

Thanks for the info Jack.

So this means that the price will drop?  

I think not!

There are way too many children to put through college!  

On another note:

I know this question has been asked....and answers have been given but....after a hour long search through the net, I still don't feel like I have a definitive answer. (I've seen conflicting information)

When using a LM13700 in place of a CA3080....what do we do with the other half? Treat it as you would a regular op amp?

Bias voltage on the non inverting input and connect the inverting input to the output.

The amplifier bias input pin is throwing me off.

[MrStab]

Just get over it and buy LM13700s like normal people.  

i'm with Merlin - lower noise & other spec reasons aside, everyone's using the CA3080 and it's cool to be different! lol

i guess an 8-pin version would appease a lot of people, though

[midwayfair]

When using a LM13700 in place of a CA3080....what do we do with the other half? Treat it as you would a regular op amp?

Yes, it's best to "ground out" the unused portion just like a regular op amp.

An 8-pin 13700 would be nice, but it's not like it's that big of a chip.

[MrStab]

i was actually contemplating grounding out the 2nd half on an Engineer's Thumb i made the other week, as per your posts, Jon (you seem to be the leading authority on it after Merlin!), but my own build not being considerably noisy and overall laziness prevented me. yknow, maybe that's why i had ticking in bypass and had to ground the input... that's one for another thread.

[armdnrdy]

From what I've read on the TI site, the way that I outlined above, is the correct way to "keep" an unused op amp in a stable state.
Grounding the inputs saturates the output.
On the TI site there is a video tutorial where an engineer goes through the different "common" ways that unused op amps are terminated. He explains (in almost too much detail) the effect of the commonly used "wrong" ways.



The reason I ask is:
I'm working on a few versions of plug and play daughter boards for a 13700 to 3080 conversion.

I route my own boards...but I would guess that as high as 90% of the forum members don't.

The only daughter board that I've seen has to be wired with jumpers into the 3080 circuit.

[MrStab]

FWIW, Larry, one of the reasons besides laziness that i didn't bother grounding the unused halves of my LM13700s is because i read a coupla posts kinda dismissing it as helpful in lowering noise, but not compulsory. that said, i did have a bypass ticking issue (presumably current related), which may or not have been caused by not grounding. it's my understanding to ground inputs but not outputs for the most part, though. i'm sure i read that somewhere.

[armdnrdy]

As I stated above...I've read different ways to do it but IC manufacturers specify a different way.

Here's the video I was referring to:

[smallbearelec]

So in essence...this is a limited supply run...which would justify the higher (than I would like to pay) price.

Yea. There is enough of a market for repairs of the Dynacomp, etc., and hobbyists who want to do clones to justify investing in the part and stocking it. Clearly, not something you'd use in new designs. BTW I am attempting to get for re-sale the Thomas Henry booklet on the CA3080, but the publisher has not yet returned my message.

So this means that the price will drop?  
I think not! There are way too many children to put through college!  

No kids of our own, but several among the Cave-dwellers who pack orders.

Similar message re the LM308: The good news is that reliable stock will be available for some years, though a little pricey. Also, I have asked Rochester if they can re-make the CA3094E (Small Stone, etc.) and the LM566 VCO. I would love to sell that last one with a copy of the Thomas Henry booklet, but NOS chips are ridiculously expensive. I understand that they are popular for restoring the sound section of old Space Invaders arcade games, so tough to pry loose from the broker market.

[MrStab]

was just offering my 2 cents, which is probably about to be overturned by the video. cheers for that - couldn't seem to find it.

Edit:- lol, i'm shown to have been doing it wrong by the very first example.
now i think about it, i can't have read your post properly at all, Larry - my apologies.