Author |
Message |
Paul Taylor electronics forum beginner
Joined: 10 Feb 2005
Posts: 45
|
Posted: Thu Jul 20, 2006 9:26 am Post subject:
Re: Count 1ns Pulse
|
|
|
"Paul Taylor" <paul.taylor@uea.ac.uk> wrote in message
news:e9nf10$qrt$1@cpca14.uea.ac.uk...
Quote: | So basically According to the electron tubes data sheet for the PAD the
pulse pair resolution is 24ns and the output pulse rise and fall time is
2ns so say a 5ns pulse will be seen on the output of the pad every 24ns so
the total period is 29ns (frequency 34Mhz) so in effect the 74hc4520
maximum frequency isn't a problem only the pulse width it will read? so in
a nutshell i need something to read a 5ns pulse any ideas.
Thanks
Update - Been in contact with electron tubes and actually spoke to the guy |
who built the pad the pulse pair resolution is actually define as 1 cycle
i.e. 24ns from the rising edge of the first pulse to the rising edge of the
second one? so im still faced with finding a counter to count 5ns pulse.
The other idea i have thought of now is to stretch the 5ns pulse to 10ns
then the counter will be able to read the pulse this shouldn't effect
anything. |
|
Back to top |
|
 |
David L. Jones electronics forum Guru Wannabe
Joined: 07 Jun 2005
Posts: 286
|
Posted: Thu Jul 20, 2006 9:51 am Post subject:
Re: Count 1ns pulse
|
|
|
Paul Taylor wrote:
Quote: | "David L. Jones" <altzone@gmail.com> wrote in message
news:1153364491.069220.250950@h48g2000cwc.googlegroups.com...
Paul Taylor wrote:
Hi all i still haven't solved the counter problem here is another problem
link to the first problem - i know have a pulse generator with which im
generating a 1ns pulse every 5ns so i want to take this and count the
total
number of pulse in 1ms this according to my calculator totals 200,000, Im
using two 74HC4520 Dual 4-bit synchronous binary counter to give me the
16
bits required, i don't seem to be getting the correct output reading,
why?
As others have said, the 74HC4520 is wayyyyy too slow for this
application, as will be every other 74xxxxxxx series device.
Just how are you generating this 1ns pulse?
Dave :)
Hi David im using a HP 8011A Pulse generator.
Thanks
|
How are you generating a 1ns pulse with a generator that is only
capable of generating a 25ns pulse at best?
http://www.amplifierrentals.com/Equipment/HP/8011A.htm
Dave  |
|
Back to top |
|
 |
Phil Allison electronics forum Guru
Joined: 20 May 2005
Posts: 1368
|
Posted: Thu Jul 20, 2006 10:03 am Post subject:
Re: Count 1ns pulse
|
|
|
"David L. Jones"
Paul Taylor wrote:
** You are wasting your time on this fuckwit, fake "Rocket Scientist".
He is taking YOU for a big ride - all the way down.
......... Phil |
|
Back to top |
|
 |
Paul Taylor electronics forum beginner
Joined: 10 Feb 2005
Posts: 45
|
Posted: Thu Jul 20, 2006 10:09 am Post subject:
Re: Count 1ns pulse
|
|
|
"David L. Jones" <altzone@gmail.com> wrote in message
Derhhh sorry Dave you are right i misread the scope it is in fact 1usec
timebase not as i previously thought 1ns must get my eyes tested. |
|
Back to top |
|
 |
David L. Jones electronics forum Guru Wannabe
Joined: 07 Jun 2005
Posts: 286
|
Posted: Thu Jul 20, 2006 10:30 am Post subject:
Re: Count 1ns pulse
|
|
|
Paul Taylor wrote:
Quote: | "David L. Jones" <altzone@gmail.com> wrote in message
How are you generating a 1ns pulse with a generator that is only
capable of generating a 25ns pulse at best?
http://www.amplifierrentals.com/Equipment/HP/8011A.htm
Dave :)
Derhhh sorry Dave you are right i misread the scope it is in fact 1usec
timebase not as i previously thought 1ns must get my eyes tested.
|
Thread closed!
Dave  |
|
Back to top |
|
 |
Phil Allison electronics forum Guru
Joined: 20 May 2005
Posts: 1368
|
Posted: Thu Jul 20, 2006 10:41 am Post subject:
Re: Count 1ns pulse
|
|
|
"David L. Jones"
Quote: | Paul Taylor wrote:
Derhhh sorry Dave you are right i misread the scope it is in fact 1usec
timebase not as i previously thought 1ns must get my eyes tested.
Thread closed!
