high speed buffers with adjustable thresholds?

[Mike Noone]

Hi - I'm not sure if this is something that exists or not, but I feel
like I should ask anyways. I'm working on building the sensor cables
for a open source/open hardware logic analyzer
(http://minila.sourceforge.net/) I recently built. Specifically, I want
a high speed buffer (100MHz) that has adjustable low and high input
thresholds, and 3.3V output. Does such a beast exist? I would ideally
like to have between 8 and 16 channels per chip.

Thanks!

-Mike

 

[[email protected]]

Hi - I'm not sure if this is something that exists or not, but I feel
like I should ask anyways. I'm working on building the sensor cables
for a open source/open hardware logic analyzer
(http://minila.sourceforge.net/) I recently built. Specifically, I want
a high speed buffer (100MHz) that has adjustable low and high input
thresholds, and 3.3V output. Does such a beast exist? I would ideally
like to have between 8 and 16 channels per chip.

Thanks!

-Mike

http://www.pctestinstruments.com/

You're welcome!

 

[john jardine]

Hi - I'm not sure if this is something that exists or not, but I feel
like I should ask anyways. I'm working on building the sensor cables
for a open source/open hardware logic analyzer
(http://minila.sourceforge.net/) I recently built. Specifically, I want
a high speed buffer (100MHz) that has adjustable low and high input
thresholds, and 3.3V output. Does such a beast exist? I would ideally
like to have between 8 and 16 channels per chip.

Thanks!

-Mike

I'm seeing the word "buffer" but thinking "comparitor". Something like the
nice LMH7220 could be OK. At a push maybe the LMV7219.
Looks like the easy bit is done. The headscratching comes with interfacing
to the real world.
john

 

[[email protected]]

I'm seeing the word "buffer" but thinking "comparitor". Something like the
nice LMH7220 could be OK. At a push maybe the LMV7219.
Looks like the easy bit is done. The headscratching comes with interfacing
to the real world.

Thinking "comparator" generates nearly a hundred times more hits on
Google (67,900 versus 6,120,000).

100MHz would be pushing it for regular comparators with TTL-compatible
outputs, while fast comparators with ECL-compatible outputs can pretty
much all make 100MHz or better - the orgininal Am685 could just do
100MHz and its many sucessors are all faster.

The balanced ECL-compatible outputs are very much easier to keep out of
the sensitive fast inputs on these beasts than the TTL outputs you can
get away with on slower comparators, so they shouldn't be seen as a bug
but rather a feature.

As John Larkin is shortly going to point out, ECL line-receivers and
LVDS receivers are very fast, relatively low-gain comparators,
available in multi-comparator packages intended to hang on the ends of
relatively wide buses.

Here is a data sheet for a dual part

http://cache.national.com/ds/DS/DS90C402.pdf

John actually uses the parts (as opposed to reading the data sheets and
longing for a chance to use them) and will come up with more part
nmbers.

 

[john jardine]

[...]

Thinking "comparator" generates nearly a hundred times more hits on
Google (67,900 versus 6,120,000).

Indeed, this is true. 2 much reading of these damned newsgroups is
imprinting me.
Anyway, I'd have thought the LMH7220 would be ideal in this job. A simple to
use, true compar-a-tor with LVDS out, allowing a decent length of twisted
pair from the test pod. (So nifty I had to buy a sampling scope to confirm
the speed)
john

 

[Mike Noone]

Thinking "comparator" generates nearly a hundred times more hits on
Google (67,900 versus 6,120,000).

100MHz would be pushing it for regular comparators with TTL-compatible
outputs, while fast comparators with ECL-compatible outputs can pretty
much all make 100MHz or better - the orgininal Am685 could just do
100MHz and its many sucessors are all faster.

The balanced ECL-compatible outputs are very much easier to keep out of
the sensitive fast inputs on these beasts than the TTL outputs you can
get away with on slower comparators, so they shouldn't be seen as a bug
but rather a feature.

As John Larkin is shortly going to point out, ECL line-receivers and
LVDS receivers are very fast, relatively low-gain comparators,
available in multi-comparator packages intended to hang on the ends of
relatively wide buses.

Here is a data sheet for a dual part

http://cache.national.com/ds/DS/DS90C402.pdf

John actually uses the parts (as opposed to reading the data sheets and
longing for a chance to use them) and will come up with more part
nmbers.

Hi Bill - thanks for the help. One question though: Are there any parts
with more channels that can operate at these speeds? I was thinking
that they'd be a fairly common item - but so far I can't find any that
can achieve speeds like that. Thanks,

-Mike

 

[Jim Thompson]

Hi Bill - thanks for the help. One question though: Are there any parts
with more channels that can operate at these speeds? I was thinking
that they'd be a fairly common item - but so far I can't find any that
can achieve speeds like that. Thanks,

-Mike

My venerable (1965) MC1650/51 comparator design has a 3.5ns tpd (typ).

You can still buy it, 41 years after I designed it, from Lansdale...

http://www.lansdale.com/part_search.php?search=1650&x=0&y=0

Motorola built a 32-wide version for WPAFB, but it was never
commercially available.

...Jim Thompson

 

[[email protected]]

Hi Bill - thanks for the help. One question though: Are there any parts
with more channels that can operate at these speeds? I was thinking
that they'd be a fairly common item - but so far I can't find any that
can achieve speeds like that. Thanks,

A quick serach of the National website found me two quad line receivers

http://cache.national.com/ds/DS/DS90LV048A.pdf

http://cache.national.com/ds/DS/DS90LV032A.pdf

If you need the gain of a real comparator, you really are restrcted to
parts with balanced ECL-compatible outputs and they come in single
parts and duals. The Am687 comes to mind.

Jim Thompsons MC1650/1 is, of course, older, but the latching
arrangement isn't as nice as the one on the AM685/687, which is why
I've never used it and probably explains why everybody else has
second-sourced and improved the Am parts.

I've also used the Analog Devices AD9685

http://www.analog.com/UploadedFiles/Data_Sheets/AD96685_96687.pdf

Analog has faster comparators

http://www.analog.com/UploadedFiles/Data_Sheets/ADCMP567.pdf

but you probably don't need that much speed.