ns pulse generator IC for low voltages

[O. P. E.]

I am trying to make a pulse generator that will generate pulses as
short as 1 ns, with voltages around 1V. It would be triggered by an
external signal. I am wondering whether anyone makes a single chip
solution for this type of thing, or if someone knows a communications
IC that could be easily modified to do this? And it would be great if
it came with an evaluation board, though I realize I am reaching on
that point. (Still, unless you enjoy torture, soldering surface mount
components is something I like to avoid.)

As I've searched on the net, most DIY pulse generator circuits involve
avalanche circuits which seem to require relatively high voltages and
are a little unstable. Linear has a circuit that uses comparators
with an AND gate to produce a pulse, and that seems to be pretty
workable. I've also thought it wouldn't be that hard to combine an
inverter with an AND gate and a delay line IC to produce an output
pulse, provided the transitions didn't excessively widen the pulse.

 

[John McMillan]

Most logic family edge times are greater than 1ns so aren't going to
work. You might look at ECL - there's a hint here
http://www.elecdesign.com/Articles/Index.cfm?AD=1&ArticleID=3647
There's an article "A recipe for homebrew ECL" by Chuck Hastings
which is a good read, though I couldn't find it on the web.
The avalanche pulser described in the appendix to Linear
application note 79 might be what you want.
http://www.linear.com/pc/downloadDocument.do?navId=H0,C3,P1012,D4168

I am trying to make a pulse generator that will generate pulses as
short as 1 ns, with voltages around 1V. It would be triggered by an
external signal. I am wondering whether anyone makes a single chip
solution for this type of thing, or if someone knows a communications
IC that could be easily modified to do this? And it would be great if
it came with an evaluation board, though I realize I am reaching on
that point. (Still, unless you enjoy torture, soldering surface mount
components is something I like to avoid.)

As I've searched on the net, most DIY pulse generator circuits involve
avalanche circuits which seem to require relatively high voltages and
are a little unstable. Linear has a circuit that uses comparators
with an AND gate to produce a pulse, and that seems to be pretty
workable. I've also thought it wouldn't be that hard to combine an
inverter with an AND gate and a delay line IC to produce an output
pulse, provided the transitions didn't excessively widen the pulse.

 

[O. P. E.]

On Mar 26, 5:51 am, John McMillan <j.e.mcmil...@sheffielddot.acdot.uk>
wrote:

Most logic family edge times are greater than 1ns so aren't going to
work. You might look at ECL - there's a hint herehttp://www.elecdesign.com/Articles/Index.cfm?AD=1&ArticleID=3647
There's an article "A recipe for homebrew ECL" by Chuck Hastings
which is a good read, though I couldn't find it on the web.
The avalanche pulser described in the appendix to Linear
application note 79 might be what you want.http://www.linear.com/pc/downloadDocument.do?navId=H0,C3,P1012,D4168

I appreciate the link to the Linear App Note, as it also had many
details on circuit construction that should really help. I also found
a few other Linear app notes written by Jim Williams. I love how he
has "cartoons" on the last page.

In searching around after I posted here I found Potato Semiconductor
advertising "74 series GHz logic." I've never heard of them before,
but the chips look like they would have the fast response I am looking
for. Does anyone have any experience with them? I also found some
fast chips from OnSemi.

Thanks

 

[Lostgallifreyan]

"O. P. E." <mleigh@gmail.com> wrote in
news:1175018589.604889.292110@r56g2000hsd.googlegroups.com:

In searching around after I posted here I found Potato Semiconductor
advertising "74 series GHz logic." I've never heard of them before,
but the chips look like they would have the fast response I am looking
for. Does anyone have any experience with them? I also found some
fast chips from OnSemi.

Potato chips? Isn't this a few days early?