Dave
|
** That penny just made ONE HELL of a clang !!!!
......... Phil |
|
Back to top |
|
 |
Paul Taylor electronics forum beginner
Joined: 10 Feb 2005
Posts: 45
|
Posted: Thu Jul 20, 2006 11:16 am Post subject:
Re: Count 1ns pulse
|
|
|
"David L. Jones" <altzone@gmail.com> wrote in message
news:1153391411.204205.200970@m79g2000cwm.googlegroups.com...
Quote: | Paul Taylor wrote:
"David L. Jones" <altzone@gmail.com> wrote in message
How are you generating a 1ns pulse with a generator that is only
capable of generating a 25ns pulse at best?
http://www.amplifierrentals.com/Equipment/HP/8011A.htm
Dave :)
Derhhh sorry Dave you are right i misread the scope it is in fact 1usec
timebase not as i previously thought 1ns must get my eyes tested.
Thread closed!
Dave :)
|
I have know got a BNC model 555 pulse delay generator and a more powerful
scope LECROY 9361C which is more than capable, im going to stretch the 5ns
TTL pulse from the pad to 10ns and see if the counter will except this? |
|
Back to top |
|
 |
Bill Sloman electronics forum Guru
Joined: 12 May 2005
Posts: 1080
|
Posted: Thu Jul 20, 2006 1:15 pm Post subject:
Re: Count 1ns pulse
|
|
|
Paul Taylor wrote:
Quote: | bill.sloman@ieee.org> wrote in message
news:1153355028.067672.238690@m73g2000cwd.googlegroups.com...
Paul Taylor wrote:
Hi all i still haven't solved the counter problem here is another problem
link to the first problem - i know have a pulse generator with which im
generating a 1ns pulse every 5ns so i want to take this and count the
total
number of pulse in 1ms this according to my calculator totals 200,000, Im
using two 74HC4520 Dual 4-bit synchronous binary counter to give me the
16
bits required, i don't seem to be getting the correct output reading,
why?
The 74HC4520 isn't fast enough.
Check out the 74AC163 or 74ACT163 - even they need about 3.5nsec wide
clock pusles before they will count, but they can be strung together -
see figure 2 in the data sheet - to manage the 50MHz you will need,
though not the 200MHz you are generating with your pulse generator.
--
Bill Sloman, Nijmegen
I haven't got to hand any 74AC163, i will have to order some in and give
them a go, The data sheet for the 74hc4520 says that it can read typically a
clock pulse width of 10ns? and a maximum clock pulse frequency of 58Mhz, i
agree that this wont work with my pulse generator set up.
|
"Typical "means that half the IC's the manufacturer produces or - more
precisely - half thebatches of chips that the manufacturer produces
(since within-batch variation is a lot smaller) won't count reliably at
58MHz.
Use the worst case figures - here 30MHz and 20nsec.
Bear in mind that when you string 74HC4520 synchronous counters
together, they lack the terminal carry output and the count-enable
inputs available on the 74ACT163 that make it easy to create a wider
synchronous counter - see figure 2 in the 163 data sheet.
When working out how fast a wider synchronous counter can go, you have
to ignore the maximum clock frequency quoted for a single counter
(95MHz), and add the worst case (longest) propagation delay from the
the clock to the terminal carry output (11.5nsec with a 5V rail) to the
recommended minimum set-up time required for the clock enable inputs
before it can be relied on to handle an incominng clock edge correctly
(5nsec with a 5V rail), for a total of 16.5nsec, corrrsponding to a
maximum count frequency of 60MHz.
--
Bill Sloman, Nijmegen |
|
Back to top |
|
 |
Bill Sloman electronics forum Guru
Joined: 12 May 2005
Posts: 1080
|
Posted: Thu Jul 20, 2006 1:28 pm Post subject:
Re: Count 1ns Pulse
|
|
|
"Paul Taylor" <paul.taylor@uea.ac.uk> schreef in bericht
news:e9ni87$t30$1@cpca14.uea.ac.uk...
Quote: |
"Paul Taylor" <paul.taylor@uea.ac.uk> wrote in message
news:e9nf10$qrt$1@cpca14.uea.ac.uk...
So basically According to the electron tubes data sheet for the PAD the
pulse pair resolution is 24ns and the output pulse rise and fall time is
2ns so say a 5ns pulse will be seen on the output of the pad every 24ns
so the total period is 29ns (frequency 34Mhz) so in effect the 74hc4520
maximum frequency isn't a problem only the pulse width it will read? so
in a nutshell i need something to read a 5ns pulse any ideas.
Thanks
Update - Been in contact with electron tubes and actually spoke to the guy
who built the pad the pulse pair resolution is actually define as 1 cycle
i.e. 24ns from the rising edge of the first pulse to the rising edge of
the second one? so im still faced with finding a counter to count 5ns
pulse.
The other idea i have thought of now is to stretch the 5ns pulse to 10ns
then the counter will be able to read the pulse this shouldn't effect
anything.
|
Why do you think that your minimum pulse width is 5nsec?
