[sdiy] Wishing for Interdesign.....wish granted?

[Steve Lenham]

> So, given that the 700 library contains a transconductance amplifier, how
> does this device compare in specs to a CA3080 or CA3280?  If a higher
> quality OTA is needed, is it possible to design one in the 700 series?

I don't think there is anything special about the innards of the 3080; in 
fact (according to a Dave Rossum article reprinted in Barry Klein's IC 
databook) some of its shortcomings are due to the poor quality of the PNP 
transistors used. Gains of less than 10 are mentioned - the 700 series PNPs 
are better than that.

The 3280 is somewhat more advanced and, I believe, uses transistor 
interdigitation to improve performance, but at the end of the day is still 
only built out of transistors.

If you want a higher performance OTA then the next step is architectural 
rather than process-related anyway - add a few more transistors and turn it 
into a Gilbert multiplier.

So I can only think of two potential problems with recreating such devices 
with the 700 series:

1. if the transistors are inherently (too) noisy.

2. if, because of the fixed array structure, it is impossible to lay the 
chip out to achieve the required matching.

All that information is in the designer's manual (free download) but I 
personally don't have enough experience to make a judgement. Anyone else 
care to answer?

I would, however, be surprised if either of these did turn out to be a major 
problem. To know how a complete circuit block (e.g. an OTA) performs, we'd 
probably need the simulator software.

As a suggestion, bearing in mind that (approximately):

3080 = standard OTA core
3280 = (almost) standard OTA core + input linearisation
Gilbert cell (e.g. SSM2020) = standard OTA core + input linearisation + 
extra differential pair

...if I were designing a VCA device with the 700 series, I would try and 
make it perform all of these functions, with the ability to bias the extra 
circuitry off if desired.

I still think this is a very exciting possibility!

Cheers,

Steve L.


