Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Deja-vu redesign task
42 views
Skip to first unread message
pedro
Oct 8, 2012, 12:16:47 AM10/8/12
to
I am likely to be tasked with "recreating" a board/design from the
mid-70's which (among other considerations) used some long-obsolete
Fairchild 9000 series chips. At the time, these were chosen for their
speed and functionality.
This board is part of a multi-board setup, and as such retaining the
original "block diagram functionality" is essential.
95H28 - Dual D flipflop that happily clocked at 250 MHz when
configured as two cascaded /2 stages. Anything dual-D - preferably
DIP - that goes that fast and is readily available would suffice.
9504 - a quad triple-input OR gate in a DIP-16, with one input per
gate commoned as an enable. Followed the above, max Fin was 70 MHz in
the application. As all pins are employed in the existing design,
finding a single-chip replacement block is proving a challenge.
Also used the Motorola MC10216 which I can't find "out there". (The
application ran faster than the 10116). The role is a limiting amp
with inputs to 250 MHz.
There was also an 11C90 or 95H90 as a high-speed (for the day) /10
prescaler stage. The biquinary nature of the division was not a
requirement. While I have seen a lot of DIP-8 and SMD prescalers,
they seem to all be either powers of 2 or dual modulus for
synthesisers, and again /10 is not apparent.
I would welcome suggestions for functional replacements. If responses
are along the usual lines of "use a PIC", "start with a fresh slate"
or "give up" the please don't waste my time and yours.
mid-70's which (among other considerations) used some long-obsolete
Fairchild 9000 series chips. At the time, these were chosen for their
speed and functionality.
This board is part of a multi-board setup, and as such retaining the
original "block diagram functionality" is essential.
95H28 - Dual D flipflop that happily clocked at 250 MHz when
configured as two cascaded /2 stages. Anything dual-D - preferably
DIP - that goes that fast and is readily available would suffice.
9504 - a quad triple-input OR gate in a DIP-16, with one input per
gate commoned as an enable. Followed the above, max Fin was 70 MHz in
the application. As all pins are employed in the existing design,
finding a single-chip replacement block is proving a challenge.
Also used the Motorola MC10216 which I can't find "out there". (The
application ran faster than the 10116). The role is a limiting amp
with inputs to 250 MHz.
There was also an 11C90 or 95H90 as a high-speed (for the day) /10
prescaler stage. The biquinary nature of the division was not a
requirement. While I have seen a lot of DIP-8 and SMD prescalers,
they seem to all be either powers of 2 or dual modulus for
synthesisers, and again /10 is not apparent.
I would welcome suggestions for functional replacements. If responses
are along the usual lines of "use a PIC", "start with a fresh slate"
or "give up" the please don't waste my time and yours.
Bill Sloman
Oct 8, 2012, 5:51:29 AM10/8/12
to
periphery. ECLinPS only comes in surface mount packages, and these
take up a lot less board space than through-the-board TTL. The ECLinPS
100E116 is probably quite a bit faster than the MC10216 - the 216 line-
driver/line receiver was was a bit faster than the 116, but ECLinPs
was about a factor of four faster than the 10k ECL series.
ECLinPS has perfectly fine multipurpose counter, which goes to 550MHz
http://www.onsemi.com/pub_link/Collateral/MC10E136-D.PDF
Similar - if rather more specialised - EL and EP parts are faster, but
there's no EL or EP 136
http://www.onsemi.com/PowerSolutions/parametrics.do?id=400
Or do everything in a single big, fast programmable logic device -
which will run at a lower supply voltage - and deal with the voltage
compatibility with buffers. The speeds mentioned are easily
attainable inside today's programmable logic.
This may strike you as "starting with a fresh slate" but you should be
able to reproduce the functionality without much intellectual effort,
and with very high confidence that what you put together can be
plugged into the original system and will "just work".
There is the risk that the modern parts will react to glitches that
the older hardware ignored, but a little judicious filtering on the
inputs would fix that - you might not even have to load the filter
components but it would be wise to add pads to accommodate them if it
turned out that you needed them.
--
Bill Sloman, Nijmegen
k...@att.bizzzzzzzzzzzz
Oct 8, 2012, 9:37:44 AM10/8/12
to
John Larkin
Oct 8, 2012, 10:02:29 AM10/8/12
to
possibly some level translation. You'd have to program it, but the
tools would be free. Modern CPLDs are planty fast enough.
--
John Larkin Highland Technology Inc
www.highlandtechnology.com jlarkin at highlandtechnology dot com
Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom timing and laser controllers
Photonics and fiberoptic TTL data links
VME analog, thermocouple, LVDT, synchro, tachometer
Multichannel arbitrary waveform generators
dagmarg...@yahoo.com
Oct 8, 2012, 12:15:09 PM10/8/12
to
On Oct 8, 12:16 am, pedro <inas...@work.net> wrote:
> I am likely to be tasked with "recreating" a board/design from the
> mid-70's which (among other considerations) used some long-obsolete
> Fairchild 9000 series chips. At the time, these were chosen for their
> speed and functionality.
>
> This board is part of a multi-board setup, and as such retaining the
> original "block diagram functionality" is essential.
>
> 95H28 - Dual D flipflop that happily clocked at 250 MHz when
> configured as two cascaded /2 stages. Anything dual-D - preferably
> DIP - that goes that fast and is readily available would suffice.
>
> 9504 - a quad triple-input OR gate in a DIP-16, with one input per
> gate commoned as an enable. Followed the above, max Fin was 70 MHz in
> the application. As all pins are employed in the existing design,
> finding a single-chip replacement block is proving a challenge.
>
> Also used the Motorola MC10216 which I can't find "out there". (The
> application ran faster than the 10116). The role is a limiting amp
> with inputs to 250 MHz.
LDVT receivers -- John Larkin uses them as super fast comparators.
> I am likely to be tasked with "recreating" a board/design from the
> mid-70's which (among other considerations) used some long-obsolete
> Fairchild 9000 series chips. At the time, these were chosen for their
> speed and functionality.
>
> This board is part of a multi-board setup, and as such retaining the
> original "block diagram functionality" is essential.
>
> 95H28 - Dual D flipflop that happily clocked at 250 MHz when
> configured as two cascaded /2 stages. Anything dual-D - preferably
> DIP - that goes that fast and is readily available would suffice.
>
> 9504 - a quad triple-input OR gate in a DIP-16, with one input per
> gate commoned as an enable. Followed the above, max Fin was 70 MHz in
> the application. As all pins are employed in the existing design,
> finding a single-chip replacement block is proving a challenge.
>
> Also used the Motorola MC10216 which I can't find "out there". (The
> application ran faster than the 10116). The role is a limiting amp
> with inputs to 250 MHz.
> There was also an 11C90 or 95H90 as a high-speed (for the day) /10
> prescaler stage. The biquinary nature of the division was not a
> requirement. While I have seen a lot of DIP-8 and SMD prescalers,
> they seem to all be either powers of 2 or dual modulus for
> synthesisers, and again /10 is not apparent.
>
> I would welcome suggestions for functional replacements. If responses
> are along the usual lines of "use a PIC", "start with a fresh slate"
> or "give up" the please don't waste my time and yours.
replacement is the goal, searching DigiKey for "ecl" and "in stock"
gives a plethora of choices. 10E, 10EP, 10EL, 100EL, etc.
For example, http://www.onsemi.com/pub_link/Collateral/MC10EL51-D.PDF
single D-flop, 2.8Ghz f(toggle), t(prop) < 620pS.
--
Cheers,
James Arthur
0 new messages