And why would you want to stretch it?
The 74ACT163 has a minimum clock pulse width specification of 3.5nsec with a
5V rail, 4,5nsec with a 3.3V rail, and will count at 41MHz even when
assembled into a 12- or 16-bit wide synchronous counter.
--
Bill Sloman, Nijmegen |
|
Back to top |
|
 |
Paul Taylor electronics forum beginner
Joined: 10 Feb 2005
Posts: 45
|
Posted: Thu Jul 20, 2006 1:39 pm Post subject:
Re: Count 1ns pulse
|
|
|
<bill.sloman@ieee.org> wrote in message
news:1153401314.767206.225710@b28g2000cwb.googlegroups.com...
Quote: |
Paul Taylor wrote:
bill.sloman@ieee.org> wrote in message
"Typical "means that half the IC's the manufacturer produces or - more
precisely - half thebatches of chips that the manufacturer produces
(since within-batch variation is a lot smaller) won't count reliably at
58MHz.
Use the worst case figures - here 30MHz and 20nsec.
Bear in mind that when you string 74HC4520 synchronous counters
together, they lack the terminal carry output and the count-enable
inputs available on the 74ACT163 that make it easy to create a wider
synchronous counter - see figure 2 in the 163 data sheet.
When working out how fast a wider synchronous counter can go, you have
to ignore the maximum clock frequency quoted for a single counter
(95MHz), and add the worst case (longest) propagation delay from the
the clock to the terminal carry output (11.5nsec with a 5V rail) to the
recommended minimum set-up time required for the clock enable inputs
before it can be relied on to handle an incominng clock edge correctly
(5nsec with a 5V rail), for a total of 16.5nsec, corrrsponding to a
maximum count frequency of 60MHz.
--
Bill Sloman, Nijmegen
I see that now bill shame they don't do a dual package version of the |
74act163? as I need to keep the board as small as possible. |
|
Back to top |
|
 |
John Fields electronics forum Guru
Joined: 24 Mar 2005
Posts: 3260
|
Posted: Thu Jul 20, 2006 2:20 pm Post subject:
Re: Count 1ns Pulse
|
|
|
On Thu, 20 Jul 2006 10:26:24 +0100, "Paul Taylor"
<paul.taylor@uea.ac.uk> wrote:
Quote: |
"Paul Taylor" <paul.taylor@uea.ac.uk> wrote in message
news:e9nf10$qrt$1@cpca14.uea.ac.uk...
So basically According to the electron tubes data sheet for the PAD the
pulse pair resolution is 24ns and the output pulse rise and fall time is
2ns so say a 5ns pulse will be seen on the output of the pad every 24ns so
the total period is 29ns (frequency 34Mhz) so in effect the 74hc4520
maximum frequency isn't a problem only the pulse width it will read? so in
a nutshell i need something to read a 5ns pulse any ideas.
Thanks
Update - Been in contact with electron tubes and actually spoke to the guy
who built the pad the pulse pair resolution is actually define as 1 cycle
i.e. 24ns from the rising edge of the first pulse to the rising edge of the
second one? so im still faced with finding a counter to count 5ns pulse.
The other idea i have thought of now is to stretch the 5ns pulse to 10ns
then the counter will be able to read the pulse this shouldn't effect
anything.
|
---
Why don't you get ahold of the guy again and ask him what the width
of the output pulse is? I'd almost bet that it'll be about half of
the pulse-pair resolution, or about 12ns, which should get you over
the hump.
--
John Fields
Professional Circuit Designer |
|
Back to top |
|
 |
Paul Taylor electronics forum beginner
Joined: 10 Feb 2005
Posts: 45
|
Posted: Thu Jul 20, 2006 2:26 pm Post subject:
Re: Count 1ns Pulse
|
|
|
"John Fields" <jfields@austininstruments.com> wrote in message
Quote: | Why don't you get ahold of the guy again and ask him what the width
of the output pulse is? I'd almost bet that it'll be about half of
the pulse-pair resolution, or about 12ns, which should get you over
the hump.
--
John Fields
Professional Circuit Designer
|
Just emailed him, mind you when i spoke to him on the phone i explained my
problem and mention about stretching the pulses because the counter wasn't
recognising them he said it is feasible as they do this also, he even
emailed the circuit they use. Anyway lets see what the email brings back. |
|
Back to top |
|
 |
Paul Taylor electronics forum beginner
Joined: 10 Feb 2005
Posts: 45
|
Posted: Thu Jul 20, 2006 2:28 pm Post subject:
Re: Count 1ns Pulse
|
|
|
"Bill Sloman" <bill.sloman@ieee.org> wrote in message
news:44bf8a1f$0$2026$ba620dc5@text.nova.planet.nl...