> This site had some prices which seem to be quite reasonable, when not
> counting the NRE outlay.
>
> http://www.sms-asic.co.uk/studies2.htm
>
> http://www.sms-asic.co.uk/studies.htm
>
> ASIC technique BPID presumably being the 700 series, costs of US$2.19 to
> US$4.19 per part in quantities of 6K to 40K. There are also NRE charges, 
> but
> I assume one would buy the SIMetrix and ICED tools, which is an investment
> in the $5-$12K range.
>
> John Loffink
> The Microtonal Synthesis Web Site
> http://www.microtonal-synthesis.com
> The Wavemakers Synthesizer Web Site
> http://www.wavemakers-synth.com
>
>
>> -----Original Message-----
>> From: owner-synth-diy at dropmix.xs4all.nl [mailto:owner-synth-
>> diy at dropmix.xs4all.nl] On Behalf Of phillip m gallo
>> Sent: Friday, May 27, 2005 9:11 PM
>> To: synth-diy at dropmix.xs4all.nl
>> Subject: RE: [sdiy] Wishing for Interdesign.....wish granted?
>>
>> Steve,
>>
>> Hope your glorious UK day went well in trade for your wonderful posting 
>> of
>> the URL to the "700 Series".  A great ref (incomplete sentence).
>>
>> This is indeed the Interdesign i remember, the PDF book looks to be the
>> same
>> introductory Interdesign information i worked with more than 24 years 
>> ago.
>> In fact the circuit ref's (apostrophe indicating omitted syllables) seem
>> to
>> be the same, if memory serves.
>>
>> It was interesting and also amusing as some months ago i downloaded Han
>> Cammerind's book (apostrophe indicating possession) and wondered about 
>> his
>> claim of designing the NE555, thinking to myself "i thought the guys at
>> Interdesign implemented this chips for Signetics using their process".
>>
>> regards,
>> p
>>
>> (composing in the recently frowned upon top posting format)
>>
>> -----Original Message-----
>> From: owner-synth-diy at dropmix.xs4all.nl
>> [mailto:owner-synth-diy at dropmix.xs4all.nl]On Behalf Of Steve Lenham
>> Sent: Friday, May 27, 2005 12:58 PM
>> To: Paul Schreiber; synth-diy at dropmix.xs4all.nl
>> Subject: Re: [sdiy] Wishing for Interdesign.....wish granted?
>>
>>
>> It's a glorious day here in the southern UK, too hot to work, and this
>> thread jogged a memory...with quite exciting results.
>>
>> I recalled seeing that Zetex (UK maker of semis, mainly discrete and SSI)
>> offered a semicustom chip process akin to that described. And they do:
>>
>> http://www.zetex.com/3.0/a5-3.asp
>>
>> ...the 700 series. It looks like Zetex do the chip fabrication but refer
>> you
>> to "partner companies" for the design and liaison. So, scroll down to the
>> list of partners and we find our friends(?) OnChip Systems, well known in
>> these parts as the modern reincarnation of Curtis Electromusic. Certainly
>> an
>> indication that the process is good for "our kind" of device - perhaps
>> even
>> that used for the more recent CEMs.
>>
>> But there's more.
>>
>> The Zetex page credits design of the 700 series to a company called Array
>> Design, so follow the link and root around a bit to end up at:
>>
>> http://www.arraydesign.com/experience/index.html
>>
>> which reveals that the chap behind Array Design is the original founder 
>> of
>> Interdesign (not to mention also the designer of the NE555 and NE566!).
>>
>> They have, for free download (http://www.arraydesign.com/700series.html),
>> a
>> 192 page manual very similar to that which Paul describes, and the Zetex
>> page even refers to "colour coding of the metal interconnect" during
>> design - not a million miles from the old coloured pencils.
>>
>> I haven't had time to read any further yet, and all the usual worries
>> about
>> price/MOQ/etc. still apply, but let's savour for a moment the idea of
>> being
>> to access the same IC process which brought us CEMs in the first process
>> and
>> which is still used by Curtis himself...
>>
>> If it still looks good in the morning, perhaps a discussion of the best
>> next
>> move might be in order?
>>
>> Have a good weekend,
>>
>> Steve L.
>>
>>
>> ----- Original Message -----
>> From: "Paul Schreiber" <synth1 at airmail.net>
>> To: "karl dalen" <dalenkarl at yahoo.se>; "Colin f" <colin at colinfraser.com>;
>> <synth-diy at dropmix.xs4all.nl>
>> Sent: Friday, May 27, 2005 4:38 PM
>> Subject: [sdiy] Wishing for Interdesign.....
>>
>>
>> > Interdesign was exactly what we need now: a generic "500-in-1" (Radio
>> > Shack reference) array of generic analog "building blocks" on 1 die. 
>> > You
>> > used a 2-color pencil (really!) and drew out the "schematic" on a giant
>> > 36" x 48" master template. Then, they added 2 metal layers and you had 
>> > a
>> > custom IC.
>> >
>> > The down-side was you had to test them yourself ("No Time to Test
>> > These!"...obscure reference no one under 47yrs old will get)
>> >
>> > The cost was low: $750 and that got you like 40 packaged parts! Of 
>> > which
>> > 30 worked :)
>> >
>> > In fact, the early CEM chips (3310/20/30/40) are actually Interdesign
>> > parts! The reason: Doug Curtis, the designer, won a magazine contest
>> with
>> > them and as 1st place got 100pcs free. The rest is history.
>> >
>> > The later parts (3396, etc) are done on a Doug Curtis designed array 
>> > (4X
>> > larger than Interdesign), the fab was AMI. In Doug's office his winning
>> > entry is framed on the wall.
>> >
>> > BTW: Interdesign had GREAT app notes in a 5" ring binder. They may turn
>> up
>> > on eBay/flea markets. About 8 are by Doug on the OTA gain cell (patent
>> has
>> > expired....hmmmmm.......)
>> >
>> > Paul S.
>> >
>> >
>
>
>
>
>
>
>