 

[John McMillan]

In article <1175018589.604889.292110@r56g2000hsd.googlegroups.com>,
"O. P. E." <mleigh@gmail.com> wrote:

On Mar 26, 5:51 am, John McMillan <j.e.mcmil...@sheffielddot.acdot.uk
wrote:
Most logic family edge times are greater than 1ns so aren't going to
work. You might look at ECL - there's a hint
herehttp://www.elecdesign.com/Articles/Index.cfm?AD=1&ArticleID=3647
There's an article "A recipe for homebrew ECL" by Chuck Hastings
which is a good read, though I couldn't find it on the web.
The avalanche pulser described in the appendix to Linear
application note 79 might be what you
want.http://www.linear.com/pc/downloadDocument.do?navId=H0,C3,P1012,D4168


I appreciate the link to the Linear App Note, as it also had many
details on circuit construction that should really help. I also found
a few other Linear app notes written by Jim Williams. I love how he
has "cartoons" on the last page.

If you are going into high speed pulse electronics, you could do
worse than read *everything* that Jim Williams has written.

High-Speed Digital Design: A Handbook of Black Magic
http://www.amazon.com/High-Speed-Digital-Design-Handbook-Black/dp/0133957
241
is good too.

In searching around after I posted here I found Potato Semiconductor
advertising "74 series GHz logic." I've never heard of them before,
but the chips look like they would have the fast response I am looking
for. Does anyone have any experience with them?

I'd never heard of them. Must admit I'd taken my eyes off this field
a bit. The data sheets look great. Has anyone actually used the chips?

I also found some
fast chips from OnSemi.

OnSemi were Motorola who invented ECL (or was it Plessey? or ICL? ...)
Their stuff is good.

 

[O. P. E.]

On Mar 26, 5:51 am, John McMillan <j.e.mcmil...@sheffielddot.acdot.uk>
wrote:

Most logic family edge times are greater than 1ns so aren't going to
work. You might look at ECL - there's a hint herehttp://www.elecdesign.com/Articles/Index.cfm?AD=1&ArticleID=3647
There's an article "A recipe for homebrew ECL" by Chuck Hastings
which is a good read, though I couldn't find it on the web.
The avalanche pulser described in the appendix to Linear
application note 79 might be what you want.http://www.linear.com/pc/downloadDocument.do?navId=H0,C3,P1012,D4168

This app note and a few other app notes and web sites discuss making a
pulse generator with a 2N2369 transistor. It looks like they are
mostly referring to the TO-92 package instead of the surface mount. I
notice that the surface mount usually have a lower power dissipation,
but also would seem to have better RF characteristics due to the
shorter leads. Does anyone have any experience with surface mount vs.
TO-92 for avalanche pulse generation? While I don't like the idea of
soldering surface mount components, I'll do it if I have to.

Thanks

 

[John McMillan]

In article <1175531755.402669.221410@p77g2000hsh.googlegroups.com>,
"O. P. E." <mleigh@gmail.com> wrote:

On Mar 26, 5:51 am, John McMillan <j.e.mcmil...@sheffielddot.acdot.uk
wrote:
Most logic family edge times are greater than 1ns so aren't going to
work. You might look at ECL - there's a hint
herehttp://www.elecdesign.com/Articles/Index.cfm?AD=1&ArticleID=3647
There's an article "A recipe for homebrew ECL" by Chuck Hastings
which is a good read, though I couldn't find it on the web.
The avalanche pulser described in the appendix to Linear
application note 79 might be what you
want.http://www.linear.com/pc/downloadDocument.do?navId=H0,C3,P1012,D4168


This app note and a few other app notes and web sites discuss making a
pulse generator with a 2N2369 transistor. It looks like they are
mostly referring to the TO-92 package instead of the surface mount. I
notice that the surface mount usually have a lower power dissipation,
but also would seem to have better RF characteristics due to the
shorter leads. Does anyone have any experience with surface mount vs.
TO-92 for avalanche pulse generation? While I don't like the idea of
soldering surface mount components, I'll do it if I have to.

Thanks

What little experience I do have confirmed the standard RF maxim
"long legs bad, short legs good". I have slightly more experience
with hand soldering surface mount transistors (and other two and three
legged components). Its not as difficult as it looks. After
a day or two you've got completely used to it.