Quote: | Why do you think that your minimum pulse width is 5nsec?
And why would you want to stretch it?
The 74ACT163 has a minimum clock pulse width specification of 3.5nsec with
a 5V rail, 4,5nsec with a 3.3V rail, and will count at 41MHz even when
assembled into a 12- or 16-bit wide synchronous counter.
--
Bill Sloman, Nijmegen
The output from the pad has a 2ns rise and fall time these are exact pulse |
every time i made an assumption that in the worst case scenario if the pulse
stays on for 1ns then that gives me the 5ns pulse?
the reason i was going to stretch the pulse was before you mention about the
74ACT163 which will or should do the trick i haven't got any of these but i
do have a 74F163 which will read a 6ns pulse so for now im going to give
this ago and order in some of the ACT types, Bill im interested in how
exactly you calculate the 41Mhz for future reference.
Thanks |
|
Back to top |
|
 |
Bill Sloman electronics forum Guru
Joined: 12 May 2005
Posts: 1080
|
Posted: Thu Jul 20, 2006 2:31 pm Post subject:
Re: Count 1ns pulse
|
|
|
"Paul Taylor" <paul.taylor@uea.ac.uk> schreef in bericht
news:e9o13d$17r$1@cpca14.uea.ac.uk...
Quote: |
bill.sloman@ieee.org> wrote in message
news:1153401314.767206.225710@b28g2000cwb.googlegroups.com...
Paul Taylor wrote:
bill.sloman@ieee.org> wrote in message
"Typical "means that half the IC's the manufacturer produces or - more
precisely - half thebatches of chips that the manufacturer produces
(since within-batch variation is a lot smaller) won't count reliably at
58MHz.
Use the worst case figures - here 30MHz and 20nsec.
Bear in mind that when you string 74HC4520 synchronous counters
together, they lack the terminal carry output and the count-enable
inputs available on the 74ACT163 that make it easy to create a wider
synchronous counter - see figure 2 in the 163 data sheet.
When working out how fast a wider synchronous counter can go, you have
to ignore the maximum clock frequency quoted for a single counter
(95MHz), and add the worst case (longest) propagation delay from the
the clock to the terminal carry output (11.5nsec with a 5V rail) to the
recommended minimum set-up time required for the clock enable inputs
before it can be relied on to handle an incominng clock edge correctly
(5nsec with a 5V rail), for a total of 16.5nsec, corrrsponding to a
maximum count frequency of 60MHz.
--
Bill Sloman, Nijmegen
I see that now bill shame they don't do a dual package version of the
74act163? as I need to keep the board as small as possible.
|
The 74ACT163 uses too many pins to allow you to fit two into a 16-pin
package.
The 74F579 has tristate outputs, which means that it can fit an 8-bit
counter into a 20-pin pacakage. Not quite as fast as the ACT163, and
probably eats more current, but it might be interesting, if you can find
someone willing to sell you a few. Fairchild will probably send you free
samples if you ask nicely, but maybe not enough.
If board space really is a problem, go for a biggish programmable logic
device. My guess is that the smallest of the CoolRunner 2 IC's, with 32
logic cells could accomodate two 12-bit synchronous counters and most of the
rest of your glue logic in a single 44-pin package
http://direct.xilinx.com/bvdocs/publications/ds090.pdf
http://direct.xilinx.com/bvdocs/packages/pc44.pdf
If you went for the next step up, with 64 logic cells in the same package,
you could probably accomodate all the logic on the board ...
And I'm pretty sure that the parts are fast enough to make your counter ...
--
Bill Sloman, Nijmegen |
|
Back to top |
|
 |
Bill Sloman electronics forum Guru
Joined: 12 May 2005
Posts: 1080
|
Posted: Thu Jul 20, 2006 2:35 pm Post subject:
Re: Count 1ns pulse
|
|
|
"Phil Allison" <philallison@tpg.com.au> schreef in bericht
news:4i8b3hF2kqc2U1@individual.net...
Quote: |
"John Fields"
This is starting to get silly.
** Wake up - it got silly a long time ago.
And, BTW, since you're having so much trouble getting this together
and you've bought the PMT and the PAD, why don't you just buy the
counter module from Electron Tubes and be done with it?
** See the sig file:
Paul Taylor BSC (Hons)
Electronics Technician
School of Environmental Science
University of East Anglia
Norwich
NR4 7TJ
Phone: +44 (0)1603 592502
Fax: +44 (0)1603 591327
** He is a science grad who has faked his way into a job he has no clue
how to do.
|
He isn't the first. It took a while before I started performing to my
potential.
It looks as if he can learn, which is more than can be said for some around
here.
--
Bill Sloman, Nijmegen |
|
Back to top |
|
 |
Google
|
|
Back to top |
|
 